A multimodal AI algorithm accurately predicts which mCRPC patients will respond to Pluvicto radioligand therapy using PSMA-PET features and laboratory values.
Key Details
- 1Researchers combined PSMA-PET imaging features with 10 core lab values from EHRs to develop the predictive model.
- 2The retrospective study analyzed 302 patients with metastatic castration-resistant prostate cancer (mCRPC) treated with at least 3 cycles of Pluvicto.
- 3Imaging features included lesion shape, standardized uptake values, and tumor uptake heterogeneity; labs included PSA and lymphocyte count.
- 4The combined model achieved an AUC of 0.80, outperforming clinical-only (0.73) and imaging-only (0.76) models.
- 5Responders identified by the model had a 30-month survival of 79% compared to 34% for nonresponders.
Why It Matters
This approach supports the use of imaging AI for personalized treatment planning in prostate cancer patients, potentially improving outcomes by identifying likely responders to PSMA-targeted radioligand therapy.
Source
AuntMinnie
Related News
•AuntMinnie
AI Boosts Sensitivity but Raises False Alarms in Microbleed Detection
AI significantly increases sensitivity for detecting cerebral microbleeds on brain MRI but at the cost of more false positives, highlighting the complementary strengths of AI and radiologists.
•AuntMinnie
Deep Learning on PSMA-PET Outperforms Tools in Prostate Cancer Survival Prediction
Deep learning models using PSMA-PET scans predict survival in metastatic prostate cancer more accurately than established clinical tools.
•AuntMinnie
AI on Chest X-Rays Matches PREVENT Score for Cardiac Risk
Deep learning models analyzing chest x-rays can predict cardiovascular risk on par with the PREVENT clinical score.