Use of diffusion tensor imaging in assessing superficial myometrial invasion by endometrial carcinoma: a preliminary study.

XML Template (2014) [8.10.2014–9:58am] //blrnas3.glyph.com/cenpro/ApplicationFiles/Journals/SAGE/3B2/ACRJ/Vol00000/140163/APPFile/SG-ACRJ140163.3d

(ACR)

[1–8] [PREPRINTER stage]

Acta Radiol OnlineFirst, published on October 20, 2014 as doi:10.1177/0284185114552293

Original Article

Use of diffusion tensor imaging in assessing superficial myometrial invasion by endometrial carcinoma: a preliminary study

Acta Radiologica 0(0) 1–8 ! The Foundation Acta Radiologica 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0284185114552293 acr.sagepub.com

Longmin Zhang, Ailian Liu, Ting Zhang, Qingwei Song, Qiang Wei and Heqing Wang

Abstract Background: Magnetic resonance imaging (MRI) remains the standard modality for local staging of gynecological malignancies, but it has several limitations, especially when differentiating a cancer limited to the endometrium from a cancer invading the superficial myometrium. Purpose: To explore 1.5 T diffusion tensor imaging (DTI) in assessing superficial myometrial infiltration by endometrial carcinoma. Material and Methods: We analyzed the sensitivity of apparent diffusion coefficient (ADC) versus fractional anisotropy (FA) in diagnosing superficial myometrial infiltration compared to DCE-MRI and T2-weighted imaging (T2WI) in 35 patients with endometrial cancer. For each patient, T2WI-DWI fusion images were generated, and five regions of interest (ROIs) were placed on corresponding DTI images. ADC and FA were calculated, and fiber tractography (FT) images for each level were obtained. ADC and FA values for the five ROIs were compared. Results: In distinguishing cancerous versus non-cancerous areas within superficial myometrium, median ADC values were significantly lower (1.16 vs. 1.48, respectively; P < 0.001) and median FA values were significantly higher (0.41 vs. 0.27; P < 0.001, respectively). ADC’s versus FA’s sensitivity, specificity, PPV, NPV, and accuracy for diagnosing superficial myometrial invasion were 74.3%, 88.6%, 86.7%, 77.5%, 81.4% versus 88.6%, 97.1%, 96.9%, 89.5%, 92.9%, respectively. T2WI and DCE-MR showed a sensitivity of 80.0% and 77.1%, respectively, in diagnosing myometrial invasion. Conclusion: Both ADC and FA were able to distinguish between cancerous verss non-cancerous areas within superficial myometrium (although FA was more sensitive based on AUC values). In addition, FA was superior to ADC, and more sensitive than T2WI and DCE-MR, in evaluating myometrial invasion. FT images provided visual confirmation of irregular arrangement and direction of the fibers due to proliferation of stromal cells caused by superficial myometrial invasion.

Keywords Diffusion tensor imaging (DTI), endometrial cancer, apparent diffusion coefficient (ADC), fractional anisotropy (FA), fiber tractography (FT) Date received: 16 June 2014; accepted: 30 August 2014

Introduction Endometrial cancer is the fourth most prevalent malignant tumor in women and has the highest incidence among cancers of the reproductive system (1). Its incidence is especially high in developed countries. The depth of myometrial invasion determines the clinical stage of endometrial carcinoma and is the most important morphologic prognostic factor. Endometrial cancer is staged using the International Federation of Gynecology and Obstetrics (FIGO) system, which recently underwent a major revision (1,2).

Magnetic resonance imaging (MRI) remains the standard modality for local staging of gynecological malignancies, but it has several limitations (3), especially Department of Radiology, The First Affiliated Hospital of Dalian Medical University, Dalian, PR China Corresponding author: Ailian Liu, Department of Radiology, The First Affiliated Hospital of Dalian Medical University, No. 222 Zhongshan Road, Xigang District, Dalian, Liaoning, 116011 PR China. Email: [email protected]

Downloaded from acr.sagepub.com at Baskent Universitesi on December 23, 2014


(ACR)


2

Acta Radiologica 0(0)

when differentiating a cancer limited to the endometrium from a cancer invading the superficial myometrium (4). Recently, Fujimoto et al. investigated the differences in the fiber architecture and other related diffusion tensor imaging (DTI) parameters among different uterine layers of the normal human uterus in vivo at 3 T (5). Using DTI and fractional anisotropy (FA), they found layer-wise differences in the microstructure of the uterus. We hoped to extend their work by applying fiber tractography (FT) and FA methodology to the evaluation of superficial myometrial invasion by endometrial cancer. Our hypothesis was that the ability of DTI to determine the fiber architecture of the uterus could be used to detect muscular fiber destruction caused by superficial myometrial invasion.

