PERSPECTIVE
Evaluation of the Clinical Photographs in the Journal of Oral and Maxillofacial Surgery: From Readers’ Perspectives Tianfu Wu, DDS,* Sulin Chen, MDS,y and Xuepeng Xiong, DDSz Purpose:
This study was designed to evaluate clinical photographs published in the Journal of Oral and Maxillofacial Surgery (JOMS) and understand the current status of oral and maxillofacial surgery.
Materials and Methods:
A total of 1,317 photographs from the JOMS Volume 69 were assessed. These photographs were scored from 1 to 10 for the following parameters: sharpness; depth of field; exposure; composition; color or grayscale; background; position; distortion; label consistency; and white balance. Then, the distributions of scores were analyzed. Each score was compared with the average score. The effects of different subjects; emergency or nonemergency situations; and intraoperative, preoperative, or postoperative conditions on the quality of photographs were analyzed by conducting a nonparametric Kruskal-Wallis test.
Results:
The total score of each photograph showed a left-skewed distribution, varying from 3 to 10, with an average score of 6.82. Four parameters, including sharpness, depth of field, exposure, and white balance, scored less than the average score. Photographs with an intraoral subject yielded the lowest score, with a significant difference (P < .05). The score of photographs taken during a nonemergency situation was significantly higher than that during an emergency situation (6.84 vs 6.03; P < .001). Photographs of an intraoperative condition yielded a score significantly lower than those of pre- and postoperative conditions (6.53 vs 7.11 and 6.75, respectively; P < .001). Approximately 45.5% of photographs (148 of 325) displayed uncovered eyes and 57.1% of specimens (40 of 70) did not appear with a plotting scale. Conclusions: Sharpness, depth of field, exposure, and white balance should be considered to a greater extent than the other parameters when oral and maxillofacial photographs are taken, particularly for intraoral conditions, emergency situations, and intraoperative conditions. Enhanced parameters and protection of a patient’s identity may significantly improve the average level of photographic quality. Crown Copyright Ó 2014 Published by Elsevier Inc on behalf of the American Association of Oral and Maxillofacial Surgeons. All rights reserved J Oral Maxillofac Surg 72:449-455, 2014
Clinical photographs in journals are important to allow readers to understand a patient’s initial presentation, treatment outcomes, and new surgical techniques.1-3 In addition to clinical understanding, clarity should be considered to provide high-quality clinical photographs
with correct exposure, white balance, proper depth of field, and background. For legal considerations, patients’ eyes in the photographs should be covered to protect their identity. (It has long been the policy of the Journal of Oral and Maxillofacial Surgery
Received from the State Key Laboratory Breeding Base of Basic
Address correspondence and reprint requests to Prof Xiong:
Science of Stomatology (Hubei-MOST) and Key Laboratory of Oral
Department of Oral and Maxillofacial Surgery, School and Hospital
Biomedicine Ministry of Education, School and Hospital of
of Stomatology, Wuhan University, Luo Yu Road #237, Wuhan
Stomatology, Wuhan University, Wuhan, China.
430072, People’s Republic of China; e-mail: [email protected]
*Resident, Department of State Key Laboratory Breeding Base of
Received June 1 2013
Basic Science of Stomatology. yResident, Oncology.
Department
Accepted June 27 2013 of
Oral
Maxillofacial-Head
Crown Copyright Ó 2014 Published by Elsevier Inc on behalf of the American
Neck
Association of Oral and Maxillofacial Surgeons. All rights reserved
zAssociate Professor; Department of Oral Maxillofacial-Head
0278-2391/13/00869-0$36.00/0
Neck Oncology.
http://dx.doi.org/10.1016/j.joms.2013.06.224
Drs Wu and Chen contributed equally to this work.
449
450
READERS’ PERSPECTIVES OF CLINICAL PHOTOGRAPHS
[JOMS] to require all patient-identifying information be removed or masked. Signed patient releases must accompany manuscripts in which there are photographs of identifiable patients unless eyes are masked to prevent identification.) With technologic advancements, simpler and more practical cameras with higher performance have been developed.4 For instance, the single lens reflex camera with an optional lens, flash lamp, aperture, and shutter provide better performance.5 However, the authors found some published medical journals with low-quality clinical photographs. Although it should be determined whether these lowquality clinical photographs can affect the readers’ understanding of a specific study, the authors recommend that high-quality clinical photographs be provided in journals. Diseases or trauma in the oral and maxillofacial region are apparent to readers compared with other clinical fields; hence, this visual nature optimizes the usefulness of photography, particularly in small anatomic regions, including the teeth, jaws, and different types of mucosa. To produce standardized photographs, a photographer requires proper training. However, oral and maxillofacial surgeons are busy and may be unable to allot extra time to study photography. A limited surgical field, complex anatomy, and emergency situations result in unsatisfactory visual reproductions of oral and maxillofacial surgery; therefore, visual reproduction should be improved.6,7 The frequency of low-quality photographs published in medical journals is seldom addressed. As a mainstream journal in the field of oral and maxillofacial surgery, the JOMS publishes articles with numerous photographs. In this study, 1,317 clinical photographs in the JOMS Volume 69 were evaluated and scored to determine the quality of such photographs and the frequency of low-quality photographs. This study also provided possible solutions to lessen common defects. The results may help improve the average quality of photographs and allow clinicians to take photographs of higher quality.
