sensors Article
Plantar Pressure Detection with Fiber Bragg Gratings Sensing System Tsair-Chun Liang 1, *, Jhe-Jhun Lin 1 and Lan-Yuen Guo 2 1 2
*
Graduate Institute of Electrical Engineering, National Kaohsiung First University of Science and Technology, Kaohsiung City 811, Taiwan; [email protected] Department of Sports Medicine, Kaohsiung Medical University, Kaohsiung City 807, Taiwan; [email protected] Correspondence: [email protected]; Tel.: +886-7-6011-000 (ext. 2712)
Academic Editor: Vittorio M. N. Passaro Received: 17 June 2016; Accepted: 19 October 2016; Published: 22 October 2016
Abstract: In this paper, a novel fiber-optic sensing system based on fiber Bragg gratings (FBGs) to measure foot plantar pressure is proposed. This study first explores the Pedar-X insole foot pressure types of the adult-size chart and then defines six measurement areas to effectively identify four foot types: neutral foot, cavus foot, supinated foot and flat foot. The plantar pressure signals are detected by only six FBGs, which are embedded in silicone rubber. The performance of the fiber optic sensing is examined and compared with a digital pressure plate of i-Step P1000 with 1024 barometric sensors. In the experiment, there are 11 participants with different foot types to participate in the test. The Pearson correlation coefficient, which is determined from the measured results of the homemade fiber-optic plantar pressure system and i-Step P1000 plantar pressure plate, reaches up to 0.671 (p < 0.01). According to the measured results from the plantar pressure data, the proposed fiber optic sensing system can successfully identify the four different foot types. Measurements of this study have demonstrated the feasibility of the proposed system so that it can be an alternative for plantar pressure detection systems. Keywords: fiber-optic sensor; fiber Bragg grating; foot plantar pressure; Pearson product-moment correlation coefficient
1. Introduction Not all feet are naturally identical since the human foot includes a lot of bones, joints, muscles, tendons, and ligaments. While the bones, joints, and tendons of a person differ little from that of another, a variety of different foot types come up substantially. In general, the human foot can be classified into five types: flat foot, neutral foot, pronated foot, supinated foot and cavus foot [1–3]. Some classification methods to identify foot type have been used in the literature [1,4,5], covering visual non-quantitative inspection, clinical values, footprint indices, and radiographic evaluation, etc. Figure 1 shows the four types of feet that are studied in this paper. As shown in Figure 1a, the flat foot is a postural deformity, in which the arch of the foot collapses on the entire sole of the foot, having complete or near-complete contact with the ground. This foot type needs extra support in the arch region to maintain a neutral gait and to dispel the shock of running [2,6,7]. If the arch of the foot lightly collapses inward, it can be called a neutral foot. Figure 1b shows the footprint of the neutral (or normal) foot type. The neutral foot is the most common type of foot and it allows the body to naturally absorb shock [2,6,7]. If the ball and heel of the footprint are connected by a thin strip on the outside or no strip at all (see Figure 1c), this is a cavus foot. This foot type needs the most assistance with shock absorption, since the rigid structure is unable to dispel impact forces very well. Sensors 2016, 16, 1766; doi:10.3390/s16101766
www.mdpi.com/journal/sensors
Sensors Sensors 2016, 16, 1766
22 of of 13 13
The pronated foot or eversion foot is the inward roll to distribute the impact force caused by the Thewhile pronated footor orrunning. eversionItfoot theof inward roll tomovements distribute the impact caused bydeal the ground walking is aispart the natural which helpforce the lower leg ground while walking or running. It is a part of the natural movements which help the lower leg with shock. Though this is not bad in itself, it does affect the way you run and may increase the deal with shock. Though thisSupination is not bad in it does affectasthe way you and1d, may increase the probability of injury [2,3,6,7]. oritself, under-pronation, depicted in run Figure is the opposite probability of injury [2,3,6,7]. Supination or under-pronation, as depicted in Figure 1d, is the opposite of pronation and refers to the outward roll of the foot during normal motion [2]. of pronation and refers to the outward roll of the foot during normal motion [2].
(a)
(b)
(c)
(d)
Figure Figure 1. 1. Schematic Schematic illustration illustration of of foot foot pressure pressure distribution distribution of of (a) (a) flat flat foot; foot; (b) (b) neutral neutral foot; foot; (c) (c) cavus cavus foot; and (d) supinated foot of left footprint when standing. foot; and (d) supinated foot of left footprint when standing.
Learning how to identify our foot type is necessary, because the foot plays a very important role Learning how to identify our foot type is necessary, because the foot plays a very important role in our daily life. We can choose the right shoes to support our foot type to avoid foot pain or injury in our daily life. We can choose the right shoes to support our foot type to avoid foot pain or injury when we are walking or running. Foot plantar pressure has been recognized as an important factor when we are walking or running. Foot plantar pressure has been recognized as an important factor in the assessment of patients with diabetes and peripheral neuropathy [8–12]. Many papers have been in the assessment of patients with diabetes and peripheral neuropathy [8–12]. Many papers have published to explore the plantar pressure distribution [13–18] but have rarely addressed the been published to explore the plantar pressure distribution [13–18] but have rarely addressed the application of fiber Bragg gratings (FBGs) on plantar pressure measurement [19]. This literature used application of fiber Bragg gratings (FBGs) on plantar pressure measurement [19]. This literature used three FBGs for forward and backward postural stability analysis. three FBGs for forward and backward postural stability analysis. In recent years, fiber-optic sensors have already been commonly utilized for numerous physical In recent years, fiber-optic sensors have already been commonly utilized for numerous physical measurements [20–26]. The fiber-optic sensor has its intrinsic advantages of light weight, low loss, measurements [20–26]. The fiber-optic sensor has its intrinsic advantages of light weight, low loss, high bandwidth, and immunity to electromagnetic interference. It can be used in a variety of harsh high bandwidth, and immunity to electromagnetic interference. It can be used in a variety of harsh environments and under high electric field strength to do environmental monitoring. Several fiber environments and under high electric field strength to do environmental monitoring. Several fiber sensor designs based on fiber Bragg gratings (FBGs) for various applications have been reported, sensor designs based on fiber Bragg gratings (FBGs) for various applications have been reported, such as biomedical measurements [19,27], pressure [28–30], displacement [31,32], temperature [33– such as biomedical measurements [19,27], pressure [28–30], displacement [31,32], temperature [33–35], 35], tilt [36,37] and ground vibration [38]. In this paper, we propose a fiber-optic sensor based on tilt [36,37] and ground vibration [38]. In this paper, we propose a fiber-optic sensor based on FBGs to FBGs to detect plantar pressure. It has the advantages of a simple architecture (only using six sensing detect plantar pressure. It has the advantages of a simple architecture (only using six sensing elements), elements), low cost, temperature insensitivity, ease of classifying foot type and high stability. It can low cost, temperature insensitivity, ease of classifying foot type and high stability. It can be used in the be used in the measurement of human plantar pressure distribution to classify foot type and measurement of human plantar pressure distribution to classify foot type and understand whether the understand whether the foot type needs to be corrected or not. foot type needs to be corrected or not. 2. General Description and Fiber-Optic Sensor Configuration 2. General Description and Fiber-Optic Sensor Configuration Figure 2 depicts an overview of the fiber-optic sensor measurement system used in plantar Figure 2 depicts an overview of the fiber-optic sensor measurement system used in plantar pressure detection. The output beam of a broadband source (BBS) is connected to the first port of the pressure detection. The output beam of a broadband source (BBS) is connected to the first port of the optical circulator (OC), and the six FBG sensors used to measure plantar pressure are connected to optical circulator (OC), and the six FBG sensors used to measure plantar pressure are connected to the the second port of the OC. The BBS enters into port 1 and leaves from port 2, as shown in Figure 2. second port of the OC. The BBS enters into port 1 and leaves from port 2, as shown in Figure 2. Some of Some of the emitting light, which satisfies the Bragg condition, is reflected back to the circulator and the emitting light, which satisfies the Bragg condition, is reflected back to the circulator and then exits then exits from port 3 into the optical spectrum analyzer (OSA). From the spectrum presented by the from port 3 into the optical spectrum analyzer (OSA). From the spectrum presented by the OSA, we can OSA, we can observe the shifts of the central wavelength of the FBGs. The sensing element of each observe the shifts of the central wavelength of the FBGs. The sensing element of each FBG is embedded FBG is embedded in silicone rubber, which is 3 cm long, 2 cm wide and 0.5 cm high, and placed on in silicone rubber, which is 3 cm long, 2 cm wide and 0.5 cm high, and placed on the insole. There are the insole. There are six polymer sensors in this configuration. Figure 3 shows the initial reflection six polymer sensors in this configuration. Figure 3 shows the initial reflection wavelength spectrum of wavelength spectrum of the six FBGs employed for plantar pressure measurement. The initial the six FBGs employed for plantar pressure measurement. The initial reflection central wavelengths reflection central wavelengths are 1539.610 nm, 1543.825 nm, 1547.960 nm, 1551.745 nm, 1555.745 nm are 1539.610 nm, 1543.825 nm, 1547.960 nm, 1551.745 nm, 1555.745 nm and 1559.785 nm at room and 1559.785 nm at room temperature for FBG1, FBG2, FBG3, FBG4, FBG5, and FBG6, respectively.
