We developed a privacy-preserving digital platform for patient-reported outcomes in inflammatory bowel disease using secure multi-party computation (SMPC), which encrypts and fragments data to ensure complete anonymity. In a prospective multicenter study of 322 patients, 277 respondents (86.0%) completed the survey; 60.1% reported greater comfort in providing candid responses under anonymized conditions. The system revealed substantial underreporting: only 65.1% of patients experiencing clinically important symptoms (including urgency and other key concerns) discussed them with physicians; medication adherence discrepancies occurred in 28.9% of patients, with 68.0% unrecognized by clinicians. This approach captured previously inaccessible information about sensitive symptoms and socioeconomic burdens, enabling a more accurate understanding of real-world disease impact. Our findings suggest that SMPC-based anonymization may help reveal patient-reported information that is often underrepresented in routine clinical care. (Registered with the University Hospital Medical Information Network Clinical Trials Registry: UMIN000039131; registered January 11, 2020).

For mouse brain PET imaging with high quantification accuracy, ultrahigh resolution is essential. Metabotropic glutamate receptors are important molecular targets, linked to various neurologic diseases. Here, we report the comparison of in vivo mouse brain images obtained with sub-0.5-mm resolution PET and those obtained with autoradiography using 4-18F-fluoro-N-[4-[6-(isopropylamino)pyrimidin-4-yl]-1,3-thiazol-2-yl]-N-methylbenzamide (18F-FITM) that binds to metabotropic glutamate receptor subtype 1. Methods: We recently developed a sub-0.5-mm resolution PET scanner having an axial coverage of 23.4 mm and an inner diameter of 48 mm. The PET scanner comprises 32 depth-of-interaction detectors, each of which has staggered 3-layer lutetium-yttrium-orthosilicate crystal arrays with a pitch of 0.8 mm. A resolution phantom was used to evaluate the imaging performance of the PET scanner. Dynamic PET imaging was performed to evaluate the concentration changes of 18F-FITM at various mouse brain regions. In vivo metabotropic glutamate receptor imaging of a mouse brain was performed using 18F-FITM and the PET scanner. Subsequently, autoradiography images of the same mouse brain were obtained using 18F-FITM. Results: The 0.45-mm rod structure was resolved with the PET scanner. Time-activity curves of various mouse brain regions were obtained. The sub-0.5-mm resolution PET visualized 18F-FITM uptake in the mouse brain with high correlation to the autoradiography images. Conclusion: The sub-0.5-mm resolution PET opens up new opportunities for translational neuroscience research using mice models.

PURPOSE: To evaluate the diagnostic performance of [68Ga]Ga-Trivehexin PET/CT for localizing hyperfunctioning parathyroid tissue in patients with primary hyperparathyroidism (PHPT) and to compare it with [99mTc]Tc-MIBI scintigraphy with SPECT/CT (MIBI scan). METHODS: In this retrospective study, patients with biochemically confirmed PHPT underwent both MIBI scan and [68Ga]Ga-Trivehexin PET/CT. Patient- and lesion-based detection rates were compared between modalities. [68Ga]Ga-Trivehexin PET/CT was acquired at approximately 40 ± 20 min after injection; delayed imaging at 80 ± 20 min was performed in a subset. Dynamic PET imaging was performed in three patients. Quantitative parameters, including SUVmax and lesion-to-background ratios, were recorded. Clinically relevant subgroups, including MEN-1-associated PHPT and persistent disease after surgery, were assessed. RESULTS: A total of 50 parathyroid-related lesions were evaluated in 38 patients. On patient-based analysis, MIBI scan and [68Ga]Ga-Trivehexin PET/CT correctly localized hyperfunctioning parathyroid tissue in 28 of 38 (74%) and 35 of 38 (92%) patients, respectively. On lesion-based analysis, [68Ga]Ga-Trivehexin PET/CT detected 49 of 50 lesions (98%), compared with 29 of 50 lesions (58%) detected by MIBI scan (P < 0.05). [68Ga]Ga-Trivehexin PET/CT demonstrated distinct focal uptake in 18 lesions with negative or equivocal MIBI findings. Dual-time-point imaging showed no significant differences in lesion SUVmax or lesion-to-background ratios between early and delayed scans. Dynamic imaging demonstrated a rapid increase in lesion contrast followed by a plateau, with no consistent improvement beyond approximately 25 min. CONCLUSION: [68Ga]Ga-Trivehexin PET/CT demonstrated higher detection rates than MIBI scan for localizing hyperfunctioning parathyroid tissue in PHPT, particularly in multiglandular and subcentimetric disease. Early acquisition appears sufficient, with limited added value of delayed imaging.

