n Genitourinar y ima Gin G
analysis of Multiparametric Prostate Mr images for
Prostate cancer Detection and aggressiveness assessment:
A Cross-Imager Study 1
Yahui Peng, PhD2 Yulei Jiang, PhD Tatjana Antic, MD Maryellen L. Giger, PhD Scott E. Eggener, MD Aytekin Oto, MD
Purpose: To validate three previously identified quantitative image features across multiparametric magnetic resonance (MR) images acquired with imagers made by two different man- ufacturers to differentiate prostate cancer (PC) from nor- mal prostatic tissue and to assess cancer aggressiveness.
Materials and
Methods: This study was HIPAA-compliant and approved by the institutional review board. Preoperative 1.5-T multipara- metric endorectal MR images of 119 PC patients (data- set A, 71 patients; dataset B, 48 patients) were analyzed, and 265 PC and normal peripheral zone regions of in- terests (ROIs) were identified through histologic and MR consensus review. The 10th percentile average apparent diffusion coefficient (ADC) value, average ADC value, and skewness of T2-weighted signal-intensity histogram were evaluated with area under the receiver operating charac- teristic curve (AUC). The image features were combined with a linear discriminant analysis classifier and evaluated both on the image dataset of each type of imager alone (leave-one-patient-out evaluation) and across the datasets (training on one dataset, testing on the other). Spearman correlation coefficient was calculated between the image features and ROI-specific Gleason scores.
Results: AUC values of the image features combined were 0.95 6 0.02 (standard error) and 0.88 6 0.03 on dataset B and dataset A alone, respectively, and 0.96 6 0.02 and 0.89 6 0.03 when training on dataset A and testing on dataset B and vice versa, respectively. Spearman correlation coeffi- cients between Gleason scores and the ADC features were between 20.27 and 20.34.
Conclusion: Consistently across images from datasets A and B, the 10th percentile ADC value, average ADC value, and T2- weighted skewness can distinguish PC from normal-tissue ROIs, and ADC features correlate moderately with ROI- specific Gleason scores.
q RSNA, 2014
1 From the Departments of Radiology (Y.P., Y.J., M.L.G., A.O.), Pathology (T.A.), and Surgery Section of Urology (S.E.E.), University of Chicago, 5841 S Maryland Ave, Chi- cago, IL 60637. Received June 7, 2013; revision requested July 25; revision received October 21; accepted November 8; final version accepted December 9. Supported in part by the U.S. Army Medical Research and Materiel Command Prostate Cancer Research Program through an Idea Devel- opment Award (PC093485). Address correspondence to Y.P. (e-mail: [email protected]).
2Current address: School of Electronic and Information Engineering, Beijing Jiaotong University, No. 3 Shangyuan- cun, Haidian District, Beijing, China, 100044 (Y.P.).
q RSNA, 2014
Accountability Act, and approved by our institutional review board with a waiver of written informed patient consent. We searched the radiologic image archive at our institution and identified 139 con- secutive patients between July 2007 and May 2010 by using the following inclusion criteria: (a) patient had biopsy-proved prostate adenocarcinoma, (b) patient underwent multiparametric MR exami- nation with an endorectal coil (includ- ing T1-weighted, T2-weighted, DW MR imaging, and DCE MR sequences), and (c) patient underwent radical prostatec- tomy. Twenty patients were excluded be- cause of the following: patient received radiation therapy before the MR exami- nation (n = 1), the radical prostatectomy was performed before the MR examina- tion (n = 2), prostatectomy histologic slides were not available for review (n = 6), the MR examination was performed with 3-T imagers (n = 5), DW MR image data were missing (n = 1), or DCE MR image data were missing (n = 5).
A total of 119 patients were included in this study. Among them, 59.7% (71 of 119; dataset A) were imaged with 1.5-T imagers (Signa; GE Healthcare) between July 2007 and October 2008, and 40.3% (48 of 119; dataset B) were imaged with 1.5-T imagers (Achieva;
amount of data that need to be analyzed and the lack of a standardized interpre- tation approach that would lead to re- producible results. Because T2-weighted and DW MR images are the most im- portant sequences that are helpful in PC detection, efforts have been made for quantitative image analysis of these im- ages by using computer-aided diagnosis (CAD) (6–9). On the basis of data from MR imagers from a single manufacturer, we previously reported (9) that the 10th percentile apparent diffusion coefficient (ADC) value within a tumor region of in- terest (ROI), the average ADC value, and T2-weighted signal-intensity skewness within an ROI were effective image fea- tures for differentiation of PC foci from normal prostatic tissue, and that both of the ADC features correlated moder- ately with the Gleason score of a tumor.
However, one of the challenges for wide- spread use of MR imaging for PC is large variations in images acquired with im- agers made by different manufacturers.
The purpose of this study was to validate the three previously identified quantita- tive image features across multiparamet- ric MR images acquired with imagers made by two different manufacturers (Signa, GE Healthcare, Waukesha, Wis;
Achieva, Philips Healthcare, Eindhoven, the Netherlands) in order to differenti- ate PC from normal prostatic tissue and to assess cancer aggressiveness.
