Multiparametric MRI-based radiomics model to predict pelvic lymph node invasion for patients with prostate cancer.

OBJECTIVE: To identify which patient with prostate cancer (PCa) could safely avoid extended pelvic lymph node dissection (ePLND) by predicting lymph node invasion (LNI), via a radiomics-based machine learning approach. METHODS: An integrative radiomics model (IRM) was proposed to predict LNI, confirmed by the histopathologic examination, integrating radiomics features, extracted from prostatic index lesion regions on MRI images, and clinical features via SVM. The study cohort comprised 244 PCa patients with MRI and followed by radical prostatectomy (RP) and ePLND within 6 months between 2010 and 2019. The proposed IRM was trained in training/validation set and evaluated in an internal independent testing set. The model's performance was measured by area under the curve (AUC), sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). AUCs were compared via Delong test with 95% confidence interval (CI), and the rest measurements were compared via chi-squared test or Fisher's exact test. RESULTS: Overall, 17 (10.6%) and 14 (16.7%) patients with LNI were included in training/validation set and testing set, respectively. Shape and first-order radiomics features showed usefulness in building the IRM. The proposed IRM achieved an AUC of 0.915 (95% CI: 0.846-0.984) in the testing set, superior to pre-existing nomograms whose AUCs were from 0.698 to 0.724 (p < 0.05). CONCLUSION: The proposed IRM could be potentially feasible to predict the risk of having LNI for patients with PCa. With the improved predictability, it could be utilized to assess which patients with PCa could safely avoid ePLND, thus reduce the number of unnecessary ePLND. KEY POINTS: • The combination of MRI-based radiomics features with clinical information improved the prediction of lymph node invasion, compared with the model using only radiomics features or clinical features. • With improved prediction performance on predicting lymph node invasion, the number of extended pelvic lymph node dissection (ePLND) could be reduced by the proposed integrative radiomics model (IRM), compared with the existing nomograms.

Optimizing Spatial Biopsy Sampling for the Detection of Prostate Cancer.

PURPOSE: The appropriate number of systematic biopsy cores to retrieve during magnetic resonance imaging (MRI)-targeted prostate biopsy is not well defined. We aimed to demonstrate a biopsy sampling approach that reduces required core count while maintaining diagnostic performance. MATERIALS AND METHODS: We collected data from a cohort of 971 men who underwent MRI-ultrasound fusion targeted biopsy for suspected prostate cancer. A regional targeted biopsy (RTB) was evaluated retrospectively; only cores within 2 cm of the margin of a radiologist-defined region of interest were considered part of the RTB. We compared detection rates for clinically significant prostate cancer (csPCa) and cancer upgrading rate on final whole mount pathology after prostatectomy between RTB, combined, MRI-targeted, and systematic biopsy. RESULTS: A total of 16,459 total cores from 971 men were included in the study data sets, of which 1,535 (9%) contained csPCa. The csPCa detection rates for systematic, MRI-targeted, combined, and RTB were 27.0% (262/971), 38.3% (372/971), 44.8% (435/971), and 44.0% (427/971), respectively. Combined biopsy detected significantly more csPCa than systematic and MRI-targeted biopsy (p <0.001 and p=0.004, respectively) but was similar to RTB (p=0.71), which used on average 3.8 (22%) fewer cores per patient. In 102 patients who underwent prostatectomy, there was no significant difference in upgrading rates between RTB and combined biopsy (p=0.84). CONCLUSIONS: A RTB approach can maintain state-of-the-art detection rates while requiring fewer retrieved cores. This result informs decision making about biopsy site selection and total retrieved core count.

Three Tesla Multiparametric Magnetic Resonance Imaging: Comparison of Performance with and without Endorectal Coil for Prostate Cancer Detection, PI-RADS™ version 2 Category and Staging with Whole Mount Histopathology Correlation.