Sensors 2016, 16, 1766 Sensors 2016, 16, 1766 Sensors 2016, 16, 1766
3 of 13 3 of 13 3 of 13
In addition, the sensitive length of the FBGs is equal to 1 cm. The channel spacing of the central In addition,isthe sensitive ofathe FBGs is equalof toany 1 cm. channel spacing of the central wavelength about 4 nm length to avoid superimposition twoThe consecutive central wavelengths. wavelength is about 4 nm to avoid a superimposition of any two consecutive central When measuring the shifts of the reflection of each FBG, we can obtainwavelengths. thesensitive relative temperature for FBG1, FBG2, FBG3, FBG4,central FBG5, wavelength and FBG6, respectively. In addition, the When measuring the shifts of reflection centralspacing wavelength each FBG, we can obtain the4relative plantar pressure distribution. length of the FBGs is equal to the 1 cm. The channel of theofcentral wavelength is about nm to plantar pressure distribution. a fiber gratingofirradiates with a broadband source,When the grating will the reflect light avoidWhen a superimposition any two consecutive central optical wavelengths. measuring shifts of a central fiber corresponding grating irradiates with a broadband optical source,plantar theby grating will distribution. reflect light with aWhen wavelength twice the grating period (Λ) relative multiplied the effective refractive the reflection wavelength oftoeach FBG, we can obtain the pressure withWhen a( wavelength corresponding to core. twice grating period (Λ)source, multiplied by the effective refractive index the grating inirradiates the fiber central wavelength ( B ) of the reflected spectral band a fiber grating withThe athe broadband optical the grating will reflect light neff ) of index ( ) of the grating in the fiber core. The central wavelength ( ) of the reflected spectral band n with a wavelength corresponding twice the grating period (Λ) multiplied by the effective refractive eff by the Bragg B is defined condition to [39]: is defined Bragg condition [39]: index (neff ) by of the grating in the fiber core. The central wavelength (λ B ) of the reflected spectral band is B 2neff (1) defined by the Bragg condition [39]: B 2neff (1) Both the fiber refractive index ( neff ) and grating λ B the = 2n (1) eff · Λ pitch ( ) vary with the changes of strain Both the fiber refractive index ( ) and the grating pitch ( ) vary with the changes of strain n eff Bragg wavelength is sensitive to strain and temperature. and temperature, consequently the Both the fiber and refractive index (neff ) and the grating pitch (Λ) vary with the changes of strain and and temperature, and consequently the Bragg wavelength is grating sensitivewill to strain and temperature temperature. Even though the reflecting central wavelength of the issensor varytemperature. with temperature, and consequently the Bragg wavelength sensitive to strain and Even Even though the reflecting central wavelength of the sensor grating will vary with temperature slightly, the effect of temperature can be excluded. The reason is that an identification is determined though the reflecting central wavelength of the sensor grating will vary with temperature slightly, slightly, the effect oflocations temperature cancentral be excluded. The reason is thatfrom an identification is determined by comparative of excluded. the wavelengths measured the FBG sensors. That is, thethe effect of temperature can be The reason is that an identification is determined by the by the comparative locations of the central wavelengths measured from the FBG sensors. That is, even under different temperatures, the comparative shifts of the central wavelengths are unchanged comparative locations of the central wavelengths measured from the FBG sensors. That is, even under eventhe under different temperatures, the to comparative shifts of the central wavelengths are unchanged and results identified irrelevant temperature. different temperatures, theare comparative shifts of the central wavelengths are unchanged and the results and the results identified are irrelevant to temperature. identified are irrelevant to temperature.
Figure 2. The configuration of the fiber sensing system for plantar pressure measurement. Figure Figure 2. 2. The The configuration configuration of of the the fiber fiber sensing sensing system systemfor forplantar plantarpressure pressuremeasurement. measurement.
Figure 3. The reflection wavelength spectrum of six FBGs without any stress. Figure 3. The reflection wavelength spectrum of six FBGs without any stress. Figure 3. The reflection wavelength spectrum of six FBGs without any stress.
3. Methods and Results 3. Methods and Results
Sensors 2016, 16, 1766 Sensors 2016, 16, 1766
4 of 13 4 of 13
3.1.3.Subjects Methods and Results The primary aim of this work is to investigate the feasibility and functionality of the FBG sensor 3.1. Subjects system to measure the plantar pressure of the foot, and thus spatially resolve the foot pressure The primary aim of this work to the investigate feasibility functionality the FBG(nine sensor distribution at different regions suchisas forefoot,the midfoot andand hindfoot. Elevenofsubjects system measure theaged plantar theweighing foot, andfrom thus47spatially the foot in pressure males andto two females), 21 topressure 25 yearsofand to 83 kg,resolve were included this distribution at different regions such as the forefoot, midfoot and hindfoot. Eleven subjects (nine work. The Institutional Review Board of Kaohsiung Medical University Chung-Ho Memorial males and two females), aged this 21 tostudy 25 years weighing from 47 to 83consent. kg, were in this Hospital, Taiwan, has approved and and all subjects gave informed Allincluded participants work. The Institutional Review Board Kaohsiung University Chung-Ho Memorial Hospital, have been diagnosed to identify their footoftypes by the Medical digital pressure plate of the i-Step P1000 before Taiwan, has approved studypressure and all subjects consent. participants have been being subjected to FBG this plantar testing.gave The informed i-Step P1000 is aAll digital pressure plate to W identify their by the digital pressure plate of the the pressure i-Step P1000 before (20diagnosed in L × 18 in × 1.5 in H) foot withtypes 1024 barometric sensors to measure exerted by being the 2. Itplantar subjected FBG pressure testing. The i-Step P1000 is a digital plate (20 in L × 18 in foot every 1tocm is a current clinical foot pressure instrument and pressure is used as a measurement W × 1.5ininthis H) study. with 1024 to measure the1,pressure exerted by the foot every 1 cm2 . reference Thebarometric results aresensors summarized in Table in which there are three participants It isnormal a current footcavus pressure and is used as a measurement in thiswith study. with feet,clinical four with feet,instrument two with flat feet, one with supinated feet,reference and the other The results are summarized in Table 1, in which there are three participants with normal feet, four with a cavus left foot and a neutral right foot. Figure 4 shows the partial measurement results of foot cavus feet, two with flat feet, one with supinated feet, and the other with a cavus left foot and a neutral pressure distribution. right foot. Figure 4 shows the partial measurement results of foot pressure distribution. Table 1. The foot category of participants using the digital pressure plate i-Step 1000. Table 1. The foot category of participants using the digital pressure plate i-Step 1000. Foot Type No. Left Right Foot Type 1 neutral No. neutral Left Rightneutral 2 neutral 1 neutral neutral 3 neutral neutral 2 neutral neutralcavus 4 cavus 3 neutral neutralcavus 5 cavus 4 cavus cavuscavus 6 cavus 5 cavus cavus 7 cavus cavus 6 cavus cavus 8 flat flat 7 cavus cavus 9 flat 8 flat flat flat 10 supinated 9 flat flatsupinated 11 cavus supinated neutral 10 supinated 11 cavus neutral
(a)
(b) Figure 4. Cont.
(c)
Sensors 16,16, 1766 Sensors 2016, 1766 Sensors 2016, 16, 17662016,
(d)
5 of 13
(d)
5 of 1313 5 of
(e)
(e)
Figure 4. 4. Foot plantar pressure distribution ofof participants: No. (b) No. 6; (c) No. (d)(d) No. 10;10; Figure 4. Foot plantar pressure distribution ofdistribution participants: (a) No. 2; (b)(a) No. 6; 2; (c)2; No. 8; (d)6; No. 10;8; 8; Figure Foot plantar pressure participants: (a) No. (b) No. (c) No. No. (e)i-Step No. 11, by i-Step P1000 digital pressure plate. (e) No. 11, by P1000 digital pressure plate. (e) No. 11, by i-Step P1000 digital pressure plate.
Plantar Foot Sensing Area Definition 3.2. Plantar3.2. Foot Definition 3.2.Sensing Plantar Area Foot Sensing Area Definition In this paper, plantardistributions pressure distributions wereaccording analyzed to according to the Pedar-X in In this paper, pressure were analyzed the Pedar-X in system Inplantar this paper, plantar pressure distributions were analyzed according to system the Pedar-X system in 11 healthy subjects. The foot can be subdivided into four segments of the forefoot, midfoot, hindfoot 11 healthy subjects. Thesubjects. foot can The be subdivided four segments of the forefoot, midfoot, hindfoot 11 healthy foot can beinto subdivided into four segments of the forefoot, midfoot, hindfoot toes. In order to measurement set the six measurement areas, first we the measured thethe length of the participant's and toes. Inand order to set the six areas, first we measured length of participant's and toes. In order to set the six measurement areas, first we measured the length of the participant's Then,ofbythe means of the general definition, can get the threeofsegments the foot: front, middle, foot. Then, foot. byfoot. means general wedefinition, can get we the three foot: of front, middle, Then, by means ofdefinition, the general we can segments get the threethe segments of the foot: front, middle, and hind. For the length of any one of the feet, the hindfoot length accounts for 30%, the midfoot and hind. For thehind. length ofthe anylength one ofofthe hindfoot length accounts for accounts 30%, the for midfoot and For anyfeet, onethe of the feet, the hindfoot length 30%, the midfoot length 30%, the forefoot portion 25%, and the toes account for 15% [40,41]. In accordance with length is 30%, theisforefoot portion is 25%, andis toes account foraccount 15% [40,41]. In[40,41]. accordance with length is 30%, the forefoot portion isthe 25%, and the toes for 15% In accordance with the the average pressure of the measurement region, we defined six sensing sub-segments [41]: hindfoot the average pressure of the measurement region, we defined sensing [41]: hindfoot average pressure of the measurement region, wesix defined sixsub-segments sensing sub-segments [41]: hindfoot of of one, midfoot offorefoot two, and forefoot of three, without the toes, as illustrated in Figure 5. of one, midfoot of two, and of three, without the toes, as illustrated in Figure 5. one, midfoot of two, and forefoot of three, without the toes, as illustrated in Figure 5.
1.
1. Medial forefoot Medial forefoot
2.
2. Central forefoot Central forefoot
3.