Materials and Methods
Patients
This study was retrospective, compliant with the Health Insurance Portability and
M
ultiparametric magnetic res- onance (MR) imaging is cur- rently the best imaging method to identify and characterize prostate cancer (PC), which is the second lead- ing cause of cancer-related death in men in the United States (1,2). The current practice of PC diagnosis based on transrectal ultrasonography-guided core needle biopsy has limitations that include the invasiveness of biopsy, like- lihood of missing anterior tumors, and inaccuracy in assessment of the aggres- siveness of PC or Gleason score (3).Multiparametric MR imaging provides anatomic information in T2-weighted images and functional information in images that are diffusion weighted (DW), dynamic contrast material en- hanced (DCE), and spectroscopic and has the potential to help improve the detection of PC and the differentiation of clinically significant PC from indolent tumors (2,4,5).
Interpretation of multiparametric MR prostate images is a challenge for radiologists because of an overwhelming
Implication for Patient Care n The quantitative ADC and
T2-weighted image features eval- uated in this study can be used consistently in multiparametric endorectal MR images acquired with 1.5-T imagers made by two different manufacturers for the differentiation of PC from nor- mal-tissue ROIs and for assess- ment of correlation strength with lesion-specific Gleason scores.
Advances in Knowledge
n The 10th percentile and the av- erage apparent diffusion coefficient (ADC) values and T2-weighted signal-intensity skewness are con- sistently effective (area under the receiver operating characteristic curve [AUC], 0.89–0.96) for dif- ferentiating prostate cancer (PC) foci from normal peripheral zone regions of interests (ROIs) in mul- tiparametric endorectal MR images acquired from 1.5-T im- agers made by two different manufacturers.
n The 10th percentile ADC value is consistently more effective than the average ADC value to differ- entiate PC foci from normal peripheral-zone tissue (AUCs, 0.92 vs 0.89 and 0.89 vs 0.87 on two independent image datasets).
n The 10th percentile and the av- erage ADC values consistently correlate moderately with lesion- specific Gleason scores (r between 20.27 and 20.34).
Published online before print
10.1148/radiol.14131320 Content code:
Radiology 2014; 271:461–471 Abbreviations:
ADC = apparent diffusion coefficient AUC = area under the ROC curve CAD = computer-aided diagnosis DCE = dynamic contrast enhanced DW = diffusion weighted PC = prostate cancer
ROC = receiver operating characteristic ROI = region of interest
Author contributions:
Guarantors of integrity of entire study, Y.P., Y.J., A.O.;
study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, Y.P., Y.J., M.L.G., A.O.; clin- ical studies, A.O.; experimental studies, Y.P., M.L.G., A.O.;
statistical analysis, Y.P., Y.J., M.L.G., A.O.; and manuscript editing, all authors
Conflicts of interest are listed at the end of this article.
considered high-grade prostatic in- traepithelial neoplasia, benign pros- tatic hyperplasia, and other benign abnormalities abnormal), unless nor- mal region could not be found on his- tologic slides (22 patients). A total of 265 ROIs were outlined, among which 102 (38.5%) were from normal periph- eral zone tissue and 163 (61.5%) were from PC tumors (Table 1). Although almost all ROIs were outlined in a T2- weighted image section, seven cancer ROIs in six cases were outlined in a DW MR image section and six other cancer ROIs in six cases were outlined in a DCE MR image section because those ROIs correlated better with his- tologic findings in those images than in the T2-weighted images. The patholo- gist also assigned a Gleason score spe- cifically to each cancer ROI (Table 1).
Transfer of ROIs between MR Images The manually drawn ROIs were trans- ferred to other MR image sequences to facilitate the analysis in those im- ages. First, assuming no patient motion throughout the entire multiparametric MR examination, the ROIs were auto- matically transferred from the sequence where the ROIs were drawn manually (A.O., 9 years of experience in pros-
tate MR imaging) established the ref- erence standard for PC and normal- tissue foci on MR images through a systematic histologic and MR correla- tive review. The pathologist identified all distinct tumor foci larger than ap- proximately 5 mm in diameter after a review of all prostatectomy tissue sec- tions. Then, by consensus, the radiolo- gist manually outlined ROIs of the cor- responding tumor foci on MR images that best correlated with the histologic findings. For tumor foci that were not discernible on MR images, their loca- tions on MR images were determined on the basis of the spatial relationship to other identifiable anatomic land- marks (eg, urethra, ejaculatory ducts, and benign prostatic hyperplasia). The radiologist drew ROIs on MR images that best aligned with the tumor that was identified on the specimen by the pathologist. In each case, tumor ROIs of the peripheral zone and/or central gland (which consists of the central and transition zones) were outlined if they were present. A normal-tissue fo- cus was also outlined in the peripheral zone in locations that the pathologist indicated as normal (the pathologist Philips Healthcare) between March
2008 and March 2010. Dataset B was reported in a previous study (9). Pa- tient age and prostate-specific antigen at diagnosis are summarized in Table 1.