MR system and scanning methods A 1.5 T MR scanner (Signa EXCITE HD, GE Healthcare, Milwaukee, WI, USA) was used to scan all 35 patients. Body coil and pelvic 8-channel phasedarray coil were selected as the transmitting and receiving coils, respectively. The following MRI sequences were used: T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), dynamic contrast-enhanced MRI (DCEMRI), DWI, and DTI. Prior to scanning, subjects were required to fast for 4–6 h to reduce artifacts caused by intestinal peristalsis. DTI data were acquired using SE-EPI sequence with the following parameters: bvalue, 0 and 600 s/mm2; six diffusion-sensitizing gradients; TR/TE, 4500/85.1; slice thickness, 6 mm; increments, 1.5 mm; field of view, 32 32; matrix size, 128 128; NEX, 6; and DTI scanning time, 3 min 35 s.

Image postprocessing

Material and Methods Patients This study was approved by our ethical committee, and written informed consent was obtained from all patients prior to their participation in the study at the pre-design level. We retrospectively analyzed all patients who visited our hospital for irregular vaginal bleeding between April 2011 and February 2013. Forty patients were diagnosed with endometrial carcinoma with myometrial invasion (29 menopausal women and 11 premenopausal women) as shown on pelvic MRI and confirmed at surgical pathology. On the basis of postoperative pathology, patients were recruited, and data were collected for measurements and analysis. For patients with suspicious endometrial cancer, the scanning sequence was fixed, and routine and dynamic contrast-enhanced MRI, diffusion weighted imaging (DWI), DTI were performed. We analyzed this imaging information and used the pathologic results as reference, in addition to the T2-DWI sequence, which was also used as reference. Of the 40 patients identified with myometrial invasion, two patients with thin endometrium and strong intestinal imaging artifacts were excluded. Three additional patients with imaging artifacts (due to a large amount of hemorrhage within the uterine tissue) were also excluded. The remaining 35 patients were included in this study. MRI examinations were performed prior to biopsy or other treatment and within 2 weeks of surgery. The primary complaint among the menopausal patients was irregular vaginal bleeding while the premenopausal patients complained about increased or prolonged menstruation or intermenstrual vaginal bleeding.

A GE AW4.4 workstation (GE Healthcare) was used for imaging analysis. The images were interpreted by two radiologists with 13 and 2 years of experience, respectively, in diagnostic abdominal MRI. The two radiologists independently assessed the sensitivity of T2WI, DCEMRI, DWI, and DTI in diagnosing superficial myometrial infiltration by endometrial carcinoma. A positive finding on the T2WI image was defined as an interruption of the uterine junctional zone (JZ) by cancer invasion. On DCE-MRI, superficial myometrial infiltration was determined by the difference in enhancement between endometrium, uterine intracavitary lesion, and myometrium. Using the T2-DWI image (Fig. 1a) to identify the level with the largest cross-sectional area of the tumor (confirmed on pathology to have superficial myometrial infiltration), ROIs (1, 2, 3, 4, and 5) were placed on the same DTI level (Fig. 1b). FA images and values were automatically generated (Fig. 1c, d). From Fig. 1, the five ROIs were: intracavitary tumor (ROI1), superficial myometrial invasion (ROI2), normal superficial myometrium (ROI3), deep myometrial invasion (ROI4), and normal deep myometrium (ROI5). ROI3 was placed on the uterine junctional zone (JZ), which appeared as a circular low-signal zone. If the location of the superficial myometrial invasion zone appeared unclear, the position opposite the normal JZ was selected. For determining the location of the deep myometrium, the ROI was placed at less than one-half the depth of the myometrium as measured from the uterine serosa to the myometrium. Since older postmenopausal women have thinner endometria, a smaller ROI was selected (10–15 mm2) for their assessment, while a 20–25 mm2 ROI was selected for premenopausal women. There were four patients who had very thin myometria, and therefore their ROIs were placed only on the intracavitary tumor and JZ. For each



(ACR)


Zhang et al.