3. Lateral forefoot Lateral forefoot
4.
4. Medial midfoot Medial midfoot
5.
5. Lateral midfoot Lateral midfoot
6.
Heel
6.
Heel
Figure The sensing area definitions this research. Figure 5. 5. The sixsix sensing area definitions inin this research. Figure 5. The six sensing area definitions in this research.
Figure 5, 5, these plantar pressure sensing areas are numbered 1 as the medial 2 as In Figure 5,Inthese plantar pressure sensing areas are numbered bynumbered 1 as the by medial forefoot, 2 forefoot, asforefoot, In Figure these plantar pressure sensing areas are by 1 as the medial 2 as central forefoot, 3 as the lateral forefoot, 4 as the medial midfoot, 5 as the lateral midfoot, and the central the forefoot, 3 as the lateral forefoot, 4 as the medial midfoot, 5 as the lateral midfoot, and 6 the central forefoot, 3 as the lateral forefoot, 4 as the medial midfoot, 5 as the lateral midfoot, and 66as as theheel. heel. When afoot different foot stands oninsole, the six regions will exhibit different as the heel. When a different type stands on the insole, the insole, six will exhibit different the When a different foot typetype stands on the theregions sixthe regions will exhibit different pressures. pressures. All subjects were asked to measure each foot five times, and then we took the average of pressures. All subjects were asked to measure each foot five times, and then we took the average of All subjects were asked to measure each foot five times, and then we took the average of the central central wavelength shift of each FBG sensor. According to the variation of the central wavelength the central the wavelength shift of each FBG sensor. According to the variation of the central wavelength wavelength shift of each FBG sensor. According to the variation of the central wavelength shift of the of the six FBG sensors, can get the relative distribution of the plantar pressure and determine shift of the shift six sensors, we can the relative distribution ofofthe pressure and determine sixFBG FBG sensors, we canget getwe the relative distribution theplantar plantar pressure and determine the subject's the subject's foot type. Figure 6 shows the prototype of the FBG sensors and a subject standing the subject's foot type. Figure 6 shows the prototype of the FBG sensors and a subject standing foot type. Figure 6 shows the prototype of the FBG sensors and a subject standing barefoot on the on the detection system. barefoot onbarefoot the detection system. detection system.
Sensors 2016, 1766 Sensors 2016, 16,16, 1766 Sensors 2016, 16, 1766
6 of 6 of 13 13 6 of 13
Figure 6. 6. Set-up of of FBG sensors and test in in action. Figure Set-up FBG sensors and test action. Figure 6. Set-up of FBG sensors and test in action.
3.3. Experimental Results 3.3. Experimental Results InIn order toto correspond toto the results ofof Figure 4, 4, the measured data from the fiber-optic sensing order correspond the results Figure the measured data from the fiber-optic sensing system for the participants are shown in this section. Figure 7 shows the plantar pressure distribution shows the the plantar plantar pressure pressure distribution system for the participants are shown in this section. Figure 7 shows ofof the left and right foot ofof the neutral feet ofof participant No. 2. 2. Figure 7a7a depicts the footprints and left and right foot ofthe the neutral feet of participant No. 2. Figure 7a depicts the footprints the left and right foot neutral feet participant No. Figure depicts the footprints and Figure 7b shows the central wavelength shift of the FBGs of the six sensing areas. In the figure, the and Figure 7b shows the central wavelength ofFBGs the FBGs of six thesensing six sensing areas. In figure, the figure, Figure 7b shows the central wavelength shiftshift of the of the areas. In the the solid represents the average values ofof five ofof each sensing element and the symbol solid symbol represents the average values ofmeasurements five measurements ofFBG each FBG sensing element solid symbol represents the average values five measurements each FBG sensing element and the line represents the measurement value range of the maximum and minimum. InIn this work, and the line represents the measurement value range ofmaximum the maximum and minimum. Inwork, thiseach work, the line represents the measurement value range of the and minimum. this each measured value slightly changes because the standing position and standing posture are not very each measured slightly changes because standing position andstanding standingposture postureare arenot not very very measured valuevalue slightly changes because thethe standing position and consistent. From Figure 7a,b, wewe can obtain the plantar pressure distribution atat allall areas ofof the foot, From Figure 7a,b, we can obtain the plantar pressure distribution at all areas of the foot, consistent. From Figure 7a,b, can obtain the plantar pressure distribution areas the foot, except forfor the medial midfoot area. That is,is, there is is nono foot pressure that can be detected by FBG4. except the medial midfoot area. That there foot pressure that can be detected by FBG4. pressure that can be detected by FBG4.
(a) (a)
(b) (b)
Figure 7. TheThe plantar pressure distribution of the neutral feet feet of participant No. 2: (a) 2: Footprint; (b) plantar pressure distribution the neutral participant No. (a) Footprint; Figure 7. The plantar pressure distribution ofofthe neutral feet of of participant No. 2: (a) Footprint; (b) central wavelength shift as a function of sensing area number. (b) central wavelength a function of sensing number. central wavelength shiftshift as aasfunction of sensing areaarea number.
Figure 8 shows the plantar pressure distribution ofof the cavus feet ofof participant No. 6, 6, one ofof the Figure distribution the cavus feet participant No. Figure 88 shows shows the the plantar plantar pressure pressure distribution of the cavus feet of participant No. 6, one one of the the four participants who has cavus foot. The cavus foot is is commonly known asas claw foot. The cavus foot four participants who has cavus foot. The cavus foot commonly known claw foot. The cavus foot four participants who has cavus foot. The cavus foot is commonly known as claw foot. The cavus foot type will make the foot hollow when bearing weight. AA high arch is is the opposite ofof a flat foot. Because type when bearing weight. high arch type will will make make the the foot foot hollow hollow when bearing weight. A high arch is the the opposite opposite of aa flat flat foot. foot. Because Because the arch area of the bow foot is relatively high, the pressure will be concentrated on the forefoot and the arch area of the bow foot is relatively high, the pressure will be concentrated on the forefoot and the arch area of the bow foot is relatively high, the pressure will be concentrated on the forefoot and heel areas. From Figure 8b, in the fifth sensing area, only a little pressure is sensed. The plantar heel areas. From sensing area, area, only only aa little little pressure pressure is is sensed. sensed. The plantar heel areas. From Figure Figure 8b, 8b, in in the the fifth fifth sensing The plantar pressure distributions are consistent among the results measured bybythe digital pressure plate ofof pressure distributions are consistent among the results measured the digital pressure plate pressure distributions are consistent among the results measured by the digital pressure plate of i-Step i-Step P1000 (Figure 4b), footprints (Figure 8a), and FBG fiber sensor (Figure 8b). i-Step 4b), footprints 8a), andfiber FBGsensor fiber sensor P1000 P1000 (Figure(Figure 4b), footprints (Figure(Figure 8a), and FBG (Figure(Figure 8b). 8b).
Sensors 2016, 16, 1766 Sensors Sensors2016, 2016,16, 16,1766 1766
7 of 13 7 7ofof1313
(a) (a)
(b) (b)
Figure 8.8.8. The plantar pressure distribution ofofthe cavus feet ofofparticipant No. Footprint; (b) Figure The plantar pressure distribution of the cavus of participant No. (a) Footprint; Figure The plantar pressure distribution the cavus feetfeet participant No.6: 6:(a) (a)6: Footprint; (b) central wavelength shift ofofsensing area (b) central wavelength shift as a function of sensing area number. central wavelength shiftas asaafunction function sensing areanumber. number.
Flat feet occur because the holding the joints in the foot together (called tendons) are Flat thetissues tissuesholding holdingthe thejoints joints together (called tendons) are Flatfeet feet occur occur because because the tissues in in thethe footfoot together (called tendons) are loose. loose. As a result, tendons do not pull together properly; there is little or no arch. Figure 9 depicts the loose. As a result, tendons do not pull together properly; there is little or no arch. Figure 9 depicts the As a result, tendons do not pull together properly; there is little or no arch. Figure 9 depicts the plantar plantar pressure distribution of a flat foot. It differs from the others foot types; that is, there will be plantar pressure distribution a flat It from differs from thefoot others footthat types; that is, there will be pressure distribution of a flatoffoot. It foot. differs the others types; is, there will be pressure pressure distribution on the medial arch area, and there will be no arch shape. The measurement pressure distribution on the medial arch area, and there will be no arch shape. The measurement distribution on the medial arch area, and there will be no arch shape. The measurement result of the result of flat No. 88of sensing system isisshown in 9b. result ofthe the flatfoot footof ofparticipant participant ofthe thefiber-optic fiber-optic sensing systemin shown9b. inFigure Figure 9b.The Theof flat foot of participant No. 8 of theNo. fiber-optic sensing system is shown Figure The pressure pressure of the fourth region (the medial midfoot) with the green circle is not zero. That phenomenon pressure of the fourth medialwith midfoot) withcircle the green not zero. That phenomenon the fourth region (theregion medial(the midfoot) the green is notcircle zero.isThat phenomenon is different isisdifferent from the types, shows aasimilar pressure as measured different fromfoot theother otherfoot foot types,and and shows similar pressuredistribution distribution asthe the measured from the other types, and shows a similar pressure distribution as the measured results of the results of the digital pressure plate of i-Step P1000 (see Figure 4c). results the digital pressure plate of i-Step P10004c). (see Figure 4c). digitalofpressure plate of i-Step P1000 (see Figure
(a) (a)
(b) (b)
Figure 9. The plantar pressure distribution of the flat feet ofofparticipant No. 8:8:(a) Footprint; (b) central Figure Figure9.9.The Theplantar plantarpressure pressuredistribution distributionofofthe theflat flatfeet feet ofparticipant participantNo. No. 8:(a) (a)Footprint; Footprint;(b) (b)central central wavelength shift as a function of sensing area number. wavelength wavelengthshift shiftasasaafunction functionofofsensing sensingarea areanumber. number.