MR Image Acquisition
An endorectal coil (Medrad; Bayer Healthcare, Warrendale, Pa) inflated with air and a phased-array surface coil were used for all MR images ex- cept for DCE MR sequences in dataset A, which were acquired with only a phased-array surface coil. Immediately before the MR examination, 1 mg glu- cagon (Glucagon; Lilly, Indianapolis, Ind) was injected intramuscularly to reduce peristalsis of the rectum. Axial, coronal, and sagittal T2-weighted im- ages, axial T1-weighted images, axial free-breathing DW MR images, and axial free-breathing DCE MR images of the entire prostate were acquired.
Orientation of axial images was per- pendicular to the rectal wall and guided by sagittal images. Acquisition of DCE MR images started 30 seconds before intravenous administration of 0.1 mmol/kg gadodiamide (Omniscan;
GE Healthcare) followed by a 20-mL saline flush at a rate of 2.0 mL/sec.
Additional image acquisition parame- ters are detailed in Tables 2 and 3 for datasets A and B, respectively.
Histologic Analysis and MR Correlation For this study, archived tissue sections from the entire prostate were re-eval- uated for use in the histologic analysis and MR correlation. All tissue sections of prostatectomy specimens from the 119 patients were retrieved from the De- partment of Pathology tissue archive of histologic and MR correlation analysis.
Prostatectomy specimens were cut se- rially into 4-mm-thick blocks from the apex to the base in transverse planes and then halved or quartered depend- ing on size. After fixation in 5% buffered formalin, each block was processed and embedded into paraffin. Finally, 4–5 mm microtomed sections were obtained and stained with hematoxylin-eosin.
A genitourinary pathologist (T.A., 8 years of experience in genitourinary pathologic analysis) and a radiologist
Table 1
Patient and ROI Characteristics
Parameter Dataset A Dataset B
No. of patients 71 48
Median patient age (y)* 61.0 (47–75) 62.5 (44–73)
Median PSA (ng/mL)* 5.3 (0.52–65.0)† 7.0 (0.8–256.0)‡
Total no. of ROIs 161 104
No. of normal ROIs 59/161 (36.6) 43/104 (41.3)
No. of cancer ROIs 102/161(63.4) 61/104 (58.7)
No. of peripheral zone cancer ROIs 94/102 (92.2) 47/61 (77.0)
No. of central gland cancer ROIs 8/102 (7.8) 14/61 (23.0)
Median ROI size (mm2)* 55.0 (16.5–312.6) 50.4 (8.9–393.3)
Patient Gleason score
6 23/102 (22.5) 14/61 (23.0)
7 57/102 (55.9) 32/61 (52.5)
8 10/102 (9.8) 9/61 (14.8)
9 12/102 (11.8) 6/61 (9.8)
Note.—Unless otherwise indicated, data in parentheses are percentages. PSA = prostate-specific antigen.
* Data in parentheses are the range.
† PSA data were missing in one patient.
‡ PSA data were missing in six patients.
to other sequences. Then, the study ra- diologist reviewed the transferred ROIs in all cases to confirm the accuracy of the ROI locations. In cases where mis- alignments were visually obvious, the ROI locations were manually shifted without modifying their size or shape.
Image Feature Analysis
Three image features were previously identified as potentially effective in the task of differentiating PC foci from nor- mal peripheral zone ROIs: 10th per- centile ADC value, average ADC value, and the skewness of T2-weighted signal- intensity histogram (9). Other image features that were previously reported in the literature, including DCE image features, were not among the stron- gest identified in the previous study (9), which is generally in agreement with current clinical experience and literature (10,11). These three image features are briefly summarized here.
Pixel-wise ADC values were calcu- lated from DW MR images by using a monoexponential model,
1 0 1
− ,
= nDW nDWb ADC
b
where DW0 and DWb are DW MR signal intensities with DW of 0 and b values, respectively, and 1n represents natural logarithm (12,13). When more than two b values were available, we estimated the pixel-wise ADC value by using a linear least-squares fit to DW MR image data of all available b values (on rare occasions it was necessary to set the pixel-wise ADC value to 0 when a fit yielded a negative value). The av- erage and 10th percentile ADC values were calculated from the histogram of pixel-wise ADC values within each ROI. The average ADC value was ex- pected to help characterize restricted water diffusion in tumors, and the 10th percentile ADC value was expected to help characterize sparse tumors (tu- mor intermixed with normal tissue) more effectively than the average ADC value.
