Multicenter Multireader Evaluation of an Artificial Intelligence-Based Attention Mapping System for the Detection of Prostate Cancer With Multiparametric MRI.

OBJECTIVE. The purpose of this study was to evaluate in a multicenter dataset the performance of an artificial intelligence (AI) detection system with attention mapping compared with multiparametric MRI (mpMRI) interpretation in the detection of prostate cancer. MATERIALS AND METHODS. MRI examinations from five institutions were included in this study and were evaluated by nine readers. In the first round, readers evaluated mpMRI studies using the Prostate Imaging Reporting and Data System version 2. After 4 weeks, images were again presented to readers along with the AI-based detection system output. Readers accepted or rejected lesions within four AI-generated attention map boxes. Additional lesions outside of boxes were excluded from detection and categorization. The performances of readers using the mpMRI-only and AI-assisted approaches were compared. RESULTS. The study population included 152 case patients and 84 control patients with 274 pathologically proven cancer lesions. The lesion-based AUC was 74.9% for MRI and 77.5% for AI with no significant difference (p = 0.095). The sensitivity for overall detection of cancer lesions was higher for AI than for mpMRI but did not reach statistical significance (57.4% vs 53.6%, p = 0.073). However, for transition zone lesions, sensitivity was higher for AI than for MRI (61.8% vs 50.8%, p = 0.001). Reading time was longer for AI than for MRI (4.66 vs 4.03 minutes, p < 0.001). There was moderate interreader agreement for AI and MRI with no significant difference (58.7% vs 58.5%, p = 0.966). CONCLUSION. Overall sensitivity was only minimally improved by use of the AI system. Significant improvement was achieved, however, in the detection of transition zone lesions with use of the AI system at the cost of a mean of 40 seconds of additional reading time.

A Prospective Accuracy Study of Prostate Imaging Reporting and Data System Version 2 on Multiparametric Magnetic Resonance Imaging in Detecting Clinically Significant Prostate Cancer With Whole-mount Pathology.

OBJECTIVE: To assess the accuracy of Prostate Imaging Reporting and Data System version 2 (PI-RADS v2) in detecting clinically significant prostate cancer (csPCa) on multiparametric magnetic resonance imaging (mpMRI) using whole-mount sections after radical prostatectomy (RP) as reference standard. METHODS: Forty-eight patients undergoing mpMRI before RP were prospectively enrolled. Two experienced radiologists independently scored and mapped imaging findings according to PI-RADS v2. One experienced uropathologist mapped cancers detected on whole-mount sections using the PI-RADS v2 sector scheme. Per-lesion and per-patient analyses were run. Primary outcomes were sensitivity and false discovery rate (FDR) in detecting csPCa using PI-RADS v2 score ≥3 and ≥4 as thresholds. Secondary outcome was inter-reader agreement. RESULTS: On the per-lesion analysis, sensitivity and FDR at the PI-RADS v2 threshold score ≥3 were 0.75 and 0.17 for Reader 1, and 0.67 and 0.13 for Reader 2, respectively. At the PI-RADS v2 threshold score ≥4, sensitivity was slightly lower, and FDR nearly halved for both readers. On the per-patient analysis, sensitivity for csPCa at the PI-RADS v2 threshold score ≥3 was 0.85 for Reader 1, and 0.78 for Reader 2. At the PI-RADS v2 threshold score ≥4, sensitivity was slightly lower for both readers. Inter-reader agreement was substantial (k 0.72 and 0.65 for PI-RADS v2 threshold score ≥3 and ≥4, respectively). CONCLUSION: In our prospective study with pathology after RP as standard of reference, PI-RADS v2 showed good sensitivity in detecting csPCa on mpMRI with substantial agreement between 2 experienced readers. Threshold score ≥4 had lower FDR.

Can computer-aided diagnosis assist in the identification of prostate cancer on prostate MRI? a multi-center, multi-reader investigation.