ASSESSMENT CONDITION
Materials and Methods INCLUSIVE AND EXCLUSIVE CRITERIA OF IMAGES
Consecutive photographs in Issues 1 to 12 of JOMS Volume 69, except Supplementary Issue 9 (which featured summaries and outlines that did not contain any photographs), were evaluated and scored one after another. Photographs in basic research or those irrelevant to oral and maxillofacial surgery were excluded, such as animal models, histopathologic images, immunofluorescence images, fiberscope views, life portraits, and computer-generated photographs. A total of 1,317 photographs were evaluated and scored.
Photographs printed in journals rather than electronic images downloaded from a database were used to avoid significant chromatic aberration from different computer screens. These images were assessed in the authors’ library under natural light. Photographs were assessed by 3 senior oral and maxillofacial surgeons with more than 5 years of clinical practice and familiarity with basic photographic terminologies. As these surgeons assessed the photographs, they were allowed to relax for 5 minutes at an interval of 30 minutes to avoid eyestrain. SCORE INDICATORS
Considering previous studies,7,8 10 indicators were selected, including sharpness, depth of field, light exposure, composition, color, background, position, distortion, label consistency, and white balance to evaluate the photographs. Each indicator was judged according to the following factors. 1) Sharpness refers to the outline of the focal point, which should be clear and sharp without blur. 2) The depth of the field is the distance from the front region to the back region of the focus. 3) Composition is the relative location between the point of interest and other objects. 4) Color was scored based on a colored or grayscale photograph. The background was given a higher score if it was gray or another color that did not influence the color of the main photograph. 5) Position corresponds to the posture of the limbs, body, or photographer. 6) Distortion refers to the undesired change in the shape or angle of the subject in a photograph. The same photograph labels or legends should be used consistently as the photographs are described in an article. 7) White balance is the process of adjusting the color of the light source to produce more realistic photographs. SCORING CRITERIA
Each indicator was assigned a score of 0 or 1 according to the aforementioned criteria. Each indicator was scored 1 if the photographs exhibited clear sharpness with good focus; adequate depth of field; appropriate light exposure; the point of interest in the composition across the center of the photograph; colored photographs; background highlighting of the object without affecting the object’s color or deviating the readers’ attention; the object or posture oriented at a horizontal or vertical direction except asymmetrically for cases with some specific diseases no inappropriate stretching or incorrect photographic angle between the camera and the object; appropriate photograph labeling; and appropriate white balance. Otherwise, the corresponding indicator was given the score of 0. Representative examples are shown in
451
WU, CHEN, AND XIONG
Figure 1.9-14 The sum of scores varying from 0 to 10 was defined as the final score of a specific photograph. STATISTICS
Data were analyzed in GraphPad Prism 5.01 (GraphPad, San Diego, CA). Nonparametric Kruskal-Wallis test was performed to compare the difference between groups. A P value less than .05 was considered statistically significant. Hierarchical clustering was performed using Cluster 3.0 (http://bonsai.hgc.jp/mdehoon/ software/cluster/) and visualized using TreeView 1.1.6r2 (http://jtreeview.sourceforge.net).
Results SCORE DISTRIBUTION AND INDICATOR VALUE
The total score of each photograph varied from 3 to 10 (mean score, 6.82 1.35) and exhibited a leftskewed distribution (Fig 2A). The average score of each indicator in the photographs varied from 0.045 to 1.0 (average score, 0.68; Fig 2B). The average scores were compared and picture composition, color, background, object position, distortion, and labeling
yielded an average score higher than 0.68. In contrast, sharpness, depth of field, exposure, and white balance (SDEW) yielded scores lower than 0.68 (Fig 2B). Hierarchical clustering showed that SDEW had similar low scores (Fig 2G). These results suggested that such indicators were unsatisfactory in oral and maxillofacial surgery photography. Among these indicators, depth of field obtained the lowest mean score, indicating that a typical oral and maxillofacial surgeon has difficulty in adjusting this parameter. In contrast, color showed the highest score (Fig 2B). POTENTIAL FACTORS RESPONSIBLE FOR IMAGE QUALITY
Different complexities in the content and subject may influence photographers and the quality of photographs. In this study, photographs were divided into 6 groups according to differences in content, including portrait, model, instrument, specimen, intraoral image, and extraoral image. The model group showed the highest score (7.97 1.1), and the intraoral group showed the lowest (6.50 1.18). Scores of the other groups were similar (Fig 2C). Kruskal-Wallis test
FIGURE 1. Representative examples. A, Inappropriate focus leads to poor photographic sharpness.9 B, Inadequate field depth leads to a fuzzy view of the distal bucca and mesial alveolar ridge mucosae.10 C, The image is underexposed.11 D, The point of interest is not in the center of the photograph.12 E, Photograph is printed in grayscale.13 F, The positions of the object and the ruler are skewed.14 Wu, Chen, and Xiong. Readers’ Perspectives of Clinical Photographs. J Oral Maxillofac Surg 2014.
