AI Mapping Tech Outperforms MRI in Predicting Prostate Cancer Spread

Avenda Health is reporting new study results that indicate its cancer mapping tool, Unfold AI, predicts cancer spread more accurately than MRI. 

The pilot study, “Prediction of Seminal Vesicle Invasion Using Artificial Intelligence Prostate Cancer Risk Mapping,” show that Unfold AI significantly improves the prediction of seminal vesicle invasion (SVI), a critical factor in prostate cancer staging and prognosis, achieving a 92% accuracy rate, compared to 52% with radiologist interpretation on a standard MRI (magnetic resonance imaging). This data lays the groundwork for further research into the diagnostic accuracy of artificial intelligence (AI) and its role in improving patient outcomes. 

The data builds on a previous study that examined Unfold AI’s ability to predict extracapsular extension risk and was named winner of the BJUI Compass Prize 2025 at the American Urological Association’s Annual Meeting earlier this month. Together, these trials demonstrate Unfold AI’s role in predicting prostate cancer spread to other organs, improving physicians’ ability to accurately diagnose and plan treatment for patients. 

Determining whether prostate cancer has spread into other nearby structures, like the seminal vesicles, is critical for accurate staging and effective treatment planning, particularly for surgery and radiation. Traditional methods like MRI are largely unreliable and inaccurate, which leads to frequent misdiagnosis of SVI with MRI alone. 

Researchers from Stanford University School of Medicine and UCLA’s David Geffen School of Medicine conducted a preliminary study of two cohorts of men, all of whom received MRI scans before undergoing prostate cancer surgery. They compared physicians’ predictions of SVI based on MRI alone to predictions made by Unfold AI, which combines MRI and clinical data to create a 3D cancer map. After surgery, the team analyzed prostate specimens for SVI to determine ground-truth and evaluate the accuracy of both methods.

Related: Study: AI Improves Physicians’ Ability to Assess Extent of Prostate Cancer

In the first cohort, pathology confirmed SVI in 25 of 147 patients. Unfold AI accurately identified 92% of these cases, while physicians using MRI identified 52%. The second cohort included 20 patients, 10 of which had SVI. MRI missed four of the 10 cases, while Unfold AI missed two. Unfold AI also produced fewer false positive rates compared to MRI in both the first (AUC 0.95 vs. 0.80, p = 0.003) and second cohorts (AUC 0.85 vs 0.74).

“These results demonstrate how Unfold AI continues to improve the diagnosis and staging of prostate cancer, enabling the physician to recommend and deliver the best therapy to the patient,” study author and Avenda Health Co-Founder/CEO Shyam Natarajan, Ph.D., said. “We are honored to be at the forefront of providing AI tools to urologists that improve the lives of patients.”

The project was funded in part by the National Cancer Institute at the National Institutes of Health.

Using deep learning, Avenda Health software maps a patient’s cancer in 3D, giving physicians precise information and the confidence to make care decisions. Dedicated to bringing this technology to all practices and physicians treating prostate cancer patients, Avenda Health aims to improve outcomes, preserve quality of life, and create a new standard of cancer care.