The skewness of T2-weighted sig- nal intensity was calculated from the Table 2 MR Image Acquisition Parameters for Dataset A SequenceSequence TypeTR (msec)TE (msec)Field of View (mm)*MatrixIn-plane Resolution (mm2)Section Thickness (mm)Flip Angle (degrees)No. of Signals Acquired T2 weighted (axial)FSE2550–875087–95140–260128–288 3 128–1920.27 3 0.27–1.02 3 1.023, 5901–2 T2 weighted (sagittal)FSE2000–496786–95150–180256–320 3 192–2240.29 3 0.29–0.70 3 0.703, 4901.5–2 T2 weighted (coronal)FSE2350–715087–90140, 150256–288 3 1920.27 3 0.27–0.59 3 0.593901.5–2 DW (axial)†EPI SE8000–1200029–93140–340128 3 1280.55 3 0.55–1.33 3 1.333–5902, 4 DCE (axial)‡FFE3.2–4.31.6–1.7240192 3 1280.94 3 0.944–6120.38–0.8 Note.— Dataset A was GE Healthcare. An array spatial sensitivity encoding technique (parallel imaging) factor of two was used in all sequences. Total imaging acquisition time was approximately 45 minutes. EPI SE = echo-planar imaging spin echo, FFE = fast field echo, FSE = fast spin echo, TE = echo time, TR = repetition time. * Listed is the length of one side of a square field of view. † Diffusion weighting factor, b values, were 0 and 500 sec/mm2 in one patient (1.4%; one of 71) and 0 and 1000 sec/mm2 in 70 patients.. ‡ Four cases were excluded because their DCE MR images were missing. Approximately 30–50 DCE MR images were acquired in 5–7 minutes at a temporal resolution of 5–12 seconds.
histogram of T2-weighted signal inten- sity within each ROI. It was defined as
3 ,
where S denotes the summation oper- ator, Si denotes T2-weighted signal in- tensity of the i pixel, m and s denote the mean and standard deviation of the T2-weighted signal-intensity histogram, and N denotes the total number of pixels within the ROI. Positive skewness would indicate that the ROI contains more dark pixels than bright pixels, which is typical in tumors, and nega- tive or zero skewness would indicate fewer or equal number of dark pixels compared with bright pixels, which is typical in normal-tissue ROIs.
MR images were retrieved from a picture archiving and communication system, and all image analyses were performed offline by using in-house computer software written in the Py- thon programming language (version 2.6.5; www.Python.com).
Statistical Analysis
Receiver operating characteristic (ROC) analysis was used to characterize the effectiveness of the image features to differentiate PC foci from normal-tissue ROIs (14). Maximum-likelihood–esti- mated proper binormal ROC curves were obtained (15), and the area under the ROC curve (AUC) was used as a figure of merit (16). Both per-ROI and per-patient analyses were performed.
In the per-ROI analysis, each ROI was used as an independent unit of analysis.
In the per-patient analysis, two or more cancer (and, independently, normal-tis- sue) ROIs within a single patient were combined. Minimum ADC features and maximum T2-weighted skewness of the ROIs (all indicative of PC) were used as the combined features.
We combined the three image features by using linear discriminant analysis classifiers (17). For analysis of dataset A or dataset B alone, a leave-one-patient-out cross-validation method was used to separate training from testing of the linear discriminant Table 3 MR Image Acquisition Parameters for Dataset B SequenceSequence TypeTR (msec)TE (msec)Field of View (mm)*MatrixIn-plane Resolution (mm2)Section Thickness (mm)Flip Angle (degrees)No. of Signals Acquired T2-weighted (axial)FSE3166–956170–120140–180216–360 3 189–3500.44 3 0.44–0.56 3 0.563, 4902–4 T2-weighted (sagittal)FSE2186–837490, 120160–200224–276 3 188–2400.47 3 0.47–0.50 3 0.503, 4902–4 T2-weighted (coronal)FSE2208–613290, 120140–180224–328 3 192–2600.44 3 0.44–0.50 3 0.503, 4902–4 DW (axial)†FSE EPI2948–861671–85160–36080–180 3 78–1780.81 3 0.81–1.28 3 1.283–6901–4 DCE (axial)‡FFE 3.3–5.41.1–2.6300–370140–292 3 136–1990.63 3 0.63–1.25 3 1.254–810–401 Note.—Dataset B was Phillips Healthcare. An effective sensitivity encoding (parallel imaging) factor of two was used in all sequences. Total imaging acquisition time was approximately 45 minutes. FFE = fast field echo, FSE = fast spin echo, FSE EPI = FSE with echo-planar imaging readout, TE = echo time, TR = repetition time. * Listed is the length of one side of a square field of view. † Diffusion weighting factor, b values, were 0, 50, 200, 1500, and 2000 sec/mm2 in 24 patients (50%; 24 of 48), and 0 and 1000 sec/mm2 in the other 24 patients. ‡ Approximately 100–120 DCE MR images were acquired in 5–10 minutes at a temporal resolution of 3–6 seconds.
information. The distributions of these feature values in Figure 4 were quali- tatively similar between the two image datasets.