In Inthis thiswork, work,the thefootprint footprintand andfiber-optic fiber-opticsensing sensingresult resultof ofthe theunique uniqueparticipant participantwith withsupinated supinated In this work, the footprint and fiber-optic sensing result of the unique participant with supinated feet is shown in Figure 10. Supination or under-pronation is a condition that occurs when feet is shown in Figure 10. Supination or under-pronation is a condition that occurs whenfeet feetroll roll feet is shown in Figure 10. Supination or under-pronation is afoot condition that occurs when feet roll outwards, placing weight on the outside of the foot. While the leans to the outside, weight outwards, placing weight on the outside of the foot. While the foot leans to the outside, weightisis outwards, placing weight onAthe outside ofsupinated the foot. While the foot leans to the outside,shockweight distributed distributedalong alongthe theoutside. outside. Aperson personwith with supinatedfeet feetwill willreduce reducethe thebody's body'snatural natural shockis distributed along the outside. A person with supinated feet will reduce the body's natural absorbing absorbingcapability. capability.From FromFigure Figure10, 10,we wecan cansee seethat thatthere thereisisno nopressure pressuredetection detectionon onarea area4,4,and and shock-absorbing capability. From This Figure 10, wepressure can see that there is no pressure detection on area 4, area 1 only has very light pressure. plantar distribution occurs at the foot supination. area 1 only has very light pressure. This plantar pressure distribution occurs at the foot supination. and areashows 1 only hasexperimental very light pressure. This plantar pressure distribution occurs at the foot supination. Figure Figure11 11 showsthe the experimentalresults resultsof ofparticipant participantNo. No.11. 11.One Oneof ofhis hisfeet feet(left (leftfoot) foot)isispes pescavus cavus and andthe theother otherisisaaneutral neutralfoot. foot.From Fromthis thisfigure, figure,we wecan caneasily easilydistinguish distinguishthe thefoot foottypes. types.
Sensors 2016, 16, 1766
8 of 13
Sensors 2016, 16, 1766 111766 shows SensorsFigure 2016, 16,
8 of 13 the experimental results of participant No. 11. One of his feet (left foot) is 8pes of 13cavus and the other is a neutral foot. From this figure, we can easily distinguish the foot types.
(a) (a)
(b) (b)
Figure 10. The plantar pressure distribution of supinated feet of participant No. 10: (a) Footprint; (b) Figure 10. The pressure distribution of supinated feet of participant No. 10: (a) Footprint; (b) Figure 10.plantar Theshift plantar of supinated central wavelength as apressure function distribution of sensing area number. feet of participant No. 10: (a) Footprint; central shift as ashift function of sensing number. (b)wavelength central wavelength as a function ofarea sensing area number.
(a) (a)
(b) (b)
Figure 11. The pressure distribution of participant No. 11, left foot cavus and right Figure 11. plantar The plantar pressure distribution of participant No.whose 11, whose left is foot is cavus and right Figure 11. The plantar pressure distribution of participant No. 11, whose left foot is cavus and right foot is neutral: (a) Footprint; (b) central wavelength shift as a function of sensing area number. foot is neutral: (a) Footprint; (b) central wavelength shift as a function of sensing area number. foot is neutral: (a) Footprint; (b) central wavelength shift as a function of sensing area number.
3.4. Sensing Stability 3.4. Sensing Stability 3.4. Sensing Stability In Figure 12, we demonstrate the sensing stability of the experimental measurement results of In Figure 12,demonstrate we demonstrate the sensing stability the experimental measurement results In Figure 12, we the sensing stability of theofexperimental measurement results of of the FBG sensors in four different foot foot types. Each subject’s foot was measured five times in this study. the sensors FBG sensors in different four different types. subject’s footmeasured was measured five times this study. the FBG in four foot types. Each Each subject’s foot was five times in thisinstudy. Although the result of each measurement is slightly changed due to the inconsistence of the standing Although the result of each measurement is slightly changed due to the inconsistence of the standing Although the result of each measurement is slightly changed due to the inconsistence of the standing position and posture each time, the overall foot pressure distribution for the foot-type classification position and posture the overall foot pressure distribution forfoot-type the foot-type classification position and posture each each time,time, the overall foot pressure distribution for the classification will will not not be affected. If the object has flat feet, it is inevitable inevitabletotodetect detectfoot foot pressure in the fourth be affected. If the object has flat feet, it is pressure in the fourth sensing will not be affected. If the object has flat feet, it is inevitable to detect foot pressure in the fourth sensing region (Figure 12a). The larger the foot pressure that is measured in the fourth sensing area, region (Figure 12a). The larger the foot pressure that is measured in the fourth sensing area, the sensing region (Figure 12a). The larger the foot pressure that is measured in the fourth sensing area,more the more severe the flat foot situation. In contrast, there is no foot pressure in the fourth sensing areaif the severe the flat situation. In contrast, there is no foot pressure inin the the more severe the foot flat foot situation. In contrast, there is no foot pressure thefourth fourthsensing sensingarea area if theparticipant participant has supinated feet. At the same time, the plantar pressure is almost close to zero in has supinated feet. At the same time, the plantar pressure is almost close to zero in if the participant has supinated feet. At the same time, the plantar pressure is almost close to zerothe in first the first sensing area for a supinated foot (Figure 12b). Moreover, if the participant has cavus feet, sensing areaarea for afor supinated foot foot (Figure 12b).12b). Moreover, if theifparticipant has cavus feet, there the first sensing a supinated (Figure Moreover, the participant has cavus feet, will therebewill be little foot pressure in the fifth sensing region (Figure 12c). Of course, the sensing value little in thein fifth regionregion (Figure 12c). Of course, the sensing value value of the foot there will be foot littlepressure foot pressure thesensing fifth sensing (Figure 12c). Of course, the sensing of the foot pressure of themidfoot medialarea midfoot area is Other also zero. Other thansituations, the above situations, the pressure of the medial is also zero. than the above the sensitive stability of the foot pressure of the medial midfoot area is also zero. Other than the above situations, the sensitive stability of aisneutral foot is shown in Figure 12d. of a neutral shownfoot in Figure 12d.in sensitive stabilityfoot of a neutral is shown Figure 12d.
Sensors 2016, 16, 1766 Sensors 2016, 16, 1766
9 of 13 9 of 13
(a)
(b)
(c)
(d) Figure 12. The central wavelength shift as a function of sensing area number for plantar pressure Figure 12. The central wavelength shift as a function of sensing area number for plantar pressure distribution of left and right foot: (a) Flat feet; (b) supinated feet; (c) cavus feet; and (d) neutral feet. distribution of left and right foot: (a) Flat feet; (b) supinated feet; (c) cavus feet; and (d) neutral feet.
Sensors 2016, 16, 1766 Sensors 2016, 16, 1766
10 of 13 10 of 13
Some the abnormal plantar pressure pattern, which is is related toto obesity Someliterature literaturehas hasdiscussed discussed the abnormal plantar pressure pattern, which related obesity ororoverweight, and can be identified with its foot type by visual non-quantitative inspection. Most ofof overweight, and can be identified with its foot type by visual non-quantitative inspection. Most the articles plantar pressure single the articlesonon plantar pressuresensors sensorsdid didnot notnormalize normalizethe thesensor sensoroutput outputwith withrespect respecttotothe the single body mass value. In this work, two subjects with flat feet of different weights, 62 kg in Figure 9b body mass value. In this work, two subjects with flat feet of different weights, 62 kg in Figure 9band and 6565kgkgininFigure Figure12a, 12a,show showthe thenon-zero non-zeropressure pressureatatthe themidfoot midfootregion. region.Two Twosubjects subjectswith withcavus cavusfeet feet ofofdifferent weights, 71 kg in Figure 8b and 52 kg in Figure 12c, show similar pressure distribution different weights, 71 kg in Figure 8b and 52 kg in Figure 12c, show similar pressure distribution patterns. curve vertically verticallyand andslightly slightlychange changethe thecurve curve pattern patterns.The Thesubject’s subject’sweight weightcan can offset offset the the curve pattern but but not the foot-type classification. not the foot-type classification. 3.5. Verification 3.5.Statistical Statistical Verification The product-moment linear ThePearson Pearson product-momentcorrelation correlationcoefficient coefficient(PPMCC (PPMCCororPCC) PCC)isisa ameasure measureofofthe the linear correlation between two variables [42]. A value of zero indicates that there is no association between correlation between two variables [42]. A value of zero indicates that there is no association between the order the ofofthe thetwo twovariables. variables.InIn ordertotounderstand understand therelationship relationship themeasurement measurementdata datafrom fromthe thedigital digital pressure i-Step P1000 P1000and and fiber-optic sensing system, wePearson’s used Pearson’s correlation pressureplate plate i-Step thethe fiber-optic sensing system, we used correlation coefficient coefficient method. we calculated the values averageofvalues digital pressure on 3ancm area cm in 2i-Step cm method. First, we First, calculated the average digitalofpressure on an area × 23 cm inP1000. i-Step This P1000. This area corresponds one the grating six fiber gratingareas, sensing areas, the central area corresponds to one oftothe sixoffiber sensing the central wavelength wavelength shift of which is measured. Figure 13 shows the correlation of all central wavelength shift of which is measured. Figure 13 shows the correlation of all central wavelength shifts of the six shifts the sixinFBG in this and the plantar pressure valuespressure of the digital FBGof sensors thissensors study and thestudy plantar pressure values of the digital platepressure of i-Stepplate P1000. ofThen, i-Stepwe P1000. Then, we the Pearsoncoefficient correlation coefficient and got a r-value 0.671which (p < calculated thecalculated Pearson correlation and got a r-value of 0.671 (p