3

Fig. 1. DTI of superficial myometrial invasion by tumor confirmed at pathology. (a) T2WI-DWI fusion image used to identify the level with the largest cross-sectional area of the tumor. (b) ROIs (1, 2, 3, 4, and 5) were placed on the same DTI level. (c, d) FA images and FA values were automatically generated. (e) Gross specimen of superficial myometrial invasion by tumor. (f) H&E stained slide under 4 magnification showing superficial myometrial invasion to the right of the intracavitary tumor in the fig. Note: the right side of the fig. corresponds to the left side of the uterus. Red arrow: intracavitary cancer; yellow arrow: superficial myometrium invasion by tumor; blue arrow: normal myometrium. Figs 1 a, b, c and d are from the same cross-sectional level. Ut, uterus.

ROI, apparent diffusion coefficient (ADC) and fractional anisotropy (FA) were calculated and expressed as medians with interquartile ranges (IQRs). In this study, ADC values were present using unit as mm2/s. FT images for each level were obtained to demonstrate the arrangement of fibers.

Statistical analysis Data were presented as medians with IQRs (the range between the 25th and 75th percentiles) for continuous

variables, and frequencies with percentages for categorical variables. Comparisons between DTI parameters (ADC and FA) for the five different ROIs were performed by the Friedman test. Given significant findings from the Friedman test, the post-hoc tests were performed by Wilcoxon signed-rank test with Bonferroni correction for the following comparisons: intracavitary tumor versus superficial myometrial invasion (ROI1 vs. ROI2), superficial myometrial invasion versus normal superficial myometrium (ROI2 vs. ROI3), as well as



(ACR)


4


deep myometrial invasion versus normal deep myometrium (ROI4 vs. ROI5). The receiver operating characteristic (ROC) curves were used to examine the diagnostic performance of DTI parameters used to distinguish cancerous areas versus non-cancerous areas within the superficial myometrium (ROI2 vs. ROI3), and the point estimate with 95% confidence interval (CI) of the area under ROC curve (AUC) was provided as an index to compare global test performance. The corresponding sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were provided at the optimal cut-off values of DTI parameters to discriminate cancerous from non-cancerous areas within the superficial myometrium (ROI2 vs. ROI3), which were determined by maximizing the sum of sensitivity and specificity. All statistical analyses were performed using SAS software version 9.2 (SAS Institute Inc., Cary, NC, USA), and a two-tailed P < 0.05 indicated statistical significance.

Results Demographics and pathological findings in study subjects

Table 1. Demographics and pathologic findings in all 35 patients. n ¼ 35 Age (years) Median (IQR) Range Pathology, n (%) Adenocarcinoma Adenosquamous carcinoma Cellular differentiation, n (%) High Medium to high Medium Low to medium Low Unspecified Stage, n (%) Ia II III Lymph node metastasis, n (%) No Yes

56 (52, 61) 45–76 33 (94.3) 2 (5.7) 7 11 6 6 1 4

(20.0) (31.4) (17.1) (17.1) (2.9) (11.4)

29 (82.9) 5 (14.3) 1(2.9) 34 (97.1) 1 (2.9)

All 35 patients were found to have superficial myometrial invasion by cancer confirmed on pathology reports. The demographics and pathologic findings in all 35 patients are presented in Table 1. The patients’ ages were in the range of 45–76 years, with a median of 56 years (IQR, 52–61 years). Regarding pathologic findings, 33 patients (94.3%) had tumors identified as adenocarcinoma and only two patients (5.7%) had cancers identified as adenosquamous carcinoma. Regarding the level of cellular differentiation, the most frequent grade found was medium to high grade (11 patients, 31.4%), followed by high grade (7 patients, 20.0%). A total of 29 patients (82.9%) were classified as Stage Ia, five (14.3%) were Stage II, and one patient (2.9%) was found to have lymph node metastasis as Stage III.

Sensitivity of T2WI and DCE-MR in detecting superficial myometrial invasion from endometrial cancer According to T2WI, 28 of 35 patients showed positive findings of superficial myometrial invasion, resulting in a sensitivity of 80.0%. Moreover, on the enhanced T1WI scan, 27 of 35 patients showed positive findings of superficial myometrial invasion, resulting in a sensitivity of 77.1%.