Combining the three image features with a linear discriminant analysis classifier separately for both datas- ets by using the leave-one-patient-out cross validation method yielded AUC values of 0.88 6 0.03 and 0.95 6 0.02 for datasets A and B, respectively (Table 4). When the linear discrim- inant analysis classifier was trained on dataset B and tested on dataset A, AUC values of 0.89 6 0.03 (per ROI) and 0.90 6 0.03 (per patient) were ob- tained. When the classifier was trained on dataset A and tested on dataset B, AUC values of 0.96 6 0.02 (per ROI) and 0.97 6 0.02 (per patient) were obtained (Fig 5). AUC value difference tissue, separately on datasets A and B,
are shown in Table 4. The 10th percen- tile ADC value was the most effective on both datasets, with AUC values of 0.89 6 0.03 (standard error) for da- taset A and 0.92 6 0.03 for dataset B (per-ROI analysis) and 0.89 6 0.03 and 0.93 6 0.03 (per-patient analysis), respectively.
The 10th percentile and the average ADC values were highly correlated (Fig 3).
Both the 10th percentile and the average ADC values were slightly greater for da- taset A than for dataset B, but the over- lap between PC and normal-tissue ROIs appeared to be similar between the two image datasets (Fig 3). The 10th percentile ADC value and T2-weighted skewness were not correlated (Fig 4), which indicates that T2-weighted skew- ness provided additional independent analysis classifier; ROIs from each pa-
tient were used, in turn, to test the linear discriminant analysis classifier while ROIs from all other patients were used to train the classifier, and subse- quently, ROC analysis was conducted on the test results of all patients in aggregate (18). Furthermore, cross- dataset validation was conducted by training a linear discriminant analysis classifier on dataset A and testing it on dataset B and vice versa. ROC curves were compared statistically between the two image datasets in terms of AUC (16).
The Spearman rank-order cor- relation coefficient was calculated to characterize the correlation strength between each image feature and the ROI-specific Gleason score. The Pear- son correlation coefficient was used to characterize the correlation strength between two image features (19).
All statistical tests were two sided, and P value less than .05 was the crit- ical value that indicated statistical sig- nificance. We calculated 10 P values for comparisons of AUC values and 14 P values for correlation coefficients; appli- cation of Bonferroni correction for mul- tiple comparisons would cause the crit- ical value for statistical significance to be adjusted to P value less than .005 for the AUC comparisons and P less than .004 for the correlation coefficients, or P value less than .002 for all. Statistical analyses were performed by using in- house computer software (Python), and ROC analysis was performed by using software developed by Metz (20).
Results
Two example cases of T2-weighted MR images and ADC maps obtained by using MR imagers from the two man- ufacturers with the corresponding his- tologic tissue sections are shown in Fig- ures 1 and 2, respectively. In both of these cases, ROIs were outlined on the T2-weighted MR images and then auto- matically transferred to the ADC maps.
The AUC values of the 10th percen- tile ADC value, average ADC value, and T2-weighted skewness in differentiating PC ROIs from normal peripheral zone
Figure 1
Figure 1: A, Example T2-weighted MR image, and, B, ADC map in a 69-year-old patient in dataset A. The patient’s tumor had a Gleason score of 3 + 4 = 7, outlined in A and then automatically transferred to the ADC map (B). C, D, Corresponding dissected tissue sections show the same tumor (arrow). (Total magnifica- tion 32.5 in C and 350 in D.) MR images are enlarged to 300% of the original size, and window and level are adjusted individually for better visualization.
ROI-specific Gleason scores (Fig 6).
The correlation appeared to be similar between the two image datasets, al- though some differences were apparent for PC ROIs with Gleason scores of 8 or 9 (Fig 6). T2-weighted skewness did not appear to correlate with ROI-specific Gleason scores, which yielded Spear- man rank-order correlation coefficients of 20.05 (P = .65) and 20.07 (P = .59) for datasets A and B, respectively.
Discussion
Validation of quantitative image fea- tures across MR imagers made by different manufacturers is important because proprietary pulse sequences, signal processing, and image process- ing can affect the appearance of images and, consequently, affect quantitative image features and their effectiveness in CAD image analysis. In this study, we evaluated the effectiveness of three image features in the differentiation of PC and normal-tissue ROIs, and in their correlation with lesion-specific Gleason scores on images acquired from MR imagers made by two different manu- facturers. Our results indicate that the 10th percentile ADC value, the average ADC value, and T2-weighted signal- intensity skewness are consistently robust image features across both the 1.5-T image datasets and can be used for prostate MR CAD development.
We estimated pixel-wise ADC values with a monoexponential model.
Given the different DW MR acquisi- tion protocols between the two im- age datasets (different b values and different numbers of b values), it is interesting to note that the discrim- inant ability of the ADC features was not significantly affected. This is consistent with previous reports that indicate that although the use of dif- ferent b values can cause systematic changes in ADC values, the diagnos- tic effectiveness of the ADC value is not necessarily degraded (21,22). The 10th percentile ADC value was con- sistently more effective than the av- erage ADC value for differentiation of PC ROI from normal peripheral zone tissue. This is probably because the Both the 10th percentile and the
average ADC values were moderately and negatively correlated with the was not statistically significant in both
per-ROI and per-patient analyses (P = .05 for both).