452
READERS’ PERSPECTIVES OF CLINICAL PHOTOGRAPHS
FIGURE 2. Statistical analysis of indicators and photograph scores. A sum of the indicator scores in 1 photograph corresponded to a photograph score. A, Score distribution of photographs. B, Mean standard deviation for each indicator score of each photograph. C, Group comparison of contents (model, instrument, portrait, specimen, extraoral, and intraoral groups), emergency (nonemergency and emergency groups), and hospital duration (pre-, intra-, and postoperative groups). D, Indicator score for intraoral group. E, Indicator score for emergency group. F, Indicator score for intraoperative group. G, Hierarchical clustering of 10 indicators in 34 countries: a single indicator from the same country was arithmetically averaged; rows represent indicators and columns represent countries; centroid linkage clustering was performed in rows and columns respectively; light red, scored 1; light green, scored 0. B, D, E, F, Dotted lines indicate the mean value. ***P < .001. (Fig 2 continued on next page.) Wu, Chen, and Xiong. Readers’ Perspectives of Clinical Photographs. J Oral Maxillofac Surg 2014.
453
WU, CHEN, AND XIONG
FIGURE 2 (cont’d). Wu, Chen, and Xiong. Readers’ Perspectives of Clinical Photographs. J Oral Maxillofac Surg 2014.
showed significant differences among the 6 groups (P < .0001; Fig 2C). Significant differences also were observed between the model and intraoral groups, between the model and portrait groups, and between the portrait and instrument groups (P < .001; data not shown). The highest score for the model group was acceptable because models are easy to capture. The lowest score for the intraoral group indicated that good intraoral photographs are difficult to take. Different situations in hospitals, such as traffic trauma, acute allergic reaction, and hemorrhage, may influence photographers and the quality of photographs. In this study, photographs were categorized as emergency and nonemergency groups. The significantly different mean scores of the emergency and nonemergency groups were 6.03 1.89 and 6.84 1.33, respectively (P < .001; Fig 2C). Given that a patient’s condition can change during different hospital stays depending on disease progression and treatment interference, the quality of the photographs may be affected. Therefore, the photographs were categorized into preoperative, intraoperative, and postoperative groups. Differences in scores were analyzed and the average scores for the 3 groups were 7.11 1.36, 6.53 1.39, and 6.75 1.23, respectively. The intraoperative group yielded significantly lower scores than the pre- and postoperative groups (P < .001; Fig 2C). Moreover, the popularity and use of high-grade cameras may vary among different countries and influence the distribution of indicator scores. To further identify the relation of potential scores among different countries and the potential distribution characteristics of indicator scores, hierarchical clustering analysis of indicator scores and indicators in each country was performed. As shown in Figure 2G, the indicator score groups were separated into 2 clusters, namely SDEW and other. These 2 clusters showed a significant differ-
ence (P < .05) based on independent-samples MannWhitney U test. However, no obvious aggregations in country clustering (Fig 2G) were found. Whether high-grade camera use influences the indicator score distribution awaits further investigation. In summary, the intraoral, emergency, and intraoperative groups showed the lowest scores in the corresponding analyses. As mentioned previously, SDEW showed scores lower than the average score for the other indicators; therefore, the authors investigated whether these indicators were the main problems in these 3 groups that prevented surgeons from taking good photographs. A similar method of analysis was used in which the unsatisfactory indicators yielded scores lower than the average (Fig 2, dotted line). Figures 2D, 2E, and 2F show that, in addition to SDEW, the main contributor to the poor performance of the 3 groups, the lower score for background may have affected the quality of photographs in the emergency and intraoperative groups. EYE COVERING AND PLOTTING SCALE IN PHOTOGRAPHS
During the scoring process, some portrait photographs did not include an eye cover and some specimen photographs did not include a scale bar. Therefore, another statistical analysis for eye cover and scale bar was performed. Approximately 45.5% of portraits (148 of 325) displayed uncovered eyes and 57.1% of specimens (40 of 70) did not appear with a scale bar.
Discussion This study aimed to determine the indicators that should be considered in oral and maxillofacial surgery photography. The authors found that the problems in quality mainly involved SDEW. Most photographs showed a score of 5 to 9 (Fig 2A), and only 1.7% of
454 photographs yielded perfect scores for the indicators considered in this study. Furthermore, the quality of photographs was affected by other factors, including the subject of photography, emergency or nonemergency situation, and different perioperative periods. The quality of an image depends on several factors, such as chromatic aberration during printing and the use of different cameras. Therefore, numerous indicators might be considered. In this study, 10 indicators with essential elements and some intermediate indicators that contribute to produce a high-quality photograph that an ordinary reader can understand were selected. This review is the first to describe the quality of photographs from the readers’ perspective; therefore, efficient changes in assessment should be implemented to improve future studies. A low indicator score indicates poor image quality, which may lessen the reproduction and presentation quality of a lesion or a scene. Inadequate sharpness, shallow depth of field, poor composition, and overexposure will lead to blurred lesion details, indistinct lesion edges and junctions between lesions and normal tissues, a poor visual field, and blown highlights. Lesions caused by lichen planus, vascular tumor, and Fordyce disease often have distinct colors. Underexposure, black-and-white images, inappropriate background selection, and adjusted white balance will result in false colors and hinder a reader’s ability to appreciate the actual lesions. In addition, the lack of a scale bar or an object to serve as a size reference in specimen photographs will make visualizing the size of a lesion difficult for readers. Studies have striven to improve the specialized techniques in medical photography by using more efficient equipment4,5 and professional cameras.6 Although these methods significantly improve the levels of photographer and image quality, common defects may be neglected during this process, particularly when surgeons are in a stressful situation. Hence, the authors determined the common errors observed in images during oral surgery to help avoid such errors. The difference between groups is shown in Figure 2, which illustrates the common defects in oral and maxillofacial surgery photography. Photographs of portraits, models, and instruments are usually taken in a normal situation, allowing clinicians to arrange the object locations and environmental conditions, such as light and background. A patient’s consent and cooperation are obtained. As expected, these indicators yielded higher scores than those for intraoral photographs, in which light and background are difficult to adjust.15 For the emergency group, a background showing bloodstains and a critical situation may result in lower scores. Similarly, the score for the intraoperative group was lower than those for the pre- and postoperative groups.