DTI parameters (ADC and FA) across five different regions of interest (ROIs) ADC values across the five different ROIs are shown in Fig. 2. The overall tests revealed that the

Fig. 2. ADCs determined by DTI across five different ROIs. The unit of ADCs: mm2/s. ROI1, intracavitary tumor; ROI2, superficial myometrial invasion; ROI3, normal superficial myometrium; ROI4, deep myometrial invasion; ROI5, normal deep myometrium.

distributions of ADCs differed significantly among the five ROIs (P < 0.001). The post-hoc tests showed ADC values were significantly lower in cancerous compared with non-cancerous areas within the



(ACR)


Zhang et al.

5

Fig. 3. The FA values determined by DTI across five different ROIs.

Fig. 4. The ROC curves of ADC and FA used to distinguish superficial myometrial invasion (ROI2) from normal superficial myometrium (ROI3). The area under the curve (AUC) was 0.874 for ADC (95% CI, 0.793–0.956) and 0.977 for FA (95% CI, 0.951– 1.000) (P ¼ 0.018).

superficial myometrium (median ROI2, 1.16 [IQR, 1.09, 1.33] vs. median ROI3, 1.48 [IQR, 1.36, 1.63]; P < 0.001). However, the comparisons in ADC values between intracavitary tumor versus superficial myometrial invasion (ROI1 vs. ROI2) and deep myometrial invasion versus normal deep myometrium (ROI4 vs. ROI5) were not statistically significant (both P > 0.05; Fig. 2). The comparisons in FA values across the five different ROIs are shown in Fig. 3. Similar to ADC results, the overall tests indicated that the distributions of FA differed significantly among the five ROIs (P < 0.001). The post-hoc tests showed that FA was significantly higher in cancerous compared with noncancerous areas within the superficial myometrium (median ROI2, 0.41 [IQR, 0.37, 0.45] vs. median ROI3, 0.27 [IQR, 0.24, 0.32]; P < 0.001) and significantly higher in superficial myometrial invasion versus intracavitary tumor (median ROI2, 0.41 [IQR, 0.37, 0.45] vs. median ROI1, 0.18 [IQR, 0.14, 0.22]; P < 0.001). No significant difference in FA was found between deep myometrial invasion versus normal deep myometrium (ROI4 vs. ROI5) (P ¼ 1.000; Fig. 3).

higher than the AUC of ADC (AUC for FA, 0.977 [95% CI, 0.951, 1.000] vs. AUC for ADC, 0.874 [95% CI, 0.793, 0.956], P ¼ 0.018), which indicated that FA was a better diagnostic tool than ADC in distinguishing superficial myometrial invasion from normal myometrium (ROI2 vs. ROI3). The corresponding sensitivity, specificity, PPV, NPV, and accuracy, at the optimal cut-off values for ADC versus FA, were 74.3%, 88.6%, 86.7%, 77.5%, and 81.4% versus 88.6%, 97.1%, 96.9%, 89.5%, and 92.9%, respectively (Table 2).

ROC curve analysis using ADC and FA to distinguish cancerous from non-cancerous areas of superficial myometrium

Discussion

Since both ADC and FA values showed significant differences between cancerous versus non-cancerous areas within the superficial myometrium (ROI2 vs. ROI3), further ROC curve analysis was performed to assess the discriminating ability between these two DTI parameters (Fig. 4). The AUC of FA was significantly

FT images FT images were reconstructed as three-dimensional images to show the arrangement, direction, and density of uterine fibers. In this study, FT images of the ROIs were generated in all patients. Cancer zones on FT images showed cancer infiltrating the myometrial fibers with uneven color and irregular arrangement and direction of fibers, while non-cancer zones on FT images showed even color and regular direction and arrangement of fibers (Fig. 5).

To the best of our knowledge, this study is the first to demonstrate the value of DTI in the in vivo evaluation of superficial myometrial infiltration by tumor. We found that both ADC and FA showed significant differences between cancerous versus non-cancerous areas within the superficial myometrium. The ADC values were significantly lower (Fig. 2), and FA values were significantly higher (Fig. 3) in the superficial



(ACR)


6


Table 2. Diagnostic performance of ADC and FA in distinguishing cancerous area (ROI2) from non-cancerous area (ROI3) within the superficial myometrium. DTI parameter

Optimal cut-off value

Sensitivity

Specificity

PPV

NPV

Accuracy

ADC (e9) FA

1.320 0.356

74.3% 88.6%

88.6% 97.1%

86.7% 96.9%

77.5% 89.5%

81.4% 92.9%

NPV, negative predictive value; PPV, positive predictive value.