Figure 2
Figure 2: A, Example T2-weighted MR image and, B, ADC map in a 65-year-old patient in dataset B. The patient had a Gleason score 4+3 = 7 tumor, outlined in A and then automatically transferred to the ADC map (B). C, D, Corresponding dissected tissue sections show the same tumor (arrow). (Total magnification, 32.5 in C and 350 in D.) MR images are enlarged to 300% of the original size and with window and level adjusted individually for better visualization.
Table 4
AUC Values of the Image Features for Distinguishing PC Foci from Normal Peripheral Zone Tissue ROIs
Image Feature Per-ROI Analysis Per-Patient Analysis
Dataset A Dataset B P Value* Dataset A Dataset B P value*
10th percentile ADC
0.89 6 0.03 0.92 6 0.03 .7 0.89 6 0.03 0.93 6 0.03 .4 Average ADC 0.87 6 0.03 0.89 6 0.03 .7 0.87 6 0.03 0.91 6 0.03 .3 T2-weighted
skewness
0.68 6 0.04 0.86 6 0.04 .003 0.71 6 0.05 0.83 6 0.04 .06 All features
combined†
0.88 6 0.03 0.95 6 0.02 .08 0.89 6 0.03 0.95 6 0.02 .07
Note.—Data shown are maximum-likelihood estimate AUC 6 standard error.
* P value is in terms of the difference in AUC values between the two datasets.
† The three features are combined with a linear discriminant analysis classifier and evaluated with the leave-one-patient-out cross-validation method.
by the AUC value, is apparently not as consistent between the two image data- sets as that of the two ADC image fea- tures. There are two possible reasons may also exist (12). Further studies
are needed to confirm this.
The effectiveness of T2-weighted signal-intensity skewness, characterized 10th percentile ADC value is more
representative of the most aggressive and densely packed portions of the lesions, between which normal tissue
Figure 3
Figure 3: Scatterplots of the 10th percentile and the average ADC values for normal (green circles) and prostate cancer (red squares) ROIs for, A, image dataset A and, B, image dataset B. The Pearson correlation coefficient and the number of ROIs are also shown. In four cancer ROIs in dataset A, the 10th percentile ADC value was set to 0 because more than 10% of the pixels in the ROIs had pixel-wise ADC value set to 0.
Figure 4
Figure 4: Scatterplots of the 10th percentile ADC value and T2-weighted signal-intensity histogram skewness for normal (green circles) and prostate cancer (red squares) ROIs for, A, image dataset A and, B, image dataset B. The Pearson correlation coefficient and the number of ROIs are also shown. In four cancer ROIs in dataset A, the 10th percentile ADC value was set to 0 because more than 10% of the pixels in the ROIs had pixel-wise ADC value set to 0.
of patients. Second, patient selection bias exists because all patients had PC and had undergone prostatectomy.
Further, normal-tissue ROIs were iden- tified from patients who had PC. This clustering of cancer and normal-tissue ROIs within a single patient could have affected our statistical analysis, which therefore requires further validation.
However, these biases are common limitations to many contemporary stud- ies of prostate MR imaging because prostate MR imaging is clinically per- formed only on select patients. Third, this study focused on differentiation of PC from normal tissue in the periph- eral zone. Differentiation of PC from normal tissue in the central gland and from benign prostatic hyperplasia and other benign abnormalities will need to be investigated in the future. Fourth, histologic analysis was done on con- ventional (ie, dissected) rather than whole-mount sections. Future studies on whole-mount sections may improve the accuracy of the MR–histologic cor- relation analysis of lesions. Fifth, the ADC value was estimated from various DW MR image acquisition protocols in the absence of a consensus in the lit- erature on how to optimize DW MR image acquisition. This variation in the acquisition protocol is expected to have affected estimated ADC values, but it does not appear to have degraded the diagnostic effectiveness of ADC values.
Finally, there are uncertainties in the selection of ROIs, and small cancer foci (,5 mm) were not included in this study. Further studies will be needed to address these limitations.
In conclusion, this validation study of quantitative multiparametric MR im- age features shows that 10th percentile ADC value, average ADC value, and T2-weighted signal-intensity skewness are consistently effective across MR imagers made by two manufacturers to distinguish PC from normal-tissue ROIs, and that the 10th percentile and the average ADC values are sig- nificantly, moderately, and negatively correlated with lesion-specific Glea- son scores, which is consistent across imagers made by two manufacturers.
Quantitative image analysis has the 10th percentile and the average ADC
values are consistent between the two image datasets. Lower ADC values are consistently associated with higher lesion-specific Gleason scores. The ap- parent inconsistencies for high-grade (Gleason score 8 or 9) tumors are probably the result of small number of patients with high-grade tumors. This observation has been previously noted in the literature (23–26). It has been hypothesized (27) that tissue that is dense with cells limits intercellular space, which restricts water molecule diffusion and, thus, results in reduced ADC values. For PC, high-grade tumors are associated with poorly differenti- ated and often packed epithelial cells compared with low-grade tumors that have at least some individual glandular structures, which preserve some (albeit reduced) intercellular space (28). This may be a reason for the ADC value to be negatively correlated with the Gleason score. However, large interpa- tient variations are expected to cause overlap in the ADC values between high-grade and low-grade tumors and may explain the moderate correlation strength (23–27).