Internet-based cognitive behavioral therapy (iCBT) is effective for depression, but its impact is constrained by low engagement and modest response rates. Personalization may address these limitations, yet a gap remains between research evidence and clinically actionable implementation. This narrative review synthesizes evidence on personalization in iCBT for depression using a three-stage framework: pre-implementation optimization, stratifying treatment based on patient characteristics, and dynamically adapting therapy using progress monitoring. Evidence was evaluated using principles of evidence grading, with attention to the volume, consistency, and directness of findings for each stage. The strongest evidence supports pre-implementation optimization of engagement and stratification of initial support based on baseline severity, treatment history, and related clinical characteristics. Evidence for dynamic adaptation is promising but less developed, with support for early identification of nonresponse and adjustment of treatment intensity, but limited iCBT-specific trials testing adaptive treatment strategies in depression. Across stages, engagement and clinical outcomes are related but distinct targets for personalization. Emerging research on responsiveness, progress monitoring, and digital biomarkers offers future opportunities for more precise and scalable personalization. More rigorous depression-specific iCBT studies are needed to determine when, how, and for whom personalized interventions improve engagement and clinical outcomes.

Medical vision-language models (MVLMs) offer promise in clinical practice but face limitations in generalizability, data quality, and clinically meaningful evaluation. We propose RadiSim-CL, an MVLM trained via curriculum learning by simulating the three-phase pathway of a radiologist: foundational knowledge understanding, anatomical knowledge, and advanced diagnostic reasoning. To support this, we curate RadiSim, a 12-million image-text pair dataset aligned to these phases. We evaluate the model using a five-stage coarse-to-fine validation framework: (1) modality recognition, (2) anatomical recognition, (3) anatomical localization, (4) abnormality and disease diagnosis, and (5) disease differentiation and grading. This framework spans 24 zero-shot subtasks across MR, CT, and DR imaging. RadiSim-CL achieves comparable performance to state-of-the-art baselines in both foundational and anatomical tasks, and demonstrates superior capabilities in complex reasoning (e.g., an AUC of 0.953 for brain tumor diagnosis and an accuracy of 0.764 for meningioma grading). Ablation studies further confirm the curriculum's effectiveness. RadiSim-CL thus offers a scalable, clinically aligned solution to enhance diagnostic precision.

Postoperative quality-of-life (QoL) outcomes after subthalamic deep brain stimulation in Parkinson's disease vary widely. Previous studies based on PDQ-39 summary scores have reported opposing relationships between baseline and postoperative QoL, reflecting analytic variability, measurement noise, and limited feature scope. To address these inconsistencies, we analyzed 130 patients using an explainable random-forest classifier with SHAP analysis trained to predict QoL changes exceeding minimal clinically important difference thresholds. Baseline variables included PDQ-39 subscores, along with demographic, motor, cognitive, and affective measures, and electrode coordinates derived from imaging. Predictors of QoL improvement included younger age, greater preoperative disease-related emotional burden, and electrode placement at the motor-associative transition in the right subthalamic nucleus. The model achieved an area under the curve of 0.70 on the held-out test set, with balanced sensitivity and specificity. Identifying interpretable cut-offs for age, emotional burden and electrode location supports individualized counseling and treatment planning, advancing outcome prediction in neuromodulation.