Fig. 5. FT image of superficial myometrial invasion by tumor in same patient as in Fig. 1. FT image showing irregular fiber direction and arrangement and uneven color on the right side of fig. (black arrows) due to the proliferation of stromal cells within the left superficial myometrium of the uterus; the left half of the fig. shows non-cancer zone (red arrows) with regular fiber direction and distribution and uniform color within the right myometrium of the uterus.

myometrium involved with tumor compared with normal superficial myometrium. In addition, FA was more sensitive than ADC in detecting superficial myometrial invasion based on AUC values (Fig. 4). FA was also more sensitive than either T2WI or DCE-MR in diagnosing myometrial invasion, in general. The FT images provided visual confirmation of irregular arrangement and direction of the fibers due to the proliferation of stromal cells caused by superficial myometrial invasion (Fig. 5). MRI is currently the most important imaging examination for preoperative evaluation of endometrial carcinoma. Findings of endometrial carcinoma on MRI include an enlarged uterus, diffuse and irregular thickening of the endometrium, hypointense T1WI signals, relatively hyperintense T2WI signals, and interruption and blurring of the JZ. Under DCE-MR, the endometrium is less enhanced than the JZ. A continuous interface between endometrium and the JZ normally exists, and the degree to which such an

interface is intact on T2WI can be used to determine invasion of the superficial myometrium. In a metaanalysis of 11 manuscripts published between 1995 and 2012, Wu et al. (6) found that the sensitivity, specificity, PPV, and NPV of T2WI in evaluating myometrial invasion by tumor were 0.87, 0.58, 0.64, and 0.84, respectively, compared to 0.81, 0.72, 0.65, and 0.85, respectively, using DCE-MRI. In some postmenopausal patients or in patients with a significantly enlarged uterine cavity, the JZ may appear unclear, leading to over- or underestimation of tumor staging based on T2WI evaluation of the JZ. In addition, morphological evaluation using MRI is not supported by quantitative data. Contrast-enhanced MRI can determine myometrial infiltration according to the difference in enhancement between the endometrium, the JZ, and the myometrium (6). However, this approach requires injection of contrast agent and is, therefore, invasive. DWI is a non-invasive method that examines the diffusion of water molecules in tissues. DWI images of tissues are low in resolution, however, and JZ signals are difficult to differentiate. Therefore, DWI images cannot serve as a stand-alone tool for staging. However, ADC is exquisitely sensitive to pathological changes and more accurate in staging compared to DCE-MR (7). ADC, therefore, is a quantitative tool which can be used to differentiate normal from cancerous tissues in the uterine cervix and endometrium. The ADC cut-off value is reportedly 1.4 103 mm2/s for normal tissue and 1.15 103 mm2/s for cancerous tissues in the uterine cervix and endometrium (8). A previous report (9) confirmed that endometrial carcinoma has significantly lower ADC values than normal endometrium (P < 0.01), and there is little overlap between the two signals. Another study utilized high b-value T2WI and DWI fusion images to evaluate the myometrial infiltration of endometrial carcinoma (10). Thus, reliance on ADC values has resolved the T2WI problem of the unclear border between endometrial carcinoma and the surrounding myometrium and has overcome the shortcoming of low signal-to-noise ratio found with high b-value DWI images. DTI is a new MRI technique that is based on DWI but is more sensitive to molecular diffusion of water.



(ACR)


Zhang et al.

