MDDI Online is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Progenics Collaborates with Veterans Affairs on AI Research Program for Medical Image Analysis

The collaboration is first in the nation to validate machine-learning tools with the aim of improving treatment management of veterans with prostate cancer.

EXINI Diagnostics, recently acquired by Progenics, specializes in medical imaging AI and the development of validated imaging biomarkers and medical devices. Its main focus is to support Progenics’ pipeline of prostate-specific membrane antigen (PSMA)–imaging radiopharmaceuticals

MD+DI/Amanda Pedersen

A close collaboration between the Veterans Affairs Greater Los Angeles Healthcare System (VAGLAHS) and Progenics Pharmaceuticals will focus on the development and validation of artificial intelligence (AI) algorithms in medical images for disease management for veterans with prostate cancer.

“The promise of AI in healthcare, specifically in this collaboration, is to increase the quality of care while driving down costs,” said Aseem Anand, PhD, vice president and site manager of EXINI Diagnostics, in an interview with MD+DI.

EXINI Diagnostics, recently acquired by Progenics, specializes in medical imaging AI and the development of validated imaging biomarkers and medical devices. Its main focus is to support Progenics’ pipeline of prostate-specific membrane antigen (PSMA)–imaging radiopharmaceuticals, Anand said.

In the initial phases of the collaboration, the VAGLAHS network will gain access to Progenics’s machine-learning platforms—the automated Bone Scan Index (aBSI) technology and PSMA platform.

The FDA-cleared aBSI technology gives a fully quantitative assessment of bone scans that incorporates inferred masses of all lesions. Anand explained that in a prospectively defined, multi-institutional, phase-three study with 721 metastatic prostate cancer patients, aBSI was found to be an independent prognostic determinant of overall survival. The study also supported aBSI in the design and eligibility for clinical trials for systemic therapies for metastatic castration-resistant prostate cancer.

PSMA AI is the first software tailored to Progenics’s PSMA-targeted imaging agents. “The imaging agent attaches to PSMA, which is overexpressed on the surface of prostate cancer cells,” said Anand. “The agent is labeled with a radioactive molecule detectable by nuclear medicine cameras, and used in conjunction with co-registered anatomical images to provide anatomical context to the functional images.”

PSMA AI automatically and quantitatively computes assessments used to detect, localize, quantify, and stage both localized and advanced prostate cancer lesions. The technology is based on deep learning and convolutional neural networks that determine a detailed anatomical context from the anatomical images. This information is combined with information from the corresponding PSMA-targeted functional image to assess cancer prevalence and growth. The performance of assessments in hybrid images on patients with prostate cancer by human experts have been statistically validated to be improved when provided with the additional information retrieved with PSMA AI, said Anand.

“PSMA imaging is an emerging technology for the diagnosis of prostate cancer, and the field has yet to establish a standardized, quantitative assessment and reporting structure with high repeatability to maximize its potential,” said Anand, noting that this is why collaboration between Progenics and the VAGLAHS is especially interesting.

The company’s algorithms will focus on parts of the workflow that are tedious, qualitative, or subject to oversight. “With the help of AI, we can measure imaging data characteristics that are invisible to the naked eye, or impractical to measure reliably in clinical routine,” said Anand. “By automating these steps and presenting results to the user in an intuitive manner, we are increasing the quality of the examination.”

In the early disease state, the focus of the algorithms would be to validate the detection and staging of veterans with a risk of prostate cancer, Anand noted. He went on to say that in later disease states, “we would explore the utility of our algorithm in monitoring and prediction of treatment response to avoid ineffective treatment options.”

“The progress made here has the potential to impact the large number of veterans with prostate cancer treated at the VA centers of care across the nation,” Anand said.

Susan Shepard

Susan Shepard

Susan Shepard is a freelance contributor to MD + DI.

Filed Under
500 characters remaining