This study has several limitations.
First, it is a retrospective study from a single institution with a limited number for this. First, the T2-weighted images
in dataset B were acquired with a larger number of averages and had better sig- nal-to-noise ratio than those in dataset A. Second, the T2-weighted images in dataset A often had stronger inhomo- geneity in sensitivity across the field of view than those in dataset B. These differences could have caused degrada- tion in the effectiveness of T2-weighted signal-intensity skewness on dataset A.
It is encouraging that the three im- age features, combined with a linear discriminant analysis classifier, appear to be consistently effective between the two datasets for distinguishing between PC and normal-tissue ROIs. The cross- validation AUC values of 0.89 6 0.03 and 0.96 6 0.02, which were obtained when the linear discriminant analysis classifier was trained on one image data- set and tested on the other, are consis- tent with the leave-one-out AUC values obtained on each image dataset alone (Table 4). Given obvious differences in DW MR image acquisition and DW MR and T2-weighted image appearance between the two datasets, the results suggest that these image features are robust across the MR imagers.
The statistically significant, moder- ate, and negative correlation between lesion-specific Gleason score and the
Figure 5
Figure 5: Estimated proper binormal (solid curves) and empirical (dotted curves), cross- validation ROC curves in the per-ROI analysis. Red curves:
training the linear discriminant analysis classifier on dataset B and testing it on dataset A;
green curves: training the linear discriminant analysis on dataset A and testing it on dataset B.
The values of AUC are also shown and their difference is not statistically significant (P = .05). The results of the per-patient analysis are similar (not shown).
4. Kozlowski P, Chang SD, Jones EC, Berean KW, Chen H, Goldenberg SL. Combined diffusion-weighted and dynamic contrast- enhanced MRI for prostate cancer diagno- sis—correlation with biopsy and histopa- thology. J Magn Reson Imaging 2006;24(1):
108–113.
5. Tanimoto A, Nakashima J, Kohno H, Shin- moto H, Kuribayashi S. Prostate cancer screening: the clinical value of diffusion- weighted imaging and dynamic MR imaging in combination with T2-weighted imaging. J Magn Reson Imaging 2007;25(1):146–152.
6. Vos PC, Barentsz JO, Karssemeijer N, Huis- man HJ. Automatic computer-aided detec- tion of prostate cancer based on multipara- metric magnetic resonance image analysis.
Phys Med Biol 2012;57(6):1527–1542.
7. Puech P, Betrouni N, Makni N, Dewalle AS, Villers A, Lemaitre L. Computer-assisted di- agnosis of prostate cancer using DCE-MRI data: design, implementation and prelimi- nary results. Int J CARS 2009;4(1):1–10.
conflicts of interest to disclose. A.O. Financial activities related to the present article: none to disclose. Financial activities not related to the present article: grant from Philips Healthcare;
payment for development of educational presen- tations from Philips Healthcare; participation in one GE Healthcare advisory board meeting on prostate cancer. Other relationships: none to disclose.
References
1. Siegel R, Naishadham D, Jemal A. Can- cer statistics, 2012. CA Cancer J Clin 2012;62(1):10–29.
2. Hoeks CM, Barentsz JO, Hambrock T, et al.
Prostate cancer: multiparametric MR imag- ing for detection, localization, and staging.
Radiology 2011;261(1):46–66.
3. Moyer VA; U.S. Preventive Services Task Force. Screening for prostate cancer: U.S.
Preventive Services Task Force recom- mendation statement. Ann Intern Med 2012;157(2):120–134.
potential to assist radiologists to better interpret multiparametric prostate MR images.
Acknowledgement: This work was supported in part by the U.S. Army Medical Research and Material Command Prostate Cancer Research Program through an Idea Development Award (PC093485).
Disclosures of Conflicts of Interest: Y.P. No relevant conflicts of interest to disclose. Y.J. No relevant conflicts of interest to disclose. T.A. No relevant conflicts of interest to disclose. M.L.G.
Financial activities related to the present article:
none to disclose. Financial activities not related to the present article: travel reimbursement and board membership from the International Society for Optics and Photonics board; board membership on Board of Quantitative Insights;
receives money for patents through university’s Tech office; author and author’s institution re- ceives royalties from IP licensed through univer- sity’s Tech office; author and author’s institution receives money from stock and stock options through the university’s Tech office. Other re- lationships: none to disclose. S.E.E. No relevant
Figure 6
Figure 6: Boxplots that show correlation between lesion specific Gleason scores and the 10th percentile ADC value (top), between lesion specific Gleason scores and the average ADC value (bottom), for dataset A (left), and for dataset B (right). The red horizontal lines denote medians, the boxes denote interquartile (the 25th percentile to the 75th percentile) range, and the data points outside the whiskers denote outliers. The Spearman correlation coefficient and the number of ROIs, n, are also shown. Normal ROIs are plotted for comparison only (not included in calculation of correlation coefficients). (In four cancer ROIs in dataset A, the 10th percentile ADC value was set to zero because more than 10% of the pixels in the ROIs had pixel-wise ADC value set to zero.)
