Objectives: The study aims to evaluate Prostate Imaging-Reporting and Data System 4 (PI-RADS 4) lesions, including those with primary diffusion restriction (diffusion-weighted imaging score 4) and those upgraded from PI-RADS 3 due to positive dynamic contrast-enhanced findings, and to compare outcomes between these groups while also assessing cancer detection rates and prostatitis between upgraded cases and PI-RADS 3 lesions. Methods: In this single-center cohort study, peripheral zone lesions classified as PI-RADS 3 and 4 by multiparametric MRI and biopsied via MRI/transrectal ultrasound-guided targeted biopsy were analyzed. Results: Of 420 peripheral zone lesions from 272 patients, 202 were PI-RADS 3, 81 PI-RADS 3 + 1, and 137 PI-RADS 4. Cancer detection rates significantly differed between PI-RADS 3 + 1 and PI-RADS 4 (38.3% vs 73% for overall cancer; 6.2% vs 30.7% for clinically significant cancer, P < .001). PI-RADS 3 + 1 lesions had significantly higher cancer detection rates than PI-RADS 3 lesions (38.3% vs 19.8%, P = .001). Prostatitis was significantly more prevalent in PI-RADS 3 (34.7%) and PI-RADS 3 + 1 (29.6%) than in PI-RADS 4 (10.9%) lesions (P < .001). Conclusions: PI-RADS 3 + 1 lesions show significantly different cancer detection rates from both PI-RADS 3 and 4, suggesting they should be managed as a distinct entity. Higher prevalence of prostatitis in PI-RADS 3 + 1 lesions indicates that clinical evaluation for inflammatory conditions may help reduce unnecessary biopsies in this subgroup. Advances in knowledge: PI-RADS 3 + 1 lesions show higher cancer detection than PI-RADS 3 but lower than PI-RADS 4, and are more frequently associated with prostatitis than PI-RADS 4. These findings support individualized management to avoid unnecessary biopsies.

Background/Objectives: Large language models (LLMs) are increasingly used as decision-support tools in radiology, yet their influence on clinical decision-making remains a significant concern. This study evaluates how artificial intelligence (AI)-generated malignancy risk estimates influence clinical decisions in breast imaging and assesses automation bias (AB) among radiology residents at different training levels. Methods: In this simulation-based study, 20 residents reviewed radiology reports for 15 synthetic breast imaging cases generated by ChatGPT-4 Turbo, each presented in either a concordant or discordant format. Participants were stratified into two groups based on their completion of a breast imaging rotation and evaluated each case by assigning a BI-RADS score, recommending management, and rating their confidence in the AI risk estimate. Diagnostic accuracy and independent predictors of AB were analyzed using Generalized Estimating Equations (GEE) to account for reader-level clustering. Results: Overall management accuracy was significantly lower in cases with discordant AI outputs (p = 0.008). A similar, albeit non-significant, decreasing trend was observed for BI-RADS accuracy. While advanced residents maintained high management accuracy regardless of AI concordance, early-stage residents showed a sharp decline in discordant cases (77.3% vs. 45.3%; p = 0.006). This decline was most critical in malignant cases for early-stage residents, where accuracy dropped from 92.0% to 48.0% (p < 0.001). In the multivariable GEE logistic regression model, training level was the only independent predictor of AB (OR: 0.09, 95% CI: 0.03–0.28, p < 0.001). Conclusions: In this simulation study, advanced clinical training was associated with lower susceptibility to AB in breast imaging, although training level alone is unlikely to be a sufficient safeguard. Safe integration of AI tools will require caution among junior trainees alongside system-level measures such as model transparency and institutional oversight.

Background/Objectives: The prognostic significance of young age in breast cancer (BC) remains controversial, with varying findings across studies. This study aimed to evaluate the imaging characteristics, pathological features, and survival outcomes of BC in young patients aged ≤ 35 years by dividing them into two groups and to compare outcomes between these subgroups to determine whether very young age alone is associated with worse prognosis. Methods: This retrospective study included patients aged ≤ 35 years with malignant breast lesions who underwent image-guided interventions at a single center between January 2011 and December 2024. Patients were divided into two groups: Group 1 (≤30 years) and Group 2 (31–35 years). Demographic, imaging, and pathological data, as well as survival outcomes, were analyzed. Results: A total of 34 patients (Group 1: n = 14, Group 2: n = 20) with malignant breast lesions underwent 40 image-guided procedures. No significant differences were observed between groups in imaging features, pathological characteristics, or pTNM stage distribution. Median follow-up was 68 months. One-year overall survival (OS) was 100% in both cohorts; 5-year OS rates were 100% and 92.3%, respectively, with no statistically significant difference (p = 0.784). Median event-free survival (EFS) was 60.5 months, also without a significant difference between groups (p = 0.897). None of the clinicopathological factors assessed were significantly associated with OS in this exploratory analysis. Conclusions: In women ≤ 35 years, tumor characteristics and patient outcomes were similar between those aged ≤ 30 and 31–35 years. OS and EFS were favorable in both groups, indicating that within the ≤35-year population studied, being ≤30 years old was not associated with worse outcomes than being 31–35 years old. Although advanced stage disease was more frequently observed in the younger group, this difference was not statistically significant.