For prostate cancer detection on prostate multiparametric MRI (mpMRI), the Prostate Imaging-Reporting and Data System version 2 (PI-RADSv2) and computer-aided diagnosis (CAD) systems aim to widely improve standardization across radiologists and centers. Our goal was to evaluate CAD assistance in prostate cancer detection compared with conventional mpMRI interpretation in a diverse dataset acquired from five institutions tested by nine readers of varying experience levels, in total representing 14 globally spread institutions. Index lesion sensitivities of mpMRI-alone were 79% (whole prostate (WP)), 84% (peripheral zone (PZ)), 71% (transition zone (TZ)), similar to CAD at 76% (WP, p=0.39), 77% (PZ, p=0.07), 79% (TZ, p=0.15). Greatest CAD benefit was in TZ for moderately-experienced readers at PI-RADSv2 <3 (84% vs mpMRI-alone 67%, p=0.055). Detection agreement was unchanged but CAD-assisted read times improved (4.6 vs 3.4 minutes, p<0.001). At PI-RADSv2 ≥ 3, CAD improved patient-level specificity (72%) compared to mpMRI-alone (45%, p<0.001). PI-RADSv2 and CAD-assisted mpMRI interpretations have similar sensitivities across multiple sites and readers while CAD has potential to improve specificity and moderately-experienced radiologists' detection of more difficult tumors in the center of the gland. The multi-institutional evidence provided is essential to future prostate MRI and CAD development.

Do clinical and radiologic features help predict malignancy of B3 breast lesions without epithelial atypia (B3a)?

OBJECTIVES: To evaluate excision histology outcome of B3a lesions diagnosed at imaging-guided core biopsy and assess whether clinical and radiologic features may predict upgrade to malignancy. METHODS: A total of 153 B3a lesions (74 papillomas, 51 radial scars and 28 fibro-epithelial lesions) that underwent surgical excision were assessed. PPV for malignancy in all B3a lesions and specific for each sub-category was evaluated. Multivariate analysis was conducted to identify association between clinical (age, family or personal history of breast cancer, symptoms), diagnostic findings (imaging modality, lesion size, final BI-RADS category) and final excision outcome. RESULTS: Eleven (7%) of 153 B3a lesions were upgraded to malignancy. All carcinomas diagnosed on excision were non-high grade DCIS. Following features were significantly associated with malignancy: lesions identified only on mammography (3/21; 14%), both mammography and ultrasound (7/44; 16%), lesion size > 10 mm (10/74; 13.5%) and BI-RADS category 4-5 (8/29; 27.6%). The absence of residual microcalcification after biopsy was associated with decreased risk for malignancy (15/16[93.7%]; p = 0.0297). Lesion size > 10 mm (OR = 9.3832; 95%; p = 0.0398) and BI-RADS category 4-5 (OR = 12.6004; 95%; p = 0.0006) were found to be independent predictors of upgrade to malignancy. CONCLUSIONS: B3a lesions are associated with low risk of malignancy at excision. Lesion size > 10 mm and BI-RADS 4-5 category may represent useful predictors of upgrade to malignancy.

Prospective comparison of PI-RADS version 2 and qualitative in-house categorization system in detection of prostate cancer.

