A machine learning tool developed by researchers at Weill Cornell Medicine and the Hospital for Special Surgery (HSS) can help distinguish between subtypes of rheumatoid arthritis (RA), potentially helping scientists find ways to improve treatment for this complex disease.
This study Nature Communicationshave demonstrated that artificial intelligence and machine learning techniques can effectively and efficiently subtype pathological samples from RA patients.
“Our tool automates the analysis of pathology slides, which may lead to more accurate and efficient disease diagnosis and personalized treatment for RA in the future,” said Fei Wang, PhD, professor of population health sciences in the Department of Population Health Sciences at Weill Cornell Medical College and founding director of the AI ​​Institute for Digital Health (AIDH). “This shows that machine learning has the potential to transform pathology assessment for many diseases.”
There is some research developing machine learning tools for the automated analysis of pathology slides in oncology, and Dr. Wang and his colleagues are working to apply this technology to other clinical specialties.
Automate time-consuming processes
In the latest study, Dr. Wang collaborated with Richard Bell, PhD, lecturer in the Arthritis and Tissue Degeneration Program and Institute at HSS and computational pathology analyst in the Molecular Histopathology Core Laboratory, and Lionel Ivashkikh, PhD, chief scientific officer and chair of the Arthritis and Tissue Degeneration Program at HSS and professor of medicine at Weill Cornell Medicine, to automate the process of subtype classification of RA tissue samples. Distinguishing between the three subtypes of RA can help clinicians select the treatment that is most likely to be effective for a particular patient.
Currently, pathologists manually classify arthritis subtypes using rubrics to identify cellular and tissue features in biopsy samples from human patients, a time-consuming process that increases study costs and can lead to inconsistencies between pathologists.
“This is an analytical bottleneck in pathology studies,” Dr. Bell said. “It's very time-consuming and tedious.”
The team first trained their algorithm on RA samples from one set of mice to optimize its ability to identify tissue and cell types within the samples and classify them by subtype. They then validated the tool on a second set of samples. The tool also provided new insights into the effectiveness of treatment in the mice, including reduced cartilage degradation within six weeks of administering a commonly used RA treatment.
The researchers then applied the tool to biopsy samples from patients in the Accelerating Medicines Partnership Rheumatoid Arthritis Research Consortium, demonstrating that it can effectively and efficiently classify human clinical samples. The researchers are currently validating the tool on additional patient samples and determining how best to incorporate this new tool into pathologists' workflows.
A step towards personalized medicine
“This is a first step towards more individualized RA treatment,” Dr. Bell said. “If we can build an algorithm that identifies patient subtypes, we can get patients the treatment they need more quickly.”
The technology has the potential to provide new insights into the disease by detecting unexpected tissue changes that humans may miss. And by saving pathologists time spent on subtype classification, the tool may also reduce the cost and increase the efficiency of clinical trials testing treatments for patients with different subtypes of RA.
“By integrating pathology slides and clinical information, this tool demonstrates the expanding impact of AI in advancing personalized medicine,” said Rainu Kaushal, PhD, senior vice dean for clinical research and chair of the Department of Population Health Sciences at Weill Cornell Medical College. “This research is particularly exciting because it opens up new avenues for detection and treatment, providing a major advancement in how we understand and care for patients with rheumatoid arthritis.”
The team is working on developing similar tools to assess osteoarthritis, disc degeneration, and tendon disorders. Additionally, Dr. Wang's team is also looking to define disease subtypes from a broader range of biomedical information. For example, they recently demonstrated that machine learning can distinguish between three subtypes of Parkinson's disease.
“We hope that our study will inspire further computational research towards developing machine learning tools for more diseases,” Dr. Wang said.
“This study represents an important advance in the analysis of RA tissue that can be applied to the benefit of patients,” Dr. Ivashkiv added.
More information:
Richard D. Bell et al. “Automated Multiscale Computational Pathology Classification (AMSCP) of Inflamed Synovial Tissue” Nature Communications (2024). DOI: 10.1038/s41467-024-51012-6
Provided by: Weill Cornell Medical College
Citation: Machine learning helps identify rheumatoid arthritis subtypes (August 29, 2024) Retrieved August 29, 2024 from https://medicalxpress.com/news/2024-08-machine-rheumatoid-arthritis-subtypes.html
This document is subject to copyright. It may not be reproduced without written permission, except for fair dealing for the purposes of personal study or research. The content is provided for informational purposes only.