PURPOSE: We investigated the performance of 3 Tesla multiparametric magnetic resonance imaging with and without an endorectal coil to detect prostate cancer with a whole mount histopathology reference. MATERIALS AND METHODS: This retrospective HIPAA (Health Insurance Portability and Accountability Act) compliant, institutional review board approved, case-control study included patients who underwent 3 Tesla multiparametric magnetic resonance imaging with and without an endorectal coil from July 2009 to December 2016 prior to prostatectomy. The tumor detection rate was calculated for total and index lesions. Lesion magnetic resonance imaging and histopathology features were compared between the 2 groups. Using SPSS®, version 24 p <0.05 was considered significant. RESULTS: A total of 871 whole mount histopathology lesions in 429 patients with a mean ± SD age of 61.8 ± 7 years were included in analysis. The subcohorts with and without an endorectal coil comprised 260 and 169 patients with a total of 529 and 342 lesions, respectively. The overall tumor detection rates in all patients, and in the endorectal coil and nonendorectal coil subcohorts were 49.6% (432 of 871 patients), 50.5% (267 of 529) and 48.2% (165 of 342), respectively. The index tumor detection rates overall, and in the endorectal coil and nonendorectal coil subcohorts were 77.6% (333 of 429 patients), 78.5% (204 of 260) and 76.3% (129 of 169), respectively. In the endorectal coil and nonendorectal coil subcohorts we detected 35.9% (66 of 184) and 48.4% (76 of 157) of anterior lesions (p = 0.019), 58% (200 of 345) and 48.1% (89 of 185) of posterior lesions (p = 0.025), 37.3% (41 of 110) and 54.4% (62 of 114) of transition zone lesions (p = 0.010), and 53.7% (225 of 419) and 45.2% (103 of 228) of peripheral lesions (p = 0.033), respectively. After adjusting for clinical and pathological factors the endorectal coil group only showed higher detection of peripheral and posterior prostate cancer. CONCLUSIONS: We found that 3 Tesla multiparametric magnetic resonance imaging with and without an endorectal coil had similar detection of overall and index prostate cancer. However, the endorectal coil subcohort had significantly higher detection of posterior and peripheral prostate cancer, and lower detection of anterior and transition zone prostate cancer.

Apparent Diffusion Coefficient (ADC) Ratio Versus Conventional ADC for Detecting Clinically Significant Prostate Cancer With 3-T MRI.

OBJECTIVE. The purpose of this study is to evaluate the performance of the apparent diffusion coefficient ratio (ADCratio; the ADC of the suspected prostate cancer [PCa] focus on MRI divided by the ADC in a noncancerous reference area) with that of conventional ADC for detection of high-grade PCa (Gleason score [GS] ≥ 3 + 4) versus low-grade PCa (GS = 3 + 3) with whole-mount (WM) histopathologic analysis used as a reference. MATERIALS AND METHODS. The cohort of this retrospective study included 218 men with 240 unilateral PCa lesions assessed by both 3-T multiparametric MRI and whole-mount histopathologic analysis. ROIs were placed on individual lesions verified by WM histopathologic analysis, to calculate the mean ADC (ADCtumor_mean) and lowest ADC within each lesion (ADCtumor_min), and within non-tumor-containing regions of the same tumor zone but on the contralateral side (ADCbenign), to calculate the background ADC. The ADCratio_mean (the ADCtumor_mean divided by the ADCbenign) was calculated. The performance of individual ADCtumor and ADCratio_mean values in discriminating PCa with a GS of 3 + 3 from PCa with a GS of 3 + 4 or greater was assessed using the AUC value. RESULTS. The ADCratio_mean had a higher AUC value for discriminating PCa lesions with a GS of 3 + 3 from those with a GS of 3 + 4 or greater (the AUC value increased from 0.70 using the ADCtumor_mean and 0.67 using the ADCtumor_min [the minimum ADC of the PCa lesion] to 0.80 for the ADCratio_mean and 0.72 for the ADCratio_min [the ADCtumor_min divided by the ADCbenign]; p = 0.043). When stratified by PCa zonal location, the ADCratio_mean had significantly more robust accuracy in the transition zone (the AUC value increased from 0.63 for ADCtumor_mean to 0.87 for ADCratio_mean; p = 0.019) compared with the peripheral zone (the AUC value increased from 0.74 for ADCtumor_mean to 0.78 for ADCratio_mean; p = 0.44). When analyzed on the basis of endorectal coil use, the ADCratio_mean performed nonsignificantly better in both the endorectal coil and non-endorectal coil subcohorts, although it performed better in the former. CONCLUSION. As an intrapatient-normalized diagnostic tool, the ADC ratio provided the best AUC value for discrimination of low-grade from high-grade PCa on 3-T MRI.

Exploring Uncertainty Measures in Bayesian Deep Attentive Neural Networks for Prostate Zonal Segmentation.