Plantar Pressure Detection with Fiber Bragg Gratings Sensing System Tsair-Chun Liang 1, *, Jhe-Jhun Lin 1 and Lan-Yuen Guo 2 1 2
*
Graduate Institute of Electrical Engineering, National Kaohsiung First University of Science and Technology, Kaohsiung City 811, Taiwan; [email protected] Department of Sports Medicine, Kaohsiung Medical University, Kaohsiung City 807, Taiwan; [email protected] Correspondence: [email protected]; Tel.: +886-7-6011-000 (ext. 2712)
Academic Editor: Vittorio M. N. Passaro Received: 17 June 2016; Accepted: 19 October 2016; Published: 22 October 2016
Abstract: In this paper, a novel fiber-optic sensing system based on fiber Bragg gratings (FBGs) to measure foot plantar pressure is proposed. This study first explores the Pedar-X insole foot pressure types of the adult-size chart and then defines six measurement areas to effectively identify four foot types: neutral foot, cavus foot, supinated foot and flat foot. The plantar pressure signals are detected by only six FBGs, which are embedded in silicone rubber. The performance of the fiber optic sensing is examined and compared with a digital pressure plate of i-Step P1000 with 1024 barometric sensors. In the experiment, there are 11 participants with different foot types to participate in the test. The Pearson correlation coefficient, which is determined from the measured results of the homemade fiber-optic plantar pressure system and i-Step P1000 plantar pressure plate, reaches up to 0.671 (p < 0.01). According to the measured results from the plantar pressure data, the proposed fiber optic sensing system can successfully identify the four different foot types. Measurements of this study have demonstrated the feasibility of the proposed system so that it can be an alternative for plantar pressure detection systems. Keywords: fiber-optic sensor; fiber Bragg grating; foot plantar pressure; Pearson product-moment correlation coefficient
1. Introduction Not all feet are naturally identical since the human foot includes a lot of bones, joints, muscles, tendons, and ligaments. While the bones, joints, and tendons of a person differ little from that of another, a variety of different foot types come up substantially. In general, the human foot can be classified into five types: flat foot, neutral foot, pronated foot, supinated foot and cavus foot [1–3]. Some classification methods to identify foot type have been used in the literature [1,4,5], covering visual non-quantitative inspection, clinical values, footprint indices, and radiographic evaluation, etc. Figure 1 shows the four types of feet that are studied in this paper. As shown in Figure 1a, the flat foot is a postural deformity, in which the arch of the foot collapses on the entire sole of the foot, having complete or near-complete contact with the ground. This foot type needs extra support in the arch region to maintain a neutral gait and to dispel the shock of running [2,6,7]. If the arch of the foot lightly collapses inward, it can be called a neutral foot. Figure 1b shows the footprint of the neutral (or normal) foot type. The neutral foot is the most common type of foot and it allows the body to naturally absorb shock [2,6,7]. If the ball and heel of the footprint are connected by a thin strip on the outside or no strip at all (see Figure 1c), this is a cavus foot. This foot type needs the most assistance with shock absorption, since the rigid structure is unable to dispel impact forces very well. Sensors 2016, 16, 1766; doi:10.3390/s16101766
www.mdpi.com/journal/sensors
Sensors Sensors 2016, 16, 1766
22 of of 13 13
The pronated foot or eversion foot is the inward roll to distribute the impact force caused by the Thewhile pronated footor orrunning. eversionItfoot theof inward roll tomovements distribute the impact caused bydeal the ground walking is aispart the natural which helpforce the lower leg ground while walking or running. It is a part of the natural movements which help the lower leg with shock. Though this is not bad in itself, it does affect the way you run and may increase the deal with shock. Though thisSupination is not bad in it does affectasthe way you and1d, may increase the probability of injury [2,3,6,7]. oritself, under-pronation, depicted in run Figure is the opposite probability of injury [2,3,6,7]. Supination or under-pronation, as depicted in Figure 1d, is the opposite of pronation and refers to the outward roll of the foot during normal motion [2]. of pronation and refers to the outward roll of the foot during normal motion [2].
(a)
(b)
(c)
(d)
Figure Figure 1. 1. Schematic Schematic illustration illustration of of foot foot pressure pressure distribution distribution of of (a) (a) flat flat foot; foot; (b) (b) neutral neutral foot; foot; (c) (c) cavus cavus foot; and (d) supinated foot of left footprint when standing. foot; and (d) supinated foot of left footprint when standing.
Learning how to identify our foot type is necessary, because the foot plays a very important role Learning how to identify our foot type is necessary, because the foot plays a very important role in our daily life. We can choose the right shoes to support our foot type to avoid foot pain or injury in our daily life. We can choose the right shoes to support our foot type to avoid foot pain or injury when we are walking or running. Foot plantar pressure has been recognized as an important factor when we are walking or running. Foot plantar pressure has been recognized as an important factor in the assessment of patients with diabetes and peripheral neuropathy [8–12]. Many papers have been in the assessment of patients with diabetes and peripheral neuropathy [8–12]. Many papers have published to explore the plantar pressure distribution [13–18] but have rarely addressed the been published to explore the plantar pressure distribution [13–18] but have rarely addressed the application of fiber Bragg gratings (FBGs) on plantar pressure measurement [19]. This literature used application of fiber Bragg gratings (FBGs) on plantar pressure measurement [19]. This literature used three FBGs for forward and backward postural stability analysis. three FBGs for forward and backward postural stability analysis. In recent years, fiber-optic sensors have already been commonly utilized for numerous physical In recent years, fiber-optic sensors have already been commonly utilized for numerous physical measurements [20–26]. The fiber-optic sensor has its intrinsic advantages of light weight, low loss, measurements [20–26]. The fiber-optic sensor has its intrinsic advantages of light weight, low loss, high bandwidth, and immunity to electromagnetic interference. It can be used in a variety of harsh high bandwidth, and immunity to electromagnetic interference. It can be used in a variety of harsh environments and under high electric field strength to do environmental monitoring. Several fiber environments and under high electric field strength to do environmental monitoring. Several fiber sensor designs based on fiber Bragg gratings (FBGs) for various applications have been reported, sensor designs based on fiber Bragg gratings (FBGs) for various applications have been reported, such as biomedical measurements [19,27], pressure [28–30], displacement [31,32], temperature [33– such as biomedical measurements [19,27], pressure [28–30], displacement [31,32], temperature [33–35], 35], tilt [36,37] and ground vibration [38]. In this paper, we propose a fiber-optic sensor based on tilt [36,37] and ground vibration [38]. In this paper, we propose a fiber-optic sensor based on FBGs to FBGs to detect plantar pressure. It has the advantages of a simple architecture (only using six sensing detect plantar pressure. It has the advantages of a simple architecture (only using six sensing elements), elements), low cost, temperature insensitivity, ease of classifying foot type and high stability. It can low cost, temperature insensitivity, ease of classifying foot type and high stability. It can be used in the be used in the measurement of human plantar pressure distribution to classify foot type and measurement of human plantar pressure distribution to classify foot type and understand whether the understand whether the foot type needs to be corrected or not. foot type needs to be corrected or not. 2. General Description and Fiber-Optic Sensor Configuration 2. General Description and Fiber-Optic Sensor Configuration Figure 2 depicts an overview of the fiber-optic sensor measurement system used in plantar Figure 2 depicts an overview of the fiber-optic sensor measurement system used in plantar pressure detection. The output beam of a broadband source (BBS) is connected to the first port of the pressure detection. The output beam of a broadband source (BBS) is connected to the first port of the optical circulator (OC), and the six FBG sensors used to measure plantar pressure are connected to optical circulator (OC), and the six FBG sensors used to measure plantar pressure are connected to the the second port of the OC. The BBS enters into port 1 and leaves from port 2, as shown in Figure 2. second port of the OC. The BBS enters into port 1 and leaves from port 2, as shown in Figure 2. Some of Some of the emitting light, which satisfies the Bragg condition, is reflected back to the circulator and the emitting light, which satisfies the Bragg condition, is reflected back to the circulator and then exits then exits from port 3 into the optical spectrum analyzer (OSA). From the spectrum presented by the from port 3 into the optical spectrum analyzer (OSA). From the spectrum presented by the OSA, we can OSA, we can observe the shifts of the central wavelength of the FBGs. The sensing element of each observe the shifts of the central wavelength of the FBGs. The sensing element of each FBG is embedded FBG is embedded in silicone rubber, which is 3 cm long, 2 cm wide and 0.5 cm high, and placed on in silicone rubber, which is 3 cm long, 2 cm wide and 0.5 cm high, and placed on the insole. There are the insole. There are six polymer sensors in this configuration. Figure 3 shows the initial reflection six polymer sensors in this configuration. Figure 3 shows the initial reflection wavelength spectrum of wavelength spectrum of the six FBGs employed for plantar pressure measurement. The initial the six FBGs employed for plantar pressure measurement. The initial reflection central wavelengths reflection central wavelengths are 1539.610 nm, 1543.825 nm, 1547.960 nm, 1551.745 nm, 1555.745 nm are 1539.610 nm, 1543.825 nm, 1547.960 nm, 1551.745 nm, 1555.745 nm and 1559.785 nm at room and 1559.785 nm at room temperature for FBG1, FBG2, FBG3, FBG4, FBG5, and FBG6, respectively.