BACKGROUND: Prostate Imaging-Reporting and Data System v. 2 (PI-RADSv2) provides standardized nomenclature for interpretation of prostate multiparametric MRI (mpMRI). Inclusion of additional features for categorization may provide benefit to stratification of disease. PURPOSE: To prospectively compare PI-RADSv2 to a qualitative in-house system for detecting prostate cancer on mpMRI. STUDY TYPE: Prospective. POPULATION: In all, 338 patients who underwent mpMRI May 2015-May 2016, with subsequent MRI/transrectal ultrasound fusion-guided biopsy. FIELD STRENGTH: 3T mpMRI (T2 W, diffusion-weighted [DW], apparent diffusion coefficient [ADC] map, b-2000 DWI acquisition, and dynamic contrast-enhanced [DCE] MRI). ASSESSMENT: One genitourinary radiologist prospectively read mpMRIs using both in-house and PI-RADSv2 5-category systems. STATISTICAL TEST: In lesion-based analysis, overall and clinically significant (CS) tumor detection rates (TDR) were calculated for all PI-RADSv2 and in-house categories. The ability of each scoring system to detect cancer was assessed by area under receiver operator characteristic curve (AUC). Within each PI-RADSv2 category, lesions were further stratified by their in-house categories to determine if TDRs can be increased by combining features of both systems. RESULTS: In 338 patients (median prostate-specific antigen [PSA] 6.5 [0.6-113.6] ng/mL; age 64 [44-84] years), 733 lesions were identified (47% tumor-positive). Predictive abilities of both systems were comparable for all (AUC 76-78%) and CS cancers (AUCs 79%). The in-house system had higher overall and CS TDRs than PI-RADSv2 for categories 3 and 4 (P < 0.01 for both), with the greatest difference between the scoring systems seen in lesions scored category 4 (CS TDRs: in-house 65%, PI-RADSv2 22.1%). For lesions categorized as PI-RADSv2 = 4, characterization of suspicious/indeterminate extraprostatic extension (EPE) and equivocal findings across all mpMRI sequences contributed to significantly different TDRs for both systems (TDR range 19-75%, P < 0.05). DATA CONCLUSION: PI-RADSv2 behaves similarly to an existing validated system that relies on the number of sequences on which a lesion is seen. This prospective evaluation suggests that sequence positivity and suspicion of EPE can enhance PI-RADSv2 category 4 cancer detection. LEVEL OF EVIDENCE: 1 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2018;47:1326-1335.

A Magnetic Resonance Imaging-Based Prediction Model for Prostate Biopsy Risk Stratification.

Importance: Multiparametric magnetic resonance imaging (MRI) in conjunction with MRI-transrectal ultrasound (TRUS) fusion-guided biopsies have improved the detection of prostate cancer. It is unclear whether MRI itself adds additional value to multivariable prediction models based on clinical parameters. Objective: To determine whether an MRI-based prediction model can reduce unnecessary biopsies in patients with suspected prostate cancer. Design, Setting, and Participants: Patients underwent MRI, MRI-TRUS fusion-guided biopsy, and 12-core systematic biopsy in 1 session. The development cohort used to derive the prediction model consisted of 400 patients from 1 institution enrolled between May 14, 2015, and August 31, 2016, and the validation cohort included 251 patients from 2 independent institutions who underwent biopsies between April 1, 2013, and June 30, 2016, at 1 institution and between July 1, 2015, and October 31, 2016, at the other institution. The MRI model included MRI-derived parameters in addition to clinical variables. Area under the curve of receiver operating characteristic curves and decision curve analysis were performed. Main Outcomes and Measures: Risk of clinically significant prostate cancer on biopsy, defined as a Gleason score of 3 + 4 or higher in at least 1 biopsy core. Results: Overall, 193 (48.3%) of the 400 patients in the development cohort (mean [SD] age at biopsy, 64.3 [7.1] years) and 96 (38.2%) of the 251 patients in the validation cohort (mean [SD] age at biopsy, 64.9 [7.2] years) had clinically significant prostate cancer, defined as a Gleason score greater than or equal to 3 + 4. By applying the model to the external validation cohort, the area under the curve increased from 64% to 84% compared with the baseline model (P < .001). At a risk threshold of 20%, the MRI model had a lower false-positive rate than the baseline model (46% [95% CI, 32%-66%] vs 92% [95% CI, 70%-100%]), with only a small reduction in the true-positive rate (89% [95% CI, 85%-96%] vs 99% [95% CI, 89%-100%]). Eighteen of 100 fewer biopsies could have been performed, with no increase in the number of patients with missed clinically significant prostate cancers. Conclusions and Relevance: The inclusion of MRI-derived parameters in a risk model could reduce the number of unnecessary biopsies while maintaining a high rate of diagnosis of clinically significant prostate cancers.

All over the map: An interobserver agreement study of tumor location based on the PI-RADSv2 sector map.