7

DTI parameters include FA, and FT. In this study, ADC and FA values were compared across five ROIs. We found no significant difference in ADC values between intracavitary cancer versus superficial myometrial invasion (P ¼ 0.553, Fig. 2). Superficial myometrial invasion, however, had a higher FA value than intracavitary tumor (P < 0.001, Fig. 3). This finding suggests that FA was more sensitive than ADC in reflecting microscopic changes caused by tumor infiltration. Compared to normal superficial myometrium in the same person, superficial myometrial invasion had significantly lower ADC values and higher FA values (P < 0.001; Figs 2 and 3, respectively). There was no difference in either FA or ADC values between deep myometrial invasion and normal deep myometrium (P ¼ 1.000). In our study, cancer infiltrated the superficial myometrium, but the deep myometrium was not infiltrated by cancer cells. Thus, the absence of significant differences in either FA or ADC values might be attributed to lack of infiltration of the deep myometrium by cancer cells. Based on combined DTI and pathological findings, Toba et al. (11) believed that cancer invasion induces growth of stromal cells within the superficial myometrium. As a consequence, anisotropic zones are formed from tight, irregular, and anisotropic cellular arrangements. According to Weinberg et al. (12), distant proliferation of cancer cells often requires formation of stromal cells; therefore, stromal cells surrounding the cancer zone should become activated prior to myometrial invasion. All subjects in this study had superficial myometrial infiltration of endometrial carcinoma. The active stromal cells, which are adjacent to the cancer zone in superficial myometrium, formed the DTI anisotropic zones, thus lead to elevated FA value and better sensitivity compared to ADC value in the zone of superficial myometrial invasion. Further invasion of the myometrium can disrupt smooth muscle structure, leading to reduced anisotropy. In this study, 10–25 mm2 ROIs were placed in areas of smooth muscle structure disruption and surrounding stromal cell growth. The results showed increased FA values in the cancer-infiltrated areas, suggesting existence of anisotropy due to the reactive growth of stromal cells. Therefore, the existence of anisotropy, as demonstrated by increased FA value in superficial myometrium involved with tumor, can be used as a sensitive method of determining myometrial invasion by tumor. The growth of superficial myometrial stromal cells restricts water molecule activity reducing ADC and increasing FA values. Thus, in undiagnosed cases, once the FA and ADC values are measured, several ROIs may be placed at different slices (or at different sites within the same slice) to acquire DTI parameters. Comparison with normal tissues (or

with prior results) for the same patient will also aid in the diagnosis. In this study, we performed ROC analysis on ADC and FA. Our results showed that FA had a higher diagnostic accuracy, sensitivity, specificity, PPV, and NPV compared to ADC. This finding suggests that the DTI technique can quantitatively assess superficial myometrial invasion by tumor through measurements of FA. As FA reflects restricted movement of water molecule in tissues, it reveals the changes in anisotropy within the myometrial tissue affected by the cancer. FT images allow observation of changes in myometrial fiber arrangement, which provides additional confirmation of superficial myometrial invasion by tumor. This study had several limitations. In addition to its retrospective nature, our sample size was small and included only patients with confirmed myometrial invasion which may have biased the radiologists’ readings. The study also lacked a uniform level for the b-value. Future prospective studies involving larger cohorts are needed to confirm our findings and hopefully extend the DTI technique to the assessment of deep myometrial invasion, cervical invasion, and pelvic lymph node metastasis. In conclusion, the DTI technique provides a safe and accurate quantitative evaluation of superficial myometrial invasion by endometrial carcinoma that is superior to conventional MR sequences. In addition, FT images offer a new morphological method which can provide visual confirmation of irregular arrangement and direction of the fibers due to the proliferation of stromal cells caused by superficial myometrial invasion. Conflict of interest None declared.

Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References 1. Beddy P, O’Neill AC, Yamamoto AK, et al. FIGO staging system for endometrial cancer: added benefits of MR imaging. Radiographics 2012;32:241–254. 2. Creasman W. Revised FIGO staging for carcinoma of the endometrium. Int J Gynaecol Obstet 2009;105:109. 3. Levy A, Medjhoul A, Caramella C, et al. Interest of diffusion-weighted echo-planar MR imaging and apparent diffusion coefficient mapping in gynecological malignancies: a review. J Magn Reson Imaging 2011;33:1020–1027. 4. Shin KE, Park BK, Kim CK, et al. MR staging accuracy for endometrial cancer based on the new FIGO stage. Acta Radiol 2011;52:818–824. 5. Fujimoto K, Kido A, Okada T, et al. Diffusion tensor imaging (DTI) of the normal human uterus in vivo at 3



(ACR)