READERS’ PERSPECTIVES OF CLINICAL PHOTOGRAPHS
Several common photograph defects were found in this study. However, such defects may not evidently influence the main content in some published images. Other unnoticed defects might exert more influence on quality. This deviation was statistically minimized by evaluating a relatively large sample of 1,317 photographs. Hence, the findings could reflect a relatively realistic state of photographic quality. Patient privacy should be protected during image processing. Midlevel and senior oral and maxillofacial surgeons are accustomed to seeing maxillofacial trauma, tumors, and facial transfiguration before and after surgery. However, patients and their friends may be unfamiliar or uncomfortable with such images, which is the reason most patients prefer to conceal their lesions or the images of such lesions to avoid scrutiny or untoward comments. The Health Insurance Portability and Accountability Act Privacy Rule gives individuals the right to protect their health information and the right to file a complaint to the Department of Health and Human Services Office for Civil Rights if their privacy is violated. Although oral and maxillofacial surgeons may be unwilling to modify surgery images, they can avoid lawsuits by obtaining patients’ informed consent before publishing patientsensitive data or by adding an image element that conceals the eyes of patients in photographs to prevent identification. In summary, although most photographs published in the JOMS exhibit high quality, some common unsatisfactory errors remain. Paying additional attention to common defects in SDEW might significantly improve the average quality of photographs and help to better reproduce and record some valuable scenes and moments in oral and maxillofacial surgery. Acknowledgments The authors thank Professor Wenfeng Zhang, Sangang He, and Jun Jia for their careful evaluation of the photographs and the library staff for their assistance in data collection.
References 1. Becker DG, Tardy ME Jr: Standardized photography in facial plastic surgery: Pearls and pitfalls. Facial Plast Surg 15:93, 1991 2. Sommer DD, Mendelsohn M: Pitfalls of nonstandardized photography in facial plastic surgery patients. Plast Reconstr Surg 114: 10, 2004 3. Riml S, Piontke A, Larcher L, et al: Quantification of faults resulting from disregard of standardised facial photography. J Plast Reconstr Aesthet Surg 64:898, 2011 4. Liu F: [Techniques of oral cavity photography II. The choice of camera and appliance for dentistry]. Zhonghua Kou Qiang Yi Xue Za Zhi 42:190, 2007 (in Chinese) 5. Peck JJ, Roofe SB, Kawasaki DK: Camera and lens selection for the facial plastic surgeon. Facial Plast Surg Clin North Am 18: 223, 2010 6. Sackstein M: [Intra-oral digital photography with the non professional camera—Simplicity and effectiveness at a low
WU, CHEN, AND XIONG
7.
8.
9.
10.
price]. Refuat Hapeh Vehashinayim 24:19, 2006 (in Hebrew) Liu F: [Techniques of oral cavity photography I. The basics of oral cavity photography]. Zhonghua Kou Qiang Yi Xue Za Zhi 42:60, 2007 (in Chinese) Han L, Zhou ZB: [Techniques of oral cavity photography V. The photography of oral maxillofacial surgery, part one]. Zhonghua Kou Qiang Yi Xue Za Zhi 42:575, 2007 (in Chinese) Furr MC, Larkin E, Blakeley R, et al: Extending multidisciplinary management of cleft palate to the developing world. J Oral Maxillofac Surg 69:237, 2011 Collette DC, Zechel MA: Novel treatment of atypical human papillomavirus-associated epithelial hyperplasia with cidofovir. J Oral Maxillofac Surg 69:2383, 2011
455 11. Zhang S, Liu X, Xu Y, et al: Application of rapid prototyping for temporomandibular joint reconstruction. J Oral Maxillofac Surg 69:432, 2011 12. Ellis E III: A study of 2 bone plating methods for fractures of the mandibular symphysis/body. J Oral Maxillofac Surg 69:1978, 2011 13. Wada T, Nakatani K, Hiraishi Y, et al: Usefulness of myofascial flap without skin in contemporary oral and maxillofacial reconstruction. J Oral Maxillofac Surg 69:1815, 2011 14. Movahed R, Weiss A, Velez I, et al: Submandibular gland MALT lymphoma associated with Sjogren’s syndrome: Case report. J Oral Maxillofac Surg 69:2924, 2011 15. Ahmad I: Digital dental photography. Part 8: Intra-oral set-ups. Br Dent J 207:151, 2009
Evaluation of the Clinical Photographs in the Journal of Oral and Maxillofacial Surgery: From Readers’ Perspectives Tianfu Wu, DDS,* Sulin Chen, MDS,y and Xuepeng Xiong, DDSz Purpose:
This study was designed to evaluate clinical photographs published in the Journal of Oral and Maxillofacial Surgery (JOMS) and understand the current status of oral and maxillofacial surgery.
