AI Predicts Aggressiveness of Early Stage Breast Cancer

Womble Bond Dickinson

Womble Bond Dickinson

[author: Gloria Malpass Ph.D.]

Womble Bond Dickinson

Image for post

Photo by National Cancer Institute on Unsplash

Approximately 20% of new breast cancers are diagnosed before they have spread from milk ducts into surrounding breast tissue. Breast cancer at this early stage is known as ductal carcinoma in situ (DCIS) or stage 0 breast cancer. While some of these cancers are cured by surgery, others are aggressive, meaning they recur and spread. The ability to predict whether an early cancer is aggressive or non-aggressive would be beneficial to physicians and patients when choosing an optimal treatment plan.

Researchers at the University of Michigan Rogel Cancer Center report that they have developed a diagnostic test using artificial intelligence (AI) that has been shown to identify aggressive versus non-aggressive DCIS with more than 90% accuracy. Their findings suggest that the location of certain key enzymes near the cell membrane, rather than the abundance of those enzymes, could be used to predict whether a stage 0 breast cancer is likely to recur and spread.

A state-of-the-art computer vision application that detects patterns in images not readily discernible by the human eye was used in the collection of microscope images of DCIS samples from 70 female patients (50 non-recurrent and 20 recurrent cancers) who had been followed for at least 10 years. The images were combined with the outcomes of the patients to build a library of DCIS images representative of aggressive and non-aggressive cancers. Then, approximately 100 micrographs not previously exposed to the program were read by a computer using machine learning software. Over time, the program has been refined and is now reported to correctly differentiate aggressive and non-aggressive disease 96% of the time. According to Howard Petty, PhD, a professor of ophthalmology and visual sciences, and of microbiology and immunology at Michigan Medicine — University of Medicine, similar predictions by a human would be accurate only approximately 70% of the time.

Although further refinement and validation of this program will be needed, these early results suggest that AI may be an effective tool for predicting the recurrence and spread of other similar types of cancer. In the meantime, the University of Michigan has filed a patent application and is exploring options to bring this technology toward the market.

References: (Main news article)

Kraft AM, Petty HR. Spatial locations of certain enzymes and transporters within pre-invasive ductal epithelial cells predict human breast cancer recurrences. Am J Physiol Cell Physiol. 2020 Sep 9. doi: 10.1152/ajpcell.00280.2020. Epub ahead of print. PMID: 32903032.

[View source.]

DISCLAIMER: Because of the generality of this update, the information provided herein may not be applicable in all situations and should not be acted upon without specific legal advice based on particular situations.

© Womble Bond Dickinson | Attorney Advertising

Written by:

Womble Bond Dickinson

Womble Bond Dickinson on:

Reporters on Deadline

"My best business intelligence, in one easy email…"

Your first step to building a free, personalized, morning email brief covering pertinent authors and topics on JD Supra:
*By using the service, you signify your acceptance of JD Supra's Privacy Policy.
Custom Email Digest
- hide
- hide

This website uses cookies to improve user experience, track anonymous site usage, store authorization tokens and permit sharing on social media networks. By continuing to browse this website you accept the use of cookies. Click here to read more about how we use cookies.