Automatic segmentation of prostatic zones on multiparametric MRI (mpMRI) can improve the diagnostic workflow of prostate cancer. We designed a spatial attentive Bayesian deep learning network for the automatic segmentation of the peripheral zone (PZ) and transition zone (TZ) of the prostate with uncertainty estimation. The proposed method was evaluated by using internal and external independent testing datasets, and overall uncertainties of the proposed model were calculated at different prostate locations (apex, middle, and base). The study cohort included 351 MRI scans, of which 304 scans were retrieved from a de-identified publicly available datasets (PROSTATEX) and 47 scans were extracted from a large U.S. tertiary referral center (external testing dataset; ETD)). All the PZ and TZ contours were drawn by research fellows under the supervision of expert genitourinary radiologists. Within the PROSTATEX dataset, 259 and 45 patients (internal testing dataset; ITD) were used to develop and validate the model. Then, the model was tested independently using the ETD only. The segmentation performance was evaluated using the Dice Similarity Coefficient (DSC). For PZ and TZ segmentation, the proposed method achieved mean DSCs of 0.80±0.05 and 0.89±0.04 on ITD, as well as 0.79±0.06 and 0.87±0.07 on ETD. For both PZ and TZ, there was no significant difference between ITD and ETD for the proposed method. This DL-based method enabled the accuracy of the PZ and TZ segmentation, which outperformed the state-of-art methods (Deeplab V3+, Attention U-Net, R2U-Net, USE-Net and U-Net). We observed that segmentation uncertainty peaked at the junction between PZ, TZ and AFS. Also, the overall uncertainties were highly consistent with the actual model performance between PZ and TZ at three clinically relevant locations of the prostate.

Characteristics of Macular Edema in Behcet Disease after Intravitreal Bevacizumab Injection.

PURPOSE: To investigate the effect of intravitreal bevacizumab (IVB) injection on macular edema (ME) secondary to Behcet's disease. METHODS: This prospective case series included 15 patients with bilateral ME due to Behcet's disease. Intravitreal bevacizumab was injected into the more severely involved eye; the contralateral eye was evaluated as the control. Patients were followed up with comprehensive ocular examination, optical coherence tomography, and fluorescein angiography (FA) for a minimum of 6 months by a single ophthalmologist. RESULTS: Patients with a mean age of 30.6 ± 7.4 years received a mean number of 3.3 IVB injections during the 6 months. The mean preinjection vision was 0.6 ± 0.3 and 0.4 ± 0.4 LogMAR in the case and control groups, respectively, with no significant improvement at 6 months. Mean central foveal thickness was 375.3 ± 132.1 and 307.2 ± 84.5 µm in the case and control groups, respectively, and these changed to 401 ± 199.9 (P = 0.65) and 307.7 ± 82.8 µm (P = 0.73) at month 6, respectively. A statistically nonsignificant improvement in ME was observed during the first 3 months in the case group. However, it did not persist up to month 6 on an as-needed basis. IVB injections caused a disproportionate decrease in the thickness of macular subfields. A reduction in disc leakage was observed on FA (P = 0.058). Logistic regression analysis revealed no statistically significant predictive factor for an improvement in visual acuity (VA) and a reduction in foveal thickness. CONCLUSION: During a 6-month period, IVB injections based on an as-needed protocol provided no statistically significant improvement in VA and ME.

Effect of growth hormone deficiency on brain MRI findings among children with growth restrictions.

OBJECTIVES: Growth hormone deficiency (GHD) is a major problem among children with short stature. In this study, the role of brain magnetic resonance imaging (MRI) in defining the underlying defects among short children with GHD is evaluated. METHODS: In a cross-sectional study, data of 158 children were evaluated. Growth hormone (GH) levels were measured using stimulating tests and brain MRI with gadolinium contrast was applied, as well. RESULTS: Some 25.3% of patients had GHD with a mean age of 8.01±3.40 years. MRI results showed 35 as normal, four with pituitary hypoplasia, and one with microadenoma. The MRI results were significantly associated with GH levels and presence of other endocrine disorders. There was a significant association between prenatal disorders and patients' bone age delay. CONCLUSIONS: In patients with severe GHD and patients with multiple pituitary hormone deficiencies, MRI is more likely to be abnormal, and bone age is much delayed in patients with history of prenatal disorders.