Materials and Methods:
A total of 1,317 photographs from the JOMS Volume 69 were assessed. These photographs were scored from 1 to 10 for the following parameters: sharpness; depth of field; exposure; composition; color or grayscale; background; position; distortion; label consistency; and white balance. Then, the distributions of scores were analyzed. Each score was compared with the average score. The effects of different subjects; emergency or nonemergency situations; and intraoperative, preoperative, or postoperative conditions on the quality of photographs were analyzed by conducting a nonparametric Kruskal-Wallis test.
Results:
The total score of each photograph showed a left-skewed distribution, varying from 3 to 10, with an average score of 6.82. Four parameters, including sharpness, depth of field, exposure, and white balance, scored less than the average score. Photographs with an intraoral subject yielded the lowest score, with a significant difference (P < .05). The score of photographs taken during a nonemergency situation was significantly higher than that during an emergency situation (6.84 vs 6.03; P < .001). Photographs of an intraoperative condition yielded a score significantly lower than those of pre- and postoperative conditions (6.53 vs 7.11 and 6.75, respectively; P < .001). Approximately 45.5% of photographs (148 of 325) displayed uncovered eyes and 57.1% of specimens (40 of 70) did not appear with a plotting scale. Conclusions: Sharpness, depth of field, exposure, and white balance should be considered to a greater extent than the other parameters when oral and maxillofacial photographs are taken, particularly for intraoral conditions, emergency situations, and intraoperative conditions. Enhanced parameters and protection of a patient’s identity may significantly improve the average level of photographic quality. Crown Copyright Ó 2014 Published by Elsevier Inc on behalf of the American Association of Oral and Maxillofacial Surgeons. All rights reserved J Oral Maxillofac Surg 72:449-455, 2014
Clinical photographs in journals are important to allow readers to understand a patient’s initial presentation, treatment outcomes, and new surgical techniques.1-3 In addition to clinical understanding, clarity should be considered to provide high-quality clinical photographs
with correct exposure, white balance, proper depth of field, and background. For legal considerations, patients’ eyes in the photographs should be covered to protect their identity. (It has long been the policy of the Journal of Oral and Maxillofacial Surgery
Received from the State Key Laboratory Breeding Base of Basic
Address correspondence and reprint requests to Prof Xiong:
Science of Stomatology (Hubei-MOST) and Key Laboratory of Oral
Department of Oral and Maxillofacial Surgery, School and Hospital
Biomedicine Ministry of Education, School and Hospital of
of Stomatology, Wuhan University, Luo Yu Road #237, Wuhan
Stomatology, Wuhan University, Wuhan, China.
430072, People’s Republic of China; e-mail: [email protected]
*Resident, Department of State Key Laboratory Breeding Base of
Received June 1 2013
Basic Science of Stomatology. yResident, Oncology.
Department
Accepted June 27 2013 of
Oral
Maxillofacial-Head
Crown Copyright Ó 2014 Published by Elsevier Inc on behalf of the American
Neck
Association of Oral and Maxillofacial Surgeons. All rights reserved
zAssociate Professor; Department of Oral Maxillofacial-Head
0278-2391/13/00869-0$36.00/0
Neck Oncology.
http://dx.doi.org/10.1016/j.joms.2013.06.224
Drs Wu and Chen contributed equally to this work.
449
450
READERS’ PERSPECTIVES OF CLINICAL PHOTOGRAPHS
[JOMS] to require all patient-identifying information be removed or masked. Signed patient releases must accompany manuscripts in which there are photographs of identifiable patients unless eyes are masked to prevent identification.) With technologic advancements, simpler and more practical cameras with higher performance have been developed.4 For instance, the single lens reflex camera with an optional lens, flash lamp, aperture, and shutter provide better performance.5 However, the authors found some published medical journals with low-quality clinical photographs. Although it should be determined whether these lowquality clinical photographs can affect the readers’ understanding of a specific study, the authors recommend that high-quality clinical photographs be provided in journals. Diseases or trauma in the oral and maxillofacial region are apparent to readers compared with other clinical fields; hence, this visual nature optimizes the usefulness of photography, particularly in small anatomic regions, including the teeth, jaws, and different types of mucosa. To produce standardized photographs, a photographer requires proper training. However, oral and maxillofacial surgeons are busy and may be unable to allot extra time to study photography. A limited surgical field, complex anatomy, and emergency situations result in unsatisfactory visual reproductions of oral and maxillofacial surgery; therefore, visual reproduction should be improved.6,7 The frequency of low-quality photographs published in medical journals is seldom addressed. As a mainstream journal in the field of oral and maxillofacial surgery, the JOMS publishes articles with numerous photographs. In this study, 1,317 clinical photographs in the JOMS Volume 69 were evaluated and scored to determine the quality of such photographs and the frequency of low-quality photographs. This study also provided possible solutions to lessen common defects. The results may help improve the average quality of photographs and allow clinicians to take photographs of higher quality.