BACKGROUND: Prostate imaging reporting and data system version 2 (PI-RADSv2) recommends a sector map for reporting findings of prostate cancer mulitparametric MRI (mpMRI). Anecdotally, radiologists may demonstrate inconsistent reproducibility with this map. PURPOSE: To evaluate interobserver agreement in defining prostate tumor location on mpMRI using the PI-RADSv2 sector map. STUDY TYPE: Retrospective. POPULATION: Thirty consecutive patients who underwent mpMRI between October, 2013 and March, 2015 and who subsequently underwent prostatectomy with whole-mount processing. FIELD STRENGTH: 3T mpMRI with T2 W, diffusion-weighted imaging (DWI) (apparent diffusion coefficient [ADC] and b-2000), dynamic contrast-enhanced (DCE). ASSESSMENT: Six radiologists (two high, two intermediate, and two low experience) from six institutions participated. Readers were blinded to lesion location and detected up to four lesions as per PI-RADSv2 guidelines. Readers marked the long-axis of lesions, saved screen-shots of each lesion, and then marked the lesion location on the PI-RADSv2 sector map. Whole-mount prostatectomy specimens registered to the MRI served as ground truth. Index lesions were defined as the highest grade lesion or largest lesion if grades were equivalent. STATISTICAL TEST: Agreement was calculated for the exact, overlap, and proportion of agreement. RESULTS: Readers detected an average of 1.9 lesions per patient (range 1.6-2.3). 96.3% (335/348) of all lesions for all readers were scored PI-RADS ≥3. Readers defined a median of 2 (range 1-18) sectors per lesion. Agreement for detecting index lesions by screen shots was 83.7% (76.1%-89.9%) vs. 71.0% (63.1-78.3%) overlap agreement on the PI-RADS sector map (P < 0.001). Exact agreement for defining sectors of detected index lesions was only 21.2% (95% confidence interval [CI]: 14.4-27.7%) and rose to 49.0% (42.4-55.3%) when overlap was considered. Agreement on defining the same level of disease (ie, apex, mid, base) was 61.4% (95% CI 50.2-71.8%). DATA CONCLUSION: Readers are highly likely to detect the same index lesion on mpMRI, but exhibit poor reproducibility when attempting to define tumor location on the PI-RADSv2 sector map. The poor agreement of the PI-RADSv2 sector map raises concerns its utility in clinical practice. LEVEL OF EVIDENCE: 3 Technical Efficacy: Stage 2 J. MAGN. RESON. IMAGING 2018;48:482-490.

18F-DCFBC Prostate-Specific Membrane Antigen-Targeted PET/CT Imaging in Localized Prostate Cancer: Correlation With Multiparametric MRI and Histopathology.

PURPOSE: To assess the ability of (N-[N-[(S)-1,3-dicarboxypropyl]carbamoyl]-4-F-fluorobenzyl-L-cysteine) (F-DCFBC), a prostate-specific membrane antigen-targeted PET agent, to detect localized prostate cancer lesions in correlation with multiparametric MRI (mpMRI) and histopathology. METHODS: This Health Insurance Portability and Accountability Act of 1996-compliant, prospective, institutional review board-approved study included 13 evaluable patients with localized prostate cancer (median age, 62.8 years [range, 51-74 years]; median prostate-specific antigen, 37.5 ng/dL [range, 3.26-216 ng/dL]). Patients underwent mpMRI and F-DCFBC PET/CT within a 3 months' window. Lesions seen on mpMRI were biopsied under transrectal ultrasound/MRI fusion-guided biopsy, or a radical prostatectomy was performed. F-DCFBC PET/CT and mpMRI were evaluated blinded and separately for tumor detection on a lesion basis. For PET image analysis, MRI and F-DCFBC PET images were fused by using software registration; imaging findings were correlated with histology, and uptake of F-DCFBC in tumors was compared with uptake in benign prostatic hyperplasia nodules and normal peripheral zone tissue using the 80% threshold SUVmax. RESULTS: A total of 25 tumor foci (mean size, 1.8 cm; median size, 1.5 cm; range, 0.6-4.7 cm) were histopathologically identified in 13 patients. Sensitivity rates of F-DCFBC PET/CT and mpMRI were 36% and 96%, respectively, for all tumors. For index lesions, the largest tumor with highest Gleason score, sensitivity rates of F-DCFBC PET/CT and mpMRI were 61.5% and 92%, respectively. The average SUVmax for primary prostate cancer was higher (5.8 ± 4.4) than that of benign prostatic hyperplasia nodules (2.1 ± 0.3) or that of normal prostate tissue (2.1 ± 0.4) at 1 hour postinjection (P = 0.0033). CONCLUSIONS: The majority of index prostate cancers are detected with F-DCFBC PET/CT, and this may be a prognostic indicator based on uptake and staging. However, for detecting prostate cancer with high sensitivity, it is important to combine prostate-specific membrane antigen PET/CT with mpMRI.

