September 24, 2017 |
High-tech Analysis of Genetic Data May Yield New Non-Surgical Test for Endometriosis
November 25, 2014  | 

Bethesda, MD - Using sophisticated computer-based technology to analyze genetic data obtained from uterine tissue, researchers have identified patterns of genetic activity that can be used to diagnose endometriosis, an often-painful condition that occurs when tissue that normally lines the inside of the uterus grows outside the uterus. The prototype diagnostic method can not only distinguish endometriosis from other disorders of the uterus, but can also identify the severity of the disease.

The finding is the first step toward the eventual development of a test to diagnose endometriosis that wouldn’t require surgery. Currently, a laparoscopy is the only definitive way to diagnose and stage endometriosis that occurs on the pelvic lining and organs.

A study by NIH researchers estimated that as many as 11 percent of women have endometriosis.

In their study, the researchers used a computer-based technology known as machine learning to analyze the gene activity of samples of tissue taken from the endometrium. Machine learning allows computers to learn from an activity without being explicitly programmed. Biomedical researchers rely on the technology to analyze the interactions that take when information on large numbers of genes is translated into proteins, a process called gene expression. The method enabled the researchers to develop a highly accurate grouping system to distinguish samples that came from women with endometriosis from samples that came from women with other conditions affecting the uterus and pelvis, and samples that came from normal controls. Continue>

Page | 1 2 3 4
Suggested Articles