ASSESSMENT CONDITION
Materials and Methods INCLUSIVE AND EXCLUSIVE CRITERIA OF IMAGES
Consecutive photographs in Issues 1 to 12 of JOMS Volume 69, except Supplementary Issue 9 (which featured summaries and outlines that did not contain any photographs), were evaluated and scored one after another. Photographs in basic research or those irrelevant to oral and maxillofacial surgery were excluded, such as animal models, histopathologic images, immunofluorescence images, fiberscope views, life portraits, and computer-generated photographs. A total of 1,317 photographs were evaluated and scored.
Photographs printed in journals rather than electronic images downloaded from a database were used to avoid significant chromatic aberration from different computer screens. These images were assessed in the authors’ library under natural light. Photographs were assessed by 3 senior oral and maxillofacial surgeons with more than 5 years of clinical practice and familiarity with basic photographic terminologies. As these surgeons assessed the photographs, they were allowed to relax for 5 minutes at an interval of 30 minutes to avoid eyestrain. SCORE INDICATORS
Considering previous studies,7,8 10 indicators were selected, including sharpness, depth of field, light exposure, composition, color, background, position, distortion, label consistency, and white balance to evaluate the photographs. Each indicator was judged according to the following factors. 1) Sharpness refers to the outline of the focal point, which should be clear and sharp without blur. 2) The depth of the field is the distance from the front region to the back region of the focus. 3) Composition is the relative location between the point of interest and other objects. 4) Color was scored based on a colored or grayscale photograph. The background was given a higher score if it was gray or another color that did not influence the color of the main photograph. 5) Position corresponds to the posture of the limbs, body, or photographer. 6) Distortion refers to the undesired change in the shape or angle of the subject in a photograph. The same photograph labels or legends should be used consistently as the photographs are described in an article. 7) White balance is the process of adjusting the color of the light source to produce more realistic photographs. SCORING CRITERIA
Each indicator was assigned a score of 0 or 1 according to the aforementioned criteria. Each indicator was scored 1 if the photographs exhibited clear sharpness with good focus; adequate depth of field; appropriate light exposure; the point of interest in the composition across the center of the photograph; colored photographs; background highlighting of the object without affecting the object’s color or deviating the readers’ attention; the object or posture oriented at a horizontal or vertical direction except asymmetrically for cases with some specific diseases no inappropriate stretching or incorrect photographic angle between the camera and the object; appropriate photograph labeling; and appropriate white balance. Otherwise, the corresponding indicator was given the score of 0. Representative examples are shown in
451
WU, CHEN, AND XIONG
Figure 1.9-14 The sum of scores varying from 0 to 10 was defined as the final score of a specific photograph. STATISTICS
Data were analyzed in GraphPad Prism 5.01 (GraphPad, San Diego, CA). Nonparametric Kruskal-Wallis test was performed to compare the difference between groups. A P value less than .05 was considered statistically significant. Hierarchical clustering was performed using Cluster 3.0 (http://bonsai.hgc.jp/mdehoon/ software/cluster/) and visualized using TreeView 1.1.6r2 (http://jtreeview.sourceforge.net).
Results SCORE DISTRIBUTION AND INDICATOR VALUE
The total score of each photograph varied from 3 to 10 (mean score, 6.82 1.35) and exhibited a leftskewed distribution (Fig 2A). The average score of each indicator in the photographs varied from 0.045 to 1.0 (average score, 0.68; Fig 2B). The average scores were compared and picture composition, color, background, object position, distortion, and labeling
yielded an average score higher than 0.68. In contrast, sharpness, depth of field, exposure, and white balance (SDEW) yielded scores lower than 0.68 (Fig 2B). Hierarchical clustering showed that SDEW had similar low scores (Fig 2G). These results suggested that such indicators were unsatisfactory in oral and maxillofacial surgery photography. Among these indicators, depth of field obtained the lowest mean score, indicating that a typical oral and maxillofacial surgeon has difficulty in adjusting this parameter. In contrast, color showed the highest score (Fig 2B). POTENTIAL FACTORS RESPONSIBLE FOR IMAGE QUALITY
Different complexities in the content and subject may influence photographers and the quality of photographs. In this study, photographs were divided into 6 groups according to differences in content, including portrait, model, instrument, specimen, intraoral image, and extraoral image. The model group showed the highest score (7.97 1.1), and the intraoral group showed the lowest (6.50 1.18). Scores of the other groups were similar (Fig 2C). Kruskal-Wallis test
FIGURE 1. Representative examples. A, Inappropriate focus leads to poor photographic sharpness.9 B, Inadequate field depth leads to a fuzzy view of the distal bucca and mesial alveolar ridge mucosae.10 C, The image is underexposed.11 D, The point of interest is not in the center of the photograph.12 E, Photograph is printed in grayscale.13 F, The positions of the object and the ruler are skewed.14 Wu, Chen, and Xiong. Readers’ Perspectives of Clinical Photographs. J Oral Maxillofac Surg 2014.