Sensors 2016, 16, 1766 Sensors 2016, 16, 1766 Sensors 2016, 16, 1766
3 of 13 3 of 13 3 of 13
In addition, the sensitive length of the FBGs is equal to 1 cm. The channel spacing of the central In addition,isthe sensitive ofathe FBGs is equalof toany 1 cm. channel spacing of the central wavelength about 4 nm length to avoid superimposition twoThe consecutive central wavelengths. wavelength is about 4 nm to avoid a superimposition of any two consecutive central When measuring the shifts of the reflection of each FBG, we can obtainwavelengths. thesensitive relative temperature for FBG1, FBG2, FBG3, FBG4,central FBG5, wavelength and FBG6, respectively. In addition, the When measuring the shifts of reflection centralspacing wavelength each FBG, we can obtain the4relative plantar pressure distribution. length of the FBGs is equal to the 1 cm. The channel of theofcentral wavelength is about nm to plantar pressure distribution. a fiber gratingofirradiates with a broadband source,When the grating will the reflect light avoidWhen a superimposition any two consecutive central optical wavelengths. measuring shifts of a central fiber corresponding grating irradiates with a broadband optical source,plantar theby grating will distribution. reflect light with aWhen wavelength twice the grating period (Λ) relative multiplied the effective refractive the reflection wavelength oftoeach FBG, we can obtain the pressure withWhen a( wavelength corresponding to core. twice grating period (Λ)source, multiplied by the effective refractive index the grating inirradiates the fiber central wavelength ( B ) of the reflected spectral band a fiber grating withThe athe broadband optical the grating will reflect light neff ) of index ( ) of the grating in the fiber core. The central wavelength ( ) of the reflected spectral band n with a wavelength corresponding twice the grating period (Λ) multiplied by the effective refractive eff by the Bragg B is defined condition to [39]: is defined Bragg condition [39]: index (neff ) by of the grating in the fiber core. The central wavelength (λ B ) of the reflected spectral band is B 2neff (1) defined by the Bragg condition [39]: B 2neff (1) Both the fiber refractive index ( neff ) and grating λ B the = 2n (1) eff · Λ pitch ( ) vary with the changes of strain Both the fiber refractive index ( ) and the grating pitch ( ) vary with the changes of strain n eff Bragg wavelength is sensitive to strain and temperature. and temperature, consequently the Both the fiber and refractive index (neff ) and the grating pitch (Λ) vary with the changes of strain and and temperature, and consequently the Bragg wavelength is grating sensitivewill to strain and temperature temperature. Even though the reflecting central wavelength of the issensor varytemperature. with temperature, and consequently the Bragg wavelength sensitive to strain and Even Even though the reflecting central wavelength of the sensor grating will vary with temperature slightly, the effect of temperature can be excluded. The reason is that an identification is determined though the reflecting central wavelength of the sensor grating will vary with temperature slightly, slightly, the effect oflocations temperature cancentral be excluded. The reason is thatfrom an identification is determined by comparative of excluded. the wavelengths measured the FBG sensors. That is, thethe effect of temperature can be The reason is that an identification is determined by the by the comparative locations of the central wavelengths measured from the FBG sensors. That is, even under different temperatures, the comparative shifts of the central wavelengths are unchanged comparative locations of the central wavelengths measured from the FBG sensors. That is, even under eventhe under different temperatures, the to comparative shifts of the central wavelengths are unchanged and results identified irrelevant temperature. different temperatures, theare comparative shifts of the central wavelengths are unchanged and the results and the results identified are irrelevant to temperature. identified are irrelevant to temperature.
Figure 2. The configuration of the fiber sensing system for plantar pressure measurement. Figure Figure 2. 2. The The configuration configuration of of the the fiber fiber sensing sensing system systemfor forplantar plantarpressure pressuremeasurement. measurement.
Figure 3. The reflection wavelength spectrum of six FBGs without any stress. Figure 3. The reflection wavelength spectrum of six FBGs without any stress. Figure 3. The reflection wavelength spectrum of six FBGs without any stress.
3. Methods and Results 3. Methods and Results
Sensors 2016, 16, 1766 Sensors 2016, 16, 1766
4 of 13 4 of 13
3.1.3.Subjects Methods and Results The primary aim of this work is to investigate the feasibility and functionality of the FBG sensor 3.1. Subjects system to measure the plantar pressure of the foot, and thus spatially resolve the foot pressure The primary aim of this work to the investigate feasibility functionality the FBG(nine sensor distribution at different regions suchisas forefoot,the midfoot andand hindfoot. Elevenofsubjects system measure theaged plantar theweighing foot, andfrom thus47spatially the foot in pressure males andto two females), 21 topressure 25 yearsofand to 83 kg,resolve were included this distribution at different regions such as the forefoot, midfoot and hindfoot. Eleven subjects (nine work. The Institutional Review Board of Kaohsiung Medical University Chung-Ho Memorial males and two females), aged this 21 tostudy 25 years weighing from 47 to 83consent. kg, were in this Hospital, Taiwan, has approved and and all subjects gave informed Allincluded participants work. The Institutional Review Board Kaohsiung University Chung-Ho Memorial Hospital, have been diagnosed to identify their footoftypes by the Medical digital pressure plate of the i-Step P1000 before Taiwan, has approved studypressure and all subjects consent. participants have been being subjected to FBG this plantar testing.gave The informed i-Step P1000 is aAll digital pressure plate to W identify their by the digital pressure plate of the the pressure i-Step P1000 before (20diagnosed in L × 18 in × 1.5 in H) foot withtypes 1024 barometric sensors to measure exerted by being the 2. Itplantar subjected FBG pressure testing. The i-Step P1000 is a digital plate (20 in L × 18 in foot every 1tocm is a current clinical foot pressure instrument and pressure is used as a measurement W × 1.5ininthis H) study. with 1024 to measure the1,pressure exerted by the foot every 1 cm2 . reference Thebarometric results aresensors summarized in Table in which there are three participants It isnormal a current footcavus pressure and is used as a measurement in thiswith study. with feet,clinical four with feet,instrument two with flat feet, one with supinated feet,reference and the other The results are summarized in Table 1, in which there are three participants with normal feet, four with a cavus left foot and a neutral right foot. Figure 4 shows the partial measurement results of foot cavus feet, two with flat feet, one with supinated feet, and the other with a cavus left foot and a neutral pressure distribution. right foot. Figure 4 shows the partial measurement results of foot pressure distribution. Table 1. The foot category of participants using the digital pressure plate i-Step 1000. Table 1. The foot category of participants using the digital pressure plate i-Step 1000. Foot Type No. Left Right Foot Type 1 neutral No. neutral Left Rightneutral 2 neutral 1 neutral neutral 3 neutral neutral 2 neutral neutralcavus 4 cavus 3 neutral neutralcavus 5 cavus 4 cavus cavuscavus 6 cavus 5 cavus cavus 7 cavus cavus 6 cavus cavus 8 flat flat 7 cavus cavus 9 flat 8 flat flat flat 10 supinated 9 flat flatsupinated 11 cavus supinated neutral 10 supinated 11 cavus neutral
(a)
(b) Figure 4. Cont.
(c)
Sensors 16,16, 1766 Sensors 2016, 1766 Sensors 2016, 16, 17662016,
(d)
5 of 13
(d)
5 of 1313 5 of
(e)
(e)
Figure 4. 4. Foot plantar pressure distribution ofof participants: No. (b) No. 6; (c) No. (d)(d) No. 10;10; Figure 4. Foot plantar pressure distribution ofdistribution participants: (a) No. 2; (b)(a) No. 6; 2; (c)2; No. 8; (d)6; No. 10;8; 8; Figure Foot plantar pressure participants: (a) No. (b) No. (c) No. No. (e)i-Step No. 11, by i-Step P1000 digital pressure plate. (e) No. 11, by P1000 digital pressure plate. (e) No. 11, by i-Step P1000 digital pressure plate.
Plantar Foot Sensing Area Definition 3.2. Plantar3.2. Foot Definition 3.2.Sensing Plantar Area Foot Sensing Area Definition In this paper, plantardistributions pressure distributions wereaccording analyzed to according to the Pedar-X in In this paper, pressure were analyzed the Pedar-X in system Inplantar this paper, plantar pressure distributions were analyzed according to system the Pedar-X system in 11 healthy subjects. The foot can be subdivided into four segments of the forefoot, midfoot, hindfoot 11 healthy subjects. Thesubjects. foot can The be subdivided four segments of the forefoot, midfoot, hindfoot 11 healthy foot can beinto subdivided into four segments of the forefoot, midfoot, hindfoot toes. In order to measurement set the six measurement areas, first we the measured thethe length of the participant's and toes. Inand order to set the six areas, first we measured length of participant's and toes. In order to set the six measurement areas, first we measured the length of the participant's Then,ofbythe means of the general definition, can get the threeofsegments the foot: front, middle, foot. Then, foot. byfoot. means general wedefinition, can get we the three foot: of front, middle, Then, by means ofdefinition, the general we can segments get the threethe segments of the foot: front, middle, and hind. For the length of any one of the feet, the hindfoot length accounts for 30%, the midfoot and hind. For thehind. length ofthe anylength one ofofthe hindfoot length accounts for accounts 30%, the for midfoot and For anyfeet, onethe of the feet, the hindfoot length 30%, the midfoot length 30%, the forefoot portion 25%, and the toes account for 15% [40,41]. In accordance with length is 30%, theisforefoot portion is 25%, andis toes account foraccount 15% [40,41]. In[40,41]. accordance with length is 30%, the forefoot portion isthe 25%, and the toes for 15% In accordance with the the average pressure of the measurement region, we defined six sensing sub-segments [41]: hindfoot the average pressure of the measurement region, we defined sensing [41]: hindfoot average pressure of the measurement region, wesix defined sixsub-segments sensing sub-segments [41]: hindfoot of of one, midfoot offorefoot two, and forefoot of three, without the toes, as illustrated in Figure 5. of one, midfoot of two, and of three, without the toes, as illustrated in Figure 5. one, midfoot of two, and forefoot of three, without the toes, as illustrated in Figure 5.
1.
1. Medial forefoot Medial forefoot
2.
2. Central forefoot Central forefoot
3.
