Mass General Brigham researchers have developed an AI model capable of identifying and measuring aggressive prostate cancer lesions.
Image Credit: Antonio Marca/Shutterstock.com
Prostate cancer is the second most common cancer among men, with nearly 300,000 new cases each year in the US alone. To establish a consistent method for estimating prostate cancer size, aiding clinicians in making more accurate treatment decisions, researchers at Mass General Brigham trained and validated an AI model using MRI scans from over 700 prostate cancer patients.
The model successfully identified and marked the boundaries of 85 % of the most radiologically aggressive prostate lesions. Tumors with larger volumes, as estimated by the AI model, correlated with a higher risk of treatment failure and metastasis, independent of traditional risk factors.
Notably, for patients who underwent radiation therapy, tumor volume proved to be a stronger predictor of metastasis than conventional risk stratification methods. Researchers believe this tool could help clinicians gauge tumor aggressiveness, personalize treatment plans, and refine radiation therapy.
Al-determined tumor volume has the potential to advance precision medicine for patients with prostate cancer by improving our ability to understand the aggressiveness of a patient's cancer and therefore recommend the most optimal treatment.
David D. Yang, MD, Study First Author and Founding Member, Department of Radiation Oncology, Mass General Brigham
MRI has significantly improved clinicians' ability to diagnose prostate cancer and is now routinely used in diagnosis and treatment. While clinicians can estimate tumor size from MRI images, these assessments are somewhat subjective and can vary between individuals.
To establish a more consistent approach, the researchers trained an AI model on MRI images from 732 prostate cancer patients treated at a single center. They then examined whether the model’s size estimates were linked to treatment outcomes 5 to 10 years after diagnosis.
As a result, the AI model was able to accurately identify and measure around 85 % of prostate tumors with a PI-RADS 5 score, indicating a high risk of clinically significant cancer within the study group. Additionally, the model’s size estimates showed promise as a prognostic tool: larger tumors correlated with a higher likelihood of prostate cancer recurrence, indicated by prostate-specific antigen (PSA) levels or metastasis in patients treated surgically or with radiation therapy.
The AI measurement itself can tell us something additional in terms of patient outcomes. For patients, this can really tell them something about what are the chances of cure, and the likelihood that their cancer will reoccur or metastasize in the future.
Martin King, MD, Ph.D., Study Senior Author, Department of Radiation Oncology, Mass General Brigham
Beyond helping clinicians and patients understand the aggressiveness of prostate cancer, the AI model could also support radiation oncologists by precisely identifying the tumor’s focal area for more targeted treatment. It offers a significantly faster alternative to current methods for assessing prostate cancer aggressiveness, which often require two weeks or more, potentially allowing patients to begin treatment sooner.
Cancer research is a cornerstone of care at Mass General Brigham. By combining research with strengths in innovation, education, and community engagement, Mass General Brigham Cancer delivers integrated, equitable cancer care. Their vision is to provide a comprehensive, research-driven approach that supports patients throughout their care journey—from prevention and early detection to treatment and survivorship.
Moving forward, the researchers plan to validate their model on a larger, multi-institutional dataset.
“We want to validate our findings, using other institutions and patient cohorts with different disease characteristics, to make sure that this approach is generalizable to all patients,” said Yang.