New UMass Amherst AI Tool Paves the Way for Personalized Prostate Cancer Care

Researchers at the University of Massachusetts Amherst have developed a groundbreaking AI tool, potentially revolutionizing how prostate cancer treatment is personalized. Radiopharmaceutical therapy (RPT) has emerged as a promising option for combating prostate cancer, yet individualizing treatment doses remains a challenge. This new AI model aims to change that by quickly and accurately calculating personalized radiation doses for patients.

Traditionally, RPT dosing has been one-size-fits-all, despite the treatment receiving FDA approval for late-stage prostate cancer in 2022. This uniform approach may not fully exploit the therapeutic potential of RPT, as it doesn't account for individual patient needs and tolerances. "Right now, everybody gets the same dose," notes Joyita Dutta, professor in the Riccio College of Engineering at UMass Amherst, emphasizing the need for personalized treatment plans to optimize therapy outcomes.

Prostate cancer is the second most common cancer among men in the United States, following skin cancer. While early-stage diagnosis usually leads to high survival rates, the five-year survival drops to just 40% for cancer diagnosed at a later stage. The AI tool integrates advancements in dosimetry, traditionally hindered by time-consuming calculations, by speeding up the process to under 23 seconds with high accuracy.

The AI model, named DiffuDose, brings together two modules: one providing a preliminary dose estimate and another refining it into a detailed radiation dose map. "Pixel by pixel in a full image, you could see how the dose was distributed across the body," Dutta explains, highlighting the tool's precision in mapping radiation uptake across organs, crucial for adjusting treatment doses safely.

Despite significant industry interest and investment in RPT, accurately calculating personalized doses to minimize toxicities has proved difficult. Individual absorbed radiation doses enable clinicians to tailor subsequent treatments, potentially directing higher doses to target tumors effectively without harming healthy tissues.

This advancement stems from cross-disciplinary efforts, integrating expertise from UMass Chan Medical School, Massachusetts General Hospital, and the Institute of Nuclear Medicine. Noteworthy contributions also come from Dutta's graduate students, including Bowen Lei and Vibha Balaji, each recognized for their work in medical imaging and radiopharmaceutical research at prestigious conferences.

Additionally, ongoing collaborations aim to refine the AI models further by incorporating patient biomarkers, promising to enhance understanding of treatment responses. These developments illustrate UMass Amherst's commitment to transforming cancer care, linking academic innovation with clinical application.

UMass researchers have made notable strides with this AI-driven solution, poised to set new standards in individualized cancer treatment and improve outcomes in prostate cancer care. As the health care landscape continues to embrace precision medicine, such innovations highlight the promise of AI in enhancing patient-specific interventions.