8. Niaf E, Rouvière O, Mège-Lechevallier F, Bratan F, Lartizien C. Computer-aided di- agnosis of prostate cancer in the peripheral zone using multiparametric MRI. Phys Med Biol 2012;57(12):3833–3851.
9. Peng Y, Jiang Y, Yang C, et al. Quantitative analysis of multiparametric prostate MR images: differentiation between prostate cancer and normal tissue and correlation with Gleason score—a computer-aided diagnosis development study. Radiology 2013;267(3):787–796.
10. Tretiakova M, Antic T, Binder D, et al. Mi- crovessel density is not increased in pros- tate cancer: digital imaging of routine sec- tions and tissue microarrays. Hum Pathol 2013;44(4):495–502.
11. Oto A, Kayhan A, Jiang Y, et al. Prostate cancer: differentiation of central gland cancer from benign prostatic hyperplasia by using diffusion-weighted and dynamic contrast-enhanced MR imaging. Radiology 2010;257(3):715–723.
12. Langer DL, van der Kwast TH, Evans AJ, et al. Intermixed normal tissue within pros- tate cancer: effect on MR imaging measure- ments of apparent diffusion coefficient and T2—sparse versus dense cancers. Radiology 2008;249(3):900–908.
13. Chenevert TL, Galbán CJ, Ivancevic MK, et al. Diffusion coefficient measurement using a temperature-controlled fluid for quality control in multicenter studies. J Magn Re- son Imaging 2011;34(4):983–987.
14. Metz CE. ROC methodology in radiologic imaging. Invest Radiol 1986;21(9):720–733.
15. Metz CE, Pan X. “Proper” binormal ROC curves: Theory and maximum-likelihood es- timation. J Math Psychol 1999;43(1):1–33.
16. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operat- ing characteristic (ROC) curve. Radiology 1982;143(1):29–36.
17. Fukunaga K. Introduction to statistical pat- tern recognition. 2nd ed. Boston, Mass: Ac- ademic Press, 1990; 131–136.
18. Fukunaga K. Introduction to statistical pat- tern recognition. 2nd ed. Boston, Mass: Ac- ademic Press, 1990; 219–224.
19. Riffenburgh RH. Statistics in medicine. 2nd ed. Amsterdam, the Netherlands: Elsevier Academic Press, 2006; 454–456.
20. Metz CE. ROC software. http://metz-roc.
uchicago.edu. Accessed June 25, 2012.
21. Thörmer G, Otto J, Reiss-Zimmermann M, et al. Diagnostic value of ADC in patients with prostate cancer: influence of the choice of b values. Eur Radiol 2012;22(8):1820–
1828.
22. Peng Y, Jiang Y, Antic T, et al. Apparent dif- fusion coefficient (ADC) for prostate cancer imaging: the impact of the b-values. AJR Am J Roentgenol (in press).
23. Tamada T, Sone T, Jo Y, et al. Apparent dif- fusion coefficient values in peripheral and transition zones of the prostate: comparison
between normal and malignant prostatic tis- sues and correlation with histologic grade.
J Magn Reson Imaging 2008;28(3):720–
726.
24. Woodfield CA, Tung GA, Grand DJ, Pezzullo JA, Machan JT, Renzulli JF 2nd. Diffusion- weighted MRI of peripheral zone prostate cancer: comparison of tumor apparent diffusion coefficient with Gleason score and percentage of tumor on core biopsy.
AJR Am J Roentgenol 2010;194(4):W316–
W322.
25. Turkbey B, Shah VP, Pang Y, et al. Is ap- parent diffusion coefficient associated with clinical risk scores for prostate cancers that are visible on 3-T MR images? Radiology 2011;258(2):488–495.
26. Verma S, Rajesh A, Morales H, et al. As- sessment of aggressiveness of prostate cancer: correlation of apparent diffusion coefficient with histologic grade after rad- ical prostatectomy. AJR Am J Roentgenol 2011;196(2):374–381.
27. Itou Y, Nakanishi K, Narumi Y, Nishizawa Y, Tsukuma H. Clinical utility of apparent diffusion coefficient (ADC) values in pa- tients with prostate cancer: can ADC values contribute to assess the aggressiveness of prostate cancer? J Magn Reson Imaging 2011;33(1):167–172.
28. Epstein JI, Netto GJ. Biopsy interpretation of the prostate. 4th ed. Philadelphia, Pa: Lip- pincott Williams & Wilkins, 2007; 175–202.
