Nucs AI Launches TrackPSMA to Provide Standardized AI Analysis for Prostate Cancer Treatment Response
San Francisco-based Nucs AI has expanded its prostate cancer platform with the launch of TrackPSMA, a tool designed to provide physicians with faster and more objective insight into how patients respond to radioligand therapy. The company notes that more than half of patients receiving this treatment may not benefit, underscoring the importance of earlier assessments to guide patients toward more effective options.
TrackPSMA, which will be showcased at RSNA 2025, uses AI to automatically compare PSMA-PET/CT scans from the same patient across different timepoints. The software quantifies disease changes and helps clinicians identify progression earlier than is often possible through traditional visual review. The goal is to replace a largely manual and inconsistent process with a standardized method that evaluates both overall disease burden and individual lesions over time.
Farid Yagubbayli, CTO of Nucs AI, said, “Being able to track disease progression lesion by lesion represents a fundamental shift in how experts understand and manage cancer over time. For the first time, we can measure therapy response objectively and reproducibly across scans, sites, and timepoints.” CEO Nijat Ahmadov added that the tool supports more precise treatment planning, noting, “By providing automated, standardized, and highly accurate treatment response assessments from PSMA-PET/CT scans, we’re able to help physicians and drug manufacturers better understand the patient-specific and tumor-specific scenarios where treatments were effective and where there may still be potential for different approaches.”
The new tool complements Nucs AI’s existing DeepPSMA and SelectPSMA solutions, which aim to streamline analysis and improve patient selection for radioligand therapy. The company positions TrackPSMA as part of a broader strategy to expand the role of AI in evidence-based imaging, developed in collaboration with leading research institutions.