452
READERS’ PERSPECTIVES OF CLINICAL PHOTOGRAPHS
FIGURE 2. Statistical analysis of indicators and photograph scores. A sum of the indicator scores in 1 photograph corresponded to a photograph score. A, Score distribution of photographs. B, Mean standard deviation for each indicator score of each photograph. C, Group comparison of contents (model, instrument, portrait, specimen, extraoral, and intraoral groups), emergency (nonemergency and emergency groups), and hospital duration (pre-, intra-, and postoperative groups). D, Indicator score for intraoral group. E, Indicator score for emergency group. F, Indicator score for intraoperative group. G, Hierarchical clustering of 10 indicators in 34 countries: a single indicator from the same country was arithmetically averaged; rows represent indicators and columns represent countries; centroid linkage clustering was performed in rows and columns respectively; light red, scored 1; light green, scored 0. B, D, E, F, Dotted lines indicate the mean value. ***P < .001. (Fig 2 continued on next page.) Wu, Chen, and Xiong. Readers’ Perspectives of Clinical Photographs. J Oral Maxillofac Surg 2014.
453
WU, CHEN, AND XIONG
FIGURE 2 (cont’d). Wu, Chen, and Xiong. Readers’ Perspectives of Clinical Photographs. J Oral Maxillofac Surg 2014.
showed significant differences among the 6 groups (P < .0001; Fig 2C). Significant differences also were observed between the model and intraoral groups, between the model and portrait groups, and between the portrait and instrument groups (P < .001; data not shown). The highest score for the model group was acceptable because models are easy to capture. The lowest score for the intraoral group indicated that good intraoral photographs are difficult to take. Different situations in hospitals, such as traffic trauma, acute allergic reaction, and hemorrhage, may influence photographers and the quality of photographs. In this study, photographs were categorized as emergency and nonemergency groups. The significantly different mean scores of the emergency and nonemergency groups were 6.03 1.89 and 6.84 1.33, respectively (P < .001; Fig 2C). Given that a patient’s condition can change during different hospital stays depending on disease progression and treatment interference, the quality of the photographs may be affected. Therefore, the photographs were categorized into preoperative, intraoperative, and postoperative groups. Differences in scores were analyzed and the average scores for the 3 groups were 7.11 1.36, 6.53 1.39, and 6.75 1.23, respectively. The intraoperative group yielded significantly lower scores than the pre- and postoperative groups (P < .001; Fig 2C). Moreover, the popularity and use of high-grade cameras may vary among different countries and influence the distribution of indicator scores. To further identify the relation of potential scores among different countries and the potential distribution characteristics of indicator scores, hierarchical clustering analysis of indicator scores and indicators in each country was performed. As shown in Figure 2G, the indicator score groups were separated into 2 clusters, namely SDEW and other. These 2 clusters showed a significant differ-
ence (P < .05) based on independent-samples MannWhitney U test. However, no obvious aggregations in country clustering (Fig 2G) were found. Whether high-grade camera use influences the indicator score distribution awaits further investigation. In summary, the intraoral, emergency, and intraoperative groups showed the lowest scores in the corresponding analyses. As mentioned previously, SDEW showed scores lower than the average score for the other indicators; therefore, the authors investigated whether these indicators were the main problems in these 3 groups that prevented surgeons from taking good photographs. A similar method of analysis was used in which the unsatisfactory indicators yielded scores lower than the average (Fig 2, dotted line). Figures 2D, 2E, and 2F show that, in addition to SDEW, the main contributor to the poor performance of the 3 groups, the lower score for background may have affected the quality of photographs in the emergency and intraoperative groups. EYE COVERING AND PLOTTING SCALE IN PHOTOGRAPHS
During the scoring process, some portrait photographs did not include an eye cover and some specimen photographs did not include a scale bar. Therefore, another statistical analysis for eye cover and scale bar was performed. Approximately 45.5% of portraits (148 of 325) displayed uncovered eyes and 57.1% of specimens (40 of 70) did not appear with a scale bar.
Discussion This study aimed to determine the indicators that should be considered in oral and maxillofacial surgery photography. The authors found that the problems in quality mainly involved SDEW. Most photographs showed a score of 5 to 9 (Fig 2A), and only 1.7% of
454 photographs yielded perfect scores for the indicators considered in this study. Furthermore, the quality of photographs was affected by other factors, including the subject of photography, emergency or nonemergency situation, and different perioperative periods. The quality of an image depends on several factors, such as chromatic aberration during printing and the use of different cameras. Therefore, numerous indicators might be considered. In this study, 10 indicators with essential elements and some intermediate indicators that contribute to produce a high-quality photograph that an ordinary reader can understand were selected. This review is the first to describe the quality of photographs from the readers’ perspective; therefore, efficient changes in assessment should be implemented to improve future studies. A low indicator score indicates poor image quality, which may lessen the reproduction and presentation quality of a lesion or a scene. Inadequate sharpness, shallow depth of field, poor composition, and overexposure will lead to blurred lesion details, indistinct lesion edges and junctions between lesions and normal tissues, a poor visual field, and blown highlights. Lesions caused by lichen planus, vascular tumor, and Fordyce disease often have distinct colors. Underexposure, black-and-white images, inappropriate background selection, and adjusted white balance will result in false colors and hinder a reader’s ability to appreciate the actual lesions. In addition, the lack of a scale bar or an object to serve as a size reference in specimen photographs will make visualizing the size of a lesion difficult for readers. Studies have striven to improve the specialized techniques in medical photography by using more efficient equipment4,5 and professional cameras.6 Although these methods significantly improve the levels of photographer and image quality, common defects may be neglected during this process, particularly when surgeons are in a stressful situation. Hence, the authors determined the common errors observed in images during oral surgery to help avoid such errors. The difference between groups is shown in Figure 2, which illustrates the common defects in oral and maxillofacial surgery photography. Photographs of portraits, models, and instruments are usually taken in a normal situation, allowing clinicians to arrange the object locations and environmental conditions, such as light and background. A patient’s consent and cooperation are obtained. As expected, these indicators yielded higher scores than those for intraoral photographs, in which light and background are difficult to adjust.15 For the emergency group, a background showing bloodstains and a critical situation may result in lower scores. Similarly, the score for the intraoperative group was lower than those for the pre- and postoperative groups.
