Treatment of Prostate Cancer: Human-AI Interaction in Detection and Diagnosis: Ways to Team Up

0 Visioni

  • Grazie! Condividilo con i tuoi amici!

    Non hai gradito questo video. Grazie per il feedback!

admin
Published
Evis Sala, MD, PhD, FRCR, FRCP, Professor of Radiology, Università Cattolica del Sacro Cuore, Rome, Italy, outlines principles for human-AI interaction in prostate cancer detection and diagnosis. She emphasizes that AI must provide value by maintaining high sensitivity and negative predictive value, reducing false positives, and improving detection of clinically significant cancers while preserving radiologist productivity.

Dr. Sala describes various AI functions, including decision support, second reader applications, workflow or patient pathway triage, and high-confidence filtering. She notes that AI should not replace radiologists but should assist by standardizing interpretation, highlighting suspicious lesions, and streamlining reporting. Effective integration requires tools that generate explainable, exportable, and actionable outputs.

Examples include combining AI-based lesion maps with prostate-specific antigen (PSA) density and Prostate Imaging-Reporting and Data System (PI-RADS) scoring to increase specificity and reduce unnecessary biopsies. She reviews studies demonstrating that AI performance is non-inferior to that of radiologists and that integrating AI with clinical data further improves accuracy. AI-assisted diagnostic workflows can reduce reporting time, prioritize cases, and provide patient-friendly visual outputs for counseling.

Dr. Sala emphasizes triage on the clinical pathway as a significant area of impact, where AI can shorten diagnostic timelines by accelerating high-risk cases to biopsy while allowing lower-risk cases to be reviewed later. She asserts that successful human–AI teaming requires prospective validation, clear benchmarking, and seamless integration into multidisciplinary workflows.

Don't forget to join the GRU Community: https://grandroundsinurology.com/register/
Follow us on Twitter/X: https://x.com/GRUrology
And like and subscribe to us here on YouTube!
Categoria
Urology
Commenta per primo questo video.

Prossimo