EndoRX: Revolutionizing Early Endometriosis Diagnosis with Machine Learning

University of Pittsburgh researchers have developed EndoDx, a groundbreaking diagnostic tool that leverages machine learning to diagnose endometriosis (EM) early and non-invasively. Endometriosis affects approximately 10% of women worldwide, often leading to chronic pain and infertility. Traditional diagnosis requires invasive surgery, but EndoDx uses patient-specific data and biomarkers to provide a precise diagnosis, potentially reducing the average 6.7-year delay in diagnosis and alleviating the clinical burden.

Description

EndoDx is a software tool that employs a robust machine learning model to analyze biomarkers from menstrual blood, peripheral blood, urine, and saliva, along with patient clinical and demographic data. This model predicts the presence and severity of endometriosis, offering a non-invasive alternative to surgical diagnosis. The technology aims to automate trend recognition for various biomarkers, providing a personalized and accurate diagnosis.

Applications

• Diagnostic tool for endometriosis
• Women’s health research
• Clinical decision support
• Personalized medicine

Advantages

EndoDx offers a non-invasive diagnosis, eliminating the need for exploratory laparoscopic surgery and reducing patient risk and healthcare costs. It significantly decreases the time-to-diagnosis, improving patient outcomes. By utilizing patient-specific data, EndoDx provides an accurate and tailored diagnosis, making it a cost-effective solution that reduces the financial burden on patients and healthcare systems by avoiding unnecessary surgeries.

Invention Readiness

EndoDx is currently in the concept stage, with ongoing development and validation of the machine learning model. Experimental work includes the collection and analysis of biomarkers from menstrual blood, peripheral blood, urine, and saliva samples. Preliminary data suggests that the model can accurately predict the presence and severity of endometriosis. Further experiments are focused on refining the algorithm, expanding the dataset, and conducting clinical trials to validate the tool’s efficacy in a real-world setting.

IP Status

https://patents.google.com/patent/WO2025229620A1