READERS’ PERSPECTIVES OF CLINICAL PHOTOGRAPHS
Several common photograph defects were found in this study. However, such defects may not evidently influence the main content in some published images. Other unnoticed defects might exert more influence on quality. This deviation was statistically minimized by evaluating a relatively large sample of 1,317 photographs. Hence, the findings could reflect a relatively realistic state of photographic quality. Patient privacy should be protected during image processing. Midlevel and senior oral and maxillofacial surgeons are accustomed to seeing maxillofacial trauma, tumors, and facial transfiguration before and after surgery. However, patients and their friends may be unfamiliar or uncomfortable with such images, which is the reason most patients prefer to conceal their lesions or the images of such lesions to avoid scrutiny or untoward comments. The Health Insurance Portability and Accountability Act Privacy Rule gives individuals the right to protect their health information and the right to file a complaint to the Department of Health and Human Services Office for Civil Rights if their privacy is violated. Although oral and maxillofacial surgeons may be unwilling to modify surgery images, they can avoid lawsuits by obtaining patients’ informed consent before publishing patientsensitive data or by adding an image element that conceals the eyes of patients in photographs to prevent identification. In summary, although most photographs published in the JOMS exhibit high quality, some common unsatisfactory errors remain. Paying additional attention to common defects in SDEW might significantly improve the average quality of photographs and help to better reproduce and record some valuable scenes and moments in oral and maxillofacial surgery. Acknowledgments The authors thank Professor Wenfeng Zhang, Sangang He, and Jun Jia for their careful evaluation of the photographs and the library staff for their assistance in data collection.
References 1. Becker DG, Tardy ME Jr: Standardized photography in facial plastic surgery: Pearls and pitfalls. Facial Plast Surg 15:93, 1991 2. Sommer DD, Mendelsohn M: Pitfalls of nonstandardized photography in facial plastic surgery patients. Plast Reconstr Surg 114: 10, 2004 3. Riml S, Piontke A, Larcher L, et al: Quantification of faults resulting from disregard of standardised facial photography. J Plast Reconstr Aesthet Surg 64:898, 2011 4. Liu F: [Techniques of oral cavity photography II. The choice of camera and appliance for dentistry]. Zhonghua Kou Qiang Yi Xue Za Zhi 42:190, 2007 (in Chinese) 5. Peck JJ, Roofe SB, Kawasaki DK: Camera and lens selection for the facial plastic surgeon. Facial Plast Surg Clin North Am 18: 223, 2010 6. Sackstein M: [Intra-oral digital photography with the non professional camera—Simplicity and effectiveness at a low
WU, CHEN, AND XIONG
7.
8.
9.
10.
price]. Refuat Hapeh Vehashinayim 24:19, 2006 (in Hebrew) Liu F: [Techniques of oral cavity photography I. The basics of oral cavity photography]. Zhonghua Kou Qiang Yi Xue Za Zhi 42:60, 2007 (in Chinese) Han L, Zhou ZB: [Techniques of oral cavity photography V. The photography of oral maxillofacial surgery, part one]. Zhonghua Kou Qiang Yi Xue Za Zhi 42:575, 2007 (in Chinese) Furr MC, Larkin E, Blakeley R, et al: Extending multidisciplinary management of cleft palate to the developing world. J Oral Maxillofac Surg 69:237, 2011 Collette DC, Zechel MA: Novel treatment of atypical human papillomavirus-associated epithelial hyperplasia with cidofovir. J Oral Maxillofac Surg 69:2383, 2011
455 11. Zhang S, Liu X, Xu Y, et al: Application of rapid prototyping for temporomandibular joint reconstruction. J Oral Maxillofac Surg 69:432, 2011 12. Ellis E III: A study of 2 bone plating methods for fractures of the mandibular symphysis/body. J Oral Maxillofac Surg 69:1978, 2011 13. Wada T, Nakatani K, Hiraishi Y, et al: Usefulness of myofascial flap without skin in contemporary oral and maxillofacial reconstruction. J Oral Maxillofac Surg 69:1815, 2011 14. Movahed R, Weiss A, Velez I, et al: Submandibular gland MALT lymphoma associated with Sjogren’s syndrome: Case report. J Oral Maxillofac Surg 69:2924, 2011 15. Ahmad I: Digital dental photography. Part 8: Intra-oral set-ups. Br Dent J 207:151, 2009