3. Lateral forefoot Lateral forefoot
4.
4. Medial midfoot Medial midfoot
5.
5. Lateral midfoot Lateral midfoot
6.
Heel
6.
Heel
Figure The sensing area definitions this research. Figure 5. 5. The sixsix sensing area definitions inin this research. Figure 5. The six sensing area definitions in this research.
Figure 5, 5, these plantar pressure sensing areas are numbered 1 as the medial 2 as In Figure 5,Inthese plantar pressure sensing areas are numbered bynumbered 1 as the by medial forefoot, 2 forefoot, asforefoot, In Figure these plantar pressure sensing areas are by 1 as the medial 2 as central forefoot, 3 as the lateral forefoot, 4 as the medial midfoot, 5 as the lateral midfoot, and the central the forefoot, 3 as the lateral forefoot, 4 as the medial midfoot, 5 as the lateral midfoot, and 6 the central forefoot, 3 as the lateral forefoot, 4 as the medial midfoot, 5 as the lateral midfoot, and 66as as theheel. heel. When afoot different foot stands oninsole, the six regions will exhibit different as the heel. When a different type stands on the insole, the insole, six will exhibit different the When a different foot typetype stands on the theregions sixthe regions will exhibit different pressures. pressures. All subjects were asked to measure each foot five times, and then we took the average of pressures. All subjects were asked to measure each foot five times, and then we took the average of All subjects were asked to measure each foot five times, and then we took the average of the central central wavelength shift of each FBG sensor. According to the variation of the central wavelength the central the wavelength shift of each FBG sensor. According to the variation of the central wavelength wavelength shift of each FBG sensor. According to the variation of the central wavelength shift of the of the six FBG sensors, can get the relative distribution of the plantar pressure and determine shift of the shift six sensors, we can the relative distribution ofofthe pressure and determine sixFBG FBG sensors, we canget getwe the relative distribution theplantar plantar pressure and determine the subject's the subject's foot type. Figure 6 shows the prototype of the FBG sensors and a subject standing the subject's foot type. Figure 6 shows the prototype of the FBG sensors and a subject standing foot type. Figure 6 shows the prototype of the FBG sensors and a subject standing barefoot on the on the detection system. barefoot onbarefoot the detection system. detection system.
Sensors 2016, 1766 Sensors 2016, 16,16, 1766 Sensors 2016, 16, 1766
6 of 6 of 13 13 6 of 13
Figure 6. 6. Set-up of of FBG sensors and test in in action. Figure Set-up FBG sensors and test action. Figure 6. Set-up of FBG sensors and test in action.
3.3. Experimental Results 3.3. Experimental Results InIn order toto correspond toto the results ofof Figure 4, 4, the measured data from the fiber-optic sensing order correspond the results Figure the measured data from the fiber-optic sensing system for the participants are shown in this section. Figure 7 shows the plantar pressure distribution shows the the plantar plantar pressure pressure distribution system for the participants are shown in this section. Figure 7 shows ofof the left and right foot ofof the neutral feet ofof participant No. 2. 2. Figure 7a7a depicts the footprints and left and right foot ofthe the neutral feet of participant No. 2. Figure 7a depicts the footprints the left and right foot neutral feet participant No. Figure depicts the footprints and Figure 7b shows the central wavelength shift of the FBGs of the six sensing areas. In the figure, the and Figure 7b shows the central wavelength ofFBGs the FBGs of six thesensing six sensing areas. In figure, the figure, Figure 7b shows the central wavelength shiftshift of the of the areas. In the the solid represents the average values ofof five ofof each sensing element and the symbol solid symbol represents the average values ofmeasurements five measurements ofFBG each FBG sensing element solid symbol represents the average values five measurements each FBG sensing element and the line represents the measurement value range of the maximum and minimum. InIn this work, and the line represents the measurement value range ofmaximum the maximum and minimum. Inwork, thiseach work, the line represents the measurement value range of the and minimum. this each measured value slightly changes because the standing position and standing posture are not very each measured slightly changes because standing position andstanding standingposture postureare arenot not very very measured valuevalue slightly changes because thethe standing position and consistent. From Figure 7a,b, wewe can obtain the plantar pressure distribution atat allall areas ofof the foot, From Figure 7a,b, we can obtain the plantar pressure distribution at all areas of the foot, consistent. From Figure 7a,b, can obtain the plantar pressure distribution areas the foot, except forfor the medial midfoot area. That is,is, there is is nono foot pressure that can be detected by FBG4. except the medial midfoot area. That there foot pressure that can be detected by FBG4. pressure that can be detected by FBG4.
(a) (a)
(b) (b)
Figure 7. TheThe plantar pressure distribution of the neutral feet feet of participant No. 2: (a) 2: Footprint; (b) plantar pressure distribution the neutral participant No. (a) Footprint; Figure 7. The plantar pressure distribution ofofthe neutral feet of of participant No. 2: (a) Footprint; (b) central wavelength shift as a function of sensing area number. (b) central wavelength a function of sensing number. central wavelength shiftshift as aasfunction of sensing areaarea number.
Figure 8 shows the plantar pressure distribution ofof the cavus feet ofof participant No. 6, 6, one ofof the Figure distribution the cavus feet participant No. Figure 88 shows shows the the plantar plantar pressure pressure distribution of the cavus feet of participant No. 6, one one of the the four participants who has cavus foot. The cavus foot is is commonly known asas claw foot. The cavus foot four participants who has cavus foot. The cavus foot commonly known claw foot. The cavus foot four participants who has cavus foot. The cavus foot is commonly known as claw foot. The cavus foot type will make the foot hollow when bearing weight. AA high arch is is the opposite ofof a flat foot. Because type when bearing weight. high arch type will will make make the the foot foot hollow hollow when bearing weight. A high arch is the the opposite opposite of aa flat flat foot. foot. Because Because the arch area of the bow foot is relatively high, the pressure will be concentrated on the forefoot and the arch area of the bow foot is relatively high, the pressure will be concentrated on the forefoot and the arch area of the bow foot is relatively high, the pressure will be concentrated on the forefoot and heel areas. From Figure 8b, in the fifth sensing area, only a little pressure is sensed. The plantar heel areas. From sensing area, area, only only aa little little pressure pressure is is sensed. sensed. The plantar heel areas. From Figure Figure 8b, 8b, in in the the fifth fifth sensing The plantar pressure distributions are consistent among the results measured bybythe digital pressure plate ofof pressure distributions are consistent among the results measured the digital pressure plate pressure distributions are consistent among the results measured by the digital pressure plate of i-Step i-Step P1000 (Figure 4b), footprints (Figure 8a), and FBG fiber sensor (Figure 8b). i-Step 4b), footprints 8a), andfiber FBGsensor fiber sensor P1000 P1000 (Figure(Figure 4b), footprints (Figure(Figure 8a), and FBG (Figure(Figure 8b). 8b).
Sensors 2016, 16, 1766 Sensors Sensors2016, 2016,16, 16,1766 1766
7 of 13 7 7ofof1313
(a) (a)
(b) (b)
Figure 8.8.8. The plantar pressure distribution ofofthe cavus feet ofofparticipant No. Footprint; (b) Figure The plantar pressure distribution of the cavus of participant No. (a) Footprint; Figure The plantar pressure distribution the cavus feetfeet participant No.6: 6:(a) (a)6: Footprint; (b) central wavelength shift ofofsensing area (b) central wavelength shift as a function of sensing area number. central wavelength shiftas asaafunction function sensing areanumber. number.
Flat feet occur because the holding the joints in the foot together (called tendons) are Flat thetissues tissuesholding holdingthe thejoints joints together (called tendons) are Flatfeet feet occur occur because because the tissues in in thethe footfoot together (called tendons) are loose. loose. As a result, tendons do not pull together properly; there is little or no arch. Figure 9 depicts the loose. As a result, tendons do not pull together properly; there is little or no arch. Figure 9 depicts the As a result, tendons do not pull together properly; there is little or no arch. Figure 9 depicts the plantar plantar pressure distribution of a flat foot. It differs from the others foot types; that is, there will be plantar pressure distribution a flat It from differs from thefoot others footthat types; that is, there will be pressure distribution of a flatoffoot. It foot. differs the others types; is, there will be pressure pressure distribution on the medial arch area, and there will be no arch shape. The measurement pressure distribution on the medial arch area, and there will be no arch shape. The measurement distribution on the medial arch area, and there will be no arch shape. The measurement result of the result of flat No. 88of sensing system isisshown in 9b. result ofthe the flatfoot footof ofparticipant participant ofthe thefiber-optic fiber-optic sensing systemin shown9b. inFigure Figure 9b.The Theof flat foot of participant No. 8 of theNo. fiber-optic sensing system is shown Figure The pressure pressure of the fourth region (the medial midfoot) with the green circle is not zero. That phenomenon pressure of the fourth medialwith midfoot) withcircle the green not zero. That phenomenon the fourth region (theregion medial(the midfoot) the green is notcircle zero.isThat phenomenon is different isisdifferent from the types, shows aasimilar pressure as measured different fromfoot theother otherfoot foot types,and and shows similar pressuredistribution distribution asthe the measured from the other types, and shows a similar pressure distribution as the measured results of the results of the digital pressure plate of i-Step P1000 (see Figure 4c). results the digital pressure plate of i-Step P10004c). (see Figure 4c). digitalofpressure plate of i-Step P1000 (see Figure
(a) (a)
(b) (b)
Figure 9. The plantar pressure distribution of the flat feet ofofparticipant No. 8:8:(a) Footprint; (b) central Figure Figure9.9.The Theplantar plantarpressure pressuredistribution distributionofofthe theflat flatfeet feet ofparticipant participantNo. No. 8:(a) (a)Footprint; Footprint;(b) (b)central central wavelength shift as a function of sensing area number. wavelength wavelengthshift shiftasasaafunction functionofofsensing sensingarea areanumber. number.