8


tesla: comparison of DTI parameters in the different uterine layers. J Magn Reson Imaging 2013;38: 1494–1500. 6. Wu LM, Xu JR, Gu HY, et al. Predictive value of T2-weighted imaging and contrast-enhanced MR imaging in assessing myometrial invasion in endometrial cancer: a pooled analysis of prospective studies. Eur Radiol 2013;23: 435–449. 7. Beddy P, Moyle P, Kataoka M, et al. Evaluation of depth of myometrial invasion and overall staging in endometrial cancer: comparison of diffusion-weighted and dynamic contrast-enhanced MR imaging. Radiology 2012;262: 530–537. 8. Motoshima S, Irie H, Nakazono T, et al. Diffusionweighted MR imaging in gynecologic cancers. J Gynecol Oncol 2011;22:275–287.

9. Seo JM, Kim CK, Choi D, et al. Endometrial cancer: utility of diffusion-weighted magnetic resonance imaging with background body signal suppression at 3T. J Magn Reson Imaging 2013;37:1151–1159. 10. Feng J, Long MM, Ni HY, et al. Determination of the depth of myometrial infiltration of stage I endometrial carcinoma using DWI-T2WI fusion images. Chinese Journal of Medical Imaging Techniques 2012;28:1187–1190. 11. Toba M, Miyasaka N, Sakurai U, et al. Diagnostic possibility of diffusion tension imaging for the evaluation of myometrial invasion in endometrial cancer: an ex vivo study. J Magn Reson Imaging 2011;34:616–622. 12. Weinberg RA. Dialogue replaces monologue: heterotypic interactions and the biology of angiogenesis. In: Weinberg RA (ed.) The biology of cancer, 1st edn. New York, NY: Garland Science, 2006, p.535.


Myometrial invasion by endometrial carcinoma: assessment by MR imaging.

Effectiveness of diffusion tensor imaging in assessing disease severity in Duchenne muscular dystrophy: preliminary study.

Reproducibility of measurement of myometrial invasion in endometrial carcinoma.

Myometrial invasion by endometrial carcinoma: assessment with plain and gadolinium-enhanced MR imaging.

Preliminary Study of MR Diffusion Tensor Imaging of the Liver for the Diagnosis of Hepatocellular Carcinoma.

transrectal ultrasound for assessing myometrial invasion in endometrial cancer: a comparison of six different approaches.

MR diffusion imaging for preoperative staging of myometrial invasion in patients with endometrial cancer: a systematic review and meta-analysis.

Magnetic resonance imaging for detection of depth of myometrial invasion and cervical invasion in patients with endometrial carcinoma.

A preliminary diffusional kurtosis imaging study of Parkinson disease: comparison with conventional diffusion tensor imaging.

Endometrial Cancer: Combined MR Volumetry and Diffusion-weighted Imaging for Assessment of Myometrial and Lymphovascular Invasion and Tumor Grade.

Diffusion-weighted imaging in assessing renal pathology of chronic kidney disease: A preliminary clinical study.

Assessment of depth of myometrial invasion in endometrial adenocarcinoma.

Correction: Preliminary Study of MR Diffusion Tensor Imaging of Pancreas for the Diagnosis of Acute Pancreatitis.

Preliminary Study of MR Diffusion Tensor Imaging of Pancreas for the Diagnosis of Acute Pancreatitis.

Liver recurrence in early endometrial cancer with focal myometrial invasion.

Cockayne syndrome: a diffusion tensor imaging and volumetric study.

Anisotropic fractional diffusion tensor imaging.

Diffusion tensor imaging study in Duchenne muscular dystrophy.

Evaluation of myometrial invasion in endometrial cancer: comparison of diffusion-weighted magnetic resonance and intraoperative frozen sections.

White Matter Plasticity Induced by Psychoeducation in Bipolar Patients: A Controlled Diffusion Tensor Imaging Study.

Hysteroscopic transcervical resection is useful to diagnose myometrial invasion in atypical polypoid adenomyoma coexisting with atypical endometrial hyperplasia or endometrial cancer with suspicious myometrial invasion.

Depth of myometrial invasion in endometrial cancer: preoperative assessment by transvaginal ultrasonography.

Diffusion kurtosis imaging evaluating epithelial-mesenchymal transition in colorectal carcinoma xenografts model: a preliminary study.

Peripheral nerve tractography in soft tissue tumors: a preliminary 3-tesla diffusion tensor magnetic resonance imaging study.