Background Clinically significant prostate cancer (PCa) diagnosis at MRI requires accurate and efficient radiologic interpretation. Although artificial intelligence may assist in this task, lack of transparency has limited clinical translation. Purpose To develop an explainable artificial intelligence (XAI) model for clinically significant PCa diagnosis at biparametric MRI using Prostate Imaging Reporting and Data System (PI-RADS) features for classification justification. Materials and Methods This retrospective study included consecutive patients with histopathologic analysis-proven prostatic lesions who underwent biparametric MRI and biopsy between January 2012 and December 2017. After image annotation by two radiologists, a deep learning model was trained to detect the index lesion; classify PCa, clinically significant PCa (Gleason score ≥ 7), and benign lesions (eg, prostatitis); and justify classifications using PI-RADS features. Lesion- and patient-based performance were assessed using fivefold cross validation and areas under the receiver operating characteristic curve. Clinical feasibility was tested in a multireader study and by using the external PROSTATEx data set. Statistical evaluation of the multireader study included Mann-Whitney U and exact Fisher-Yates test. Results Overall, 1224 men (median age, 67 years; IQR, 62-73 years) had 3260 prostatic lesions (372 lesions with Gleason score of 6; 743 lesions with Gleason score of ≥ 7; 2145 benign lesions). XAI reliably detected clinically significant PCa in internal (area under the receiver operating characteristic curve, 0.89) and external test sets (area under the receiver operating characteristic curve, 0.87) with a sensitivity of 93% (95% CI: 87, 98) and an average of one false-positive finding per patient. Accuracy of the visual and textual explanations of XAI classifications was 80% (1080 of 1352), confirmed by experts. XAI-assisted readings improved the confidence (4.1 vs 3.4 on a five-point Likert scale; P = .007) of nonexperts in assessing PI-RADS 3 lesions, reducing reading time by 58 seconds (P = .009). Conclusion The explainable AI model reliably detected and classified clinically significant prostate cancer and improved the confidence and reading time of nonexperts while providing visual and textual explanations using well-established imaging features. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Chapiro in this issue.

MRI–transrectal US fusion–guided irreversible electroporation is a safe and effective procedure for patients with localized low- to intermediate-risk prostate cancer with promising midterm results,... BackgroundIrreversible electroporation (IRE) is a nonthermal ablative method based on the formation of nanoscale defects in cell membranes leading to cell death. Clinical experience with the techni...

The literature suggests implantation of the humeral component in reverse total shoulder arthroplasty (RTSA) in 0°-40° of retrotorsion without further specification. We hypothesized that optimal humeral component retrotorsion to avoid notching and gain balanced rotational capacity would depend on scapular position and posture. We investigated 200 shoulders in 100 patients with available whole-body computed tomography scans and created 3-dimensional models. Implantation of a humeral component in 20° of retrotorsion was simulated, and a correction angle (CA) to yield perfect opposition to the glenosphere was calculated. Patient-specific variables such as age, sex, posture, and scapular orientation parameters were correlated with this CA. Scapular orientation showed large interindividual differences. A highly significant correlation was seen between the CA and scapular internal rotation (R = 0.71, P < .001) and protraction (R = 0.39, P < .001). When the CA was adjusted for glenoid retroversion, the correlation coefficient of scapular internal rotation increased even further (R = 0.91, P < .001). Scapular internal rotation itself showed a correlation with thoracic kyphosis (R = 0.27, P < .001), protraction (R = 0.57, P < .001), tilt (R = 0.29, P < .001), and scapular translation (R = -0.23, P < .001). Scapular orientation and posture should be integrated into the determination process of humeral component retrotorsion in RTSA. In theory, implantation of the humeral component with increased retrotorsion leads to improved neutral opposition of the RTSA components in patients with extensive internal rotation of the scapula. On the basis of varying scapular internal rotation, we propose the distinction of 3 different posture types (A-C) for enhanced appraisal of scapulothoracic orientation.

68Gallium prostate specific membrane antigen (PSMA) ligand positron emission tomography (PET) is an increasingly used imaging modality in prostate cancer, especially in cases of tumor recurrence after curative intended therapy. Owed to the novelty of the PSMA-targeting tracers, clinical evidence on the value of PSMA-PET is moderate but rapidly increasing. State of the art imaging is pivotal for radiotherapy treatment planning as it may affect dose prescription, target delineation and use of concomitant therapy. This review summarizes the evidence on PSMA-PET imaging from a radiation oncologist’s point of view. Additionally a short survey containing twelve examples of patients and 6 additional questions was performed in seven mayor academic centers with experience in PSMA ligand imaging and the findings are reported here.

Purpose of this study was to evaluate the diagnostic performance of T1 relaxation time (T1) for differentiating prostate cancer (PCa) from benign tissue as well as high- from low-grade PCa. Twenty-three patients with suspicion for PCa were included in this prospective study. 3 T MRI including a Modified Look-Locker inversion recovery sequence was acquired. Subsequent targeted and systematic prostate biopsy served as a reference standard. T1 and apparent diffusion coefficient (ADC) value in PCa and reference regions without malignancy as well as high- and low-grade PCa were compared using the Mann-Whitney U test. The performance of T1, ADC value, and a combination of both to differentiate PCa and reference regions was assessed by receiver operating characteristic (ROC) analysis. T1 and ADC value were lower in PCa compared to reference regions in the peripheral and transition zone (p < 0.001). ROC analysis revealed high AUCs for T1 (0.92; 95%-CI, 0.87–0.98) and ADC value (0.97; 95%-CI, 0.94 to 1.0) when differentiating PCa and reference regions. A combination of T1 and ADC value yielded an even higher AUC. The difference was statistically significant comparing it to the AUC for ADC value alone (p = 0.02). No significant differences were found between high- and low-grade PCa for T1 (p = 0.31) and ADC value (p = 0.8). T1 relaxation time differs significantly between PCa and benign prostate tissue with lower T1 in PCa. It could represent an imaging biomarker for PCa.

Pancreatic neuroendocrine neoplasms (pNEN) are rare malignancies arising from neuroendocrine cells of the pancreas. Functional tumors can present with specific clinical syndromes due to hormonal secretion. These tumors can present as incidental findings on imaging performed for unrelated purposes or they are diagnosed when workup is initiated in patients with specific syndromes or metastases. This article presents an overview of available imaging techniques focusing on computed tomography and magnetic resonance imaging. Recommendations regarding examination protocols are given. Typical imaging features of pNEN and metastases are described. Their potential value for the evaluation of prognosis as well as tumor response under treatment is discussed.