In Inthis thiswork, work,the thefootprint footprintand andfiber-optic fiber-opticsensing sensingresult resultof ofthe theunique uniqueparticipant participantwith withsupinated supinated In this work, the footprint and fiber-optic sensing result of the unique participant with supinated feet is shown in Figure 10. Supination or under-pronation is a condition that occurs when feet is shown in Figure 10. Supination or under-pronation is a condition that occurs whenfeet feetroll roll feet is shown in Figure 10. Supination or under-pronation is afoot condition that occurs when feet roll outwards, placing weight on the outside of the foot. While the leans to the outside, weight outwards, placing weight on the outside of the foot. While the foot leans to the outside, weightisis outwards, placing weight onAthe outside ofsupinated the foot. While the foot leans to the outside,shockweight distributed distributedalong alongthe theoutside. outside. Aperson personwith with supinatedfeet feetwill willreduce reducethe thebody's body'snatural natural shockis distributed along the outside. A person with supinated feet will reduce the body's natural absorbing absorbingcapability. capability.From FromFigure Figure10, 10,we wecan cansee seethat thatthere thereisisno nopressure pressuredetection detectionon onarea area4,4,and and shock-absorbing capability. From This Figure 10, wepressure can see that there is no pressure detection on area 4, area 1 only has very light pressure. plantar distribution occurs at the foot supination. area 1 only has very light pressure. This plantar pressure distribution occurs at the foot supination. and areashows 1 only hasexperimental very light pressure. This plantar pressure distribution occurs at the foot supination. Figure Figure11 11 showsthe the experimentalresults resultsof ofparticipant participantNo. No.11. 11.One Oneof ofhis hisfeet feet(left (leftfoot) foot)isispes pescavus cavus and andthe theother otherisisaaneutral neutralfoot. foot.From Fromthis thisfigure, figure,we wecan caneasily easilydistinguish distinguishthe thefoot foottypes. types.
Sensors 2016, 16, 1766
8 of 13
Sensors 2016, 16, 1766 111766 shows SensorsFigure 2016, 16,
8 of 13 the experimental results of participant No. 11. One of his feet (left foot) is 8pes of 13cavus and the other is a neutral foot. From this figure, we can easily distinguish the foot types.
(a) (a)
(b) (b)
Figure 10. The plantar pressure distribution of supinated feet of participant No. 10: (a) Footprint; (b) Figure 10. The pressure distribution of supinated feet of participant No. 10: (a) Footprint; (b) Figure 10.plantar Theshift plantar of supinated central wavelength as apressure function distribution of sensing area number. feet of participant No. 10: (a) Footprint; central shift as ashift function of sensing number. (b)wavelength central wavelength as a function ofarea sensing area number.
(a) (a)
(b) (b)
Figure 11. The pressure distribution of participant No. 11, left foot cavus and right Figure 11. plantar The plantar pressure distribution of participant No.whose 11, whose left is foot is cavus and right Figure 11. The plantar pressure distribution of participant No. 11, whose left foot is cavus and right foot is neutral: (a) Footprint; (b) central wavelength shift as a function of sensing area number. foot is neutral: (a) Footprint; (b) central wavelength shift as a function of sensing area number. foot is neutral: (a) Footprint; (b) central wavelength shift as a function of sensing area number.
3.4. Sensing Stability 3.4. Sensing Stability 3.4. Sensing Stability In Figure 12, we demonstrate the sensing stability of the experimental measurement results of In Figure 12,demonstrate we demonstrate the sensing stability the experimental measurement results In Figure 12, we the sensing stability of theofexperimental measurement results of of the FBG sensors in four different foot foot types. Each subject’s foot was measured five times in this study. the sensors FBG sensors in different four different types. subject’s footmeasured was measured five times this study. the FBG in four foot types. Each Each subject’s foot was five times in thisinstudy. Although the result of each measurement is slightly changed due to the inconsistence of the standing Although the result of each measurement is slightly changed due to the inconsistence of the standing Although the result of each measurement is slightly changed due to the inconsistence of the standing position and posture each time, the overall foot pressure distribution for the foot-type classification position and posture the overall foot pressure distribution forfoot-type the foot-type classification position and posture each each time,time, the overall foot pressure distribution for the classification will will not not be affected. If the object has flat feet, it is inevitable inevitabletotodetect detectfoot foot pressure in the fourth be affected. If the object has flat feet, it is pressure in the fourth sensing will not be affected. If the object has flat feet, it is inevitable to detect foot pressure in the fourth sensing region (Figure 12a). The larger the foot pressure that is measured in the fourth sensing area, region (Figure 12a). The larger the foot pressure that is measured in the fourth sensing area, the sensing region (Figure 12a). The larger the foot pressure that is measured in the fourth sensing area,more the more severe the flat foot situation. In contrast, there is no foot pressure in the fourth sensing areaif the severe the flat situation. In contrast, there is no foot pressure inin the the more severe the foot flat foot situation. In contrast, there is no foot pressure thefourth fourthsensing sensingarea area if theparticipant participant has supinated feet. At the same time, the plantar pressure is almost close to zero in has supinated feet. At the same time, the plantar pressure is almost close to zero in if the participant has supinated feet. At the same time, the plantar pressure is almost close to zerothe in first the first sensing area for a supinated foot (Figure 12b). Moreover, if the participant has cavus feet, sensing areaarea for afor supinated foot foot (Figure 12b).12b). Moreover, if theifparticipant has cavus feet, there the first sensing a supinated (Figure Moreover, the participant has cavus feet, will therebewill be little foot pressure in the fifth sensing region (Figure 12c). Of course, the sensing value little in thein fifth regionregion (Figure 12c). Of course, the sensing value value of the foot there will be foot littlepressure foot pressure thesensing fifth sensing (Figure 12c). Of course, the sensing of the foot pressure of themidfoot medialarea midfoot area is Other also zero. Other thansituations, the above situations, the pressure of the medial is also zero. than the above the sensitive stability of the foot pressure of the medial midfoot area is also zero. Other than the above situations, the sensitive stability of aisneutral foot is shown in Figure 12d. of a neutral shownfoot in Figure 12d.in sensitive stabilityfoot of a neutral is shown Figure 12d.
Sensors 2016, 16, 1766 Sensors 2016, 16, 1766
9 of 13 9 of 13
(a)
(b)
(c)
(d) Figure 12. The central wavelength shift as a function of sensing area number for plantar pressure Figure 12. The central wavelength shift as a function of sensing area number for plantar pressure distribution of left and right foot: (a) Flat feet; (b) supinated feet; (c) cavus feet; and (d) neutral feet. distribution of left and right foot: (a) Flat feet; (b) supinated feet; (c) cavus feet; and (d) neutral feet.
Sensors 2016, 16, 1766 Sensors 2016, 16, 1766
10 of 13 10 of 13
Some the abnormal plantar pressure pattern, which is is related toto obesity Someliterature literaturehas hasdiscussed discussed the abnormal plantar pressure pattern, which related obesity ororoverweight, and can be identified with its foot type by visual non-quantitative inspection. Most ofof overweight, and can be identified with its foot type by visual non-quantitative inspection. Most the articles plantar pressure single the articlesonon plantar pressuresensors sensorsdid didnot notnormalize normalizethe thesensor sensoroutput outputwith withrespect respecttotothe the single body mass value. In this work, two subjects with flat feet of different weights, 62 kg in Figure 9b body mass value. In this work, two subjects with flat feet of different weights, 62 kg in Figure 9band and 6565kgkgininFigure Figure12a, 12a,show showthe thenon-zero non-zeropressure pressureatatthe themidfoot midfootregion. region.Two Twosubjects subjectswith withcavus cavusfeet feet ofofdifferent weights, 71 kg in Figure 8b and 52 kg in Figure 12c, show similar pressure distribution different weights, 71 kg in Figure 8b and 52 kg in Figure 12c, show similar pressure distribution patterns. curve vertically verticallyand andslightly slightlychange changethe thecurve curve pattern patterns.The Thesubject’s subject’sweight weightcan can offset offset the the curve pattern but but not the foot-type classification. not the foot-type classification. 3.5. Verification 3.5.Statistical Statistical Verification The product-moment linear ThePearson Pearson product-momentcorrelation correlationcoefficient coefficient(PPMCC (PPMCCororPCC) PCC)isisa ameasure measureofofthe the linear correlation between two variables [42]. A value of zero indicates that there is no association between correlation between two variables [42]. A value of zero indicates that there is no association between the order the ofofthe thetwo twovariables. variables.InIn ordertotounderstand understand therelationship relationship themeasurement measurementdata datafrom fromthe thedigital digital pressure i-Step P1000 P1000and and fiber-optic sensing system, wePearson’s used Pearson’s correlation pressureplate plate i-Step thethe fiber-optic sensing system, we used correlation coefficient coefficient method. we calculated the values averageofvalues digital pressure on 3ancm area cm in 2i-Step cm method. First, we First, calculated the average digitalofpressure on an area × 23 cm inP1000. i-Step This P1000. This area corresponds one the grating six fiber gratingareas, sensing areas, the central area corresponds to one oftothe sixoffiber sensing the central wavelength wavelength shift of which is measured. Figure 13 shows the correlation of all central wavelength shift of which is measured. Figure 13 shows the correlation of all central wavelength shifts of the six shifts the sixinFBG in this and the plantar pressure valuespressure of the digital FBGof sensors thissensors study and thestudy plantar pressure values of the digital platepressure of i-Stepplate P1000. ofThen, i-Stepwe P1000. Then, we the Pearsoncoefficient correlation coefficient and got a r-value 0.671which (p < calculated thecalculated Pearson correlation and got a r-value of 0.671 (p