Positron emission tomography (PET) in primary prostate cancer staging and risk assessment.

Prostate cancer (PCa) is one of the few neoplasms that are not well served by 18F-Fluorodeoxyglucose (FDG) positron emission tomography (PET). As a result, a number of PET tracers have been developed to target particular biological features of PCa. Such agents can be used for diagnosis, staging, identification of biochemical recurrence (BCR) and evaluation of metastatic disease. Here, we focus on primary disease and local staging. To date, magnetic resonance imaging (MRI) has proven superior to PET in the imaging of primary PCa. However, some PET agents have shown remarkable promise in staging high-risk PCa (defined as any combination of a clinical T3, a PSA score >20 ng/mL, or a Gleason score of 8-10), as well as biochemical relapse after definitive therapy and metastatic PCa. PET agents can be divided into those that interrogate tumor metabolism (18F-FDG, 11C-Choline, 18F-Choline, 11C-Acetate, 18F-FACBC), hormone receptors (18F-FDHT), and other targets such as prostate specific membrane antigen (PSMA) (68Ga-PSMA, 18F-DCFBC, 18F-DCFPyl) or gastric releasing peptide (18F-GRP or 18F-Bombesin). In this review, we compare the available PCa targeted PET tracers utilized in staging of high risk tumors.

Prospective Evaluation of PI-RADS™ Version 2 Using the International Society of Urological Pathology Prostate Cancer Grade Group System.

PURPOSE: The PI-RADS™ (Prostate Imaging Reporting and Data System), version 2 scoring system, introduced in 2015, is based on expert consensus. In the same time frame ISUP (International Society of Urological Pathology) introduced a new pathological scoring system for prostate cancer. Our goal was to prospectively evaluate the cancer detection rates for each PI-RADS, version 2 category and compare them to ISUP group scores in patients undergoing systematic biopsy and magnetic resonance imaging-transrectal ultrasound fusion guided biopsy. MATERIALS AND METHODS: A total of 339 treatment naïve patients prospectively underwent multiparametric magnetic resonance imaging evaluated with PI-RADS, version 2 with subsequent systematic and fusion guided biopsy from May 2015 to May 2016. ISUP scores were applied to pathological specimens. An ISUP score of 2 or greater (ie Gleason 3 + 4 or greater) was defined as clinically significant prostate cancer. Cancer detection rates were determined for each PI-RADS, version 2 category as well as for the T2 weighted PI-RADS, version 2 categories in the peripheral zone. RESULTS: The cancer detection rate for PI-RADS, version 2 categories 1, 2, 3, 4 and 5 was 25%, 20.2%, 24.8%, 39.1% and 86.9% for all prostate cancer, and 0%, 9.6%, 12%, 22.1% and 72.4% for clinically significant prostate cancer, respectively. On T2-weighted magnetic resonance imaging the cancer detection rate in the peripheral zone was significantly higher for PI-RADS, version 2 category 4 than for overall PI-RADS, version 2 category 4 in the peripheral zone (all prostate cancer 36.6% vs 48.1%, p = 0.001, and clinically significant prostate cancer 22.9% vs 32.6%, p = 0.002). CONCLUSIONS: The cancer detection rate increases with higher PI-RADS, version 2 categories.