Nikhil Prasad Fact checked by:Thailand Medical News Team May 19, 2024 5 months, 3 weeks, 5 days, 7 hours, 22 minutes ago
AI In Medicine: In a groundbreaking advancement for medical diagnostics, Australian scientists have developed an innovative artificial intelligence (AI) platform designed to significantly improve the accuracy and speed of classifying central nervous system (CNS) tumors. This
AI In Medicine development, spearheaded by researchers from The Australian National University (ANU), promises to revolutionize the way brain tumors are diagnosed and treated.
Revolutionary AI Tool Enhances Brain Tumor Classification
The Need for Precision in Tumor Classification
Accurate classification of brain tumors is critical for effective patient treatment. Misclassification or delays in diagnosis can lead to inappropriate treatment plans, adversely affecting patient outcomes. Traditionally, the gold standard for identifying different types of brain tumors has been DNA methylation-based profiling. DNA methylation involves the addition of methyl groups to the DNA molecule, which can switch genes on or off, influencing gene activity.
According to Dr Danh-Tai Hoang, a key researcher in this project, while DNA methylation profiling is highly accurate, it has significant drawbacks. "The time it takes to do this kind of testing can be a major drawback, often requiring several weeks or more when patients might be relying on quick decisions on therapies," he explained. Moreover, the availability of such tests is limited, with most hospitals worldwide lacking the necessary facilities.
Introducing DEPLOY: A Breakthrough in AI-Based Diagnostics
To address these challenges, the ANU researchers, in collaboration with experts from the National Cancer Institute in the United States, developed a novel AI platform named DEPLOY (Deep lEarning from histoPathoLOgy and methYlation). This tool leverages deep learning techniques to predict DNA methylation and classify brain tumors into ten major subtypes, significantly reducing the time required for accurate diagnosis.
DEPLOY utilizes histopathology images - microscopic pictures of a patient’s tissue - trained on large datasets comprising approximately 4,000 patients from the U.S. and Europe.
Unprecedented Accuracy and Clinical Relevance
The results achieved by DEPLOY are remarkable. Dr Hoang highlighted that "DEPLOY achieved an unprecedented accuracy of 95%." This high level of precision was particularly evident when DEPLOY was tested on a subset of 309 challenging samples. In these cases, DEPLOY provided diagnoses that were more clinically relevant than those initially provided by pathologists, underscoring its potential as a complementary diagnostic tool.
"DEPLOY shows the potential future role of AI in enhancing pathologists' initial diagnoses, or even prompting re-evaluation in the case of disparities," Dr Hoang added. The implications of such a tool are profound, potentially extending beyond CNS tumors to aid in the classification of other cancer types as well.
;
How DEPLOY Works
DEPLOY's innovative approach integrates three distinct components:
-Direct Model: Classifies CNS tumors directly from slide images.
-Indirect Model: Generates predictions for DNA methylation beta values from histopathology images, which are subsequently used for tumor classification.
-Patient Demographics Model: Classifies tumor types directly from routinely available patient demographics.
First, DEPLOY accurately predicts beta values from histopathology images. Then, using a ten-class model trained on an internal dataset of 1,796 patients, it predicts tumor categories in three independent external test datasets, including 2,156 patients. The overall accuracy of 95% and balanced accuracy of 91% on high-confidence samples showcases DEPLOY's potential to assist pathologists in diagnosing CNS tumors within a clinically relevant short timeframe.
Future Prospects and Broader Applications
The successful development and implementation of DEPLOY marks a significant step forward in the field of medical diagnostics. The researchers believe that DEPLOY's capabilities could be expanded to classify other types of cancer, thereby broadening its impact on oncology and patient care. The potential for rapid, accurate, and accessible diagnostics could transform how cancer is diagnosed and treated worldwide.
Conclusion
The development of DEPLOY by ANU researchers represents a monumental achievement in the use of AI for medical diagnostics. By significantly improving the speed and accuracy of brain tumor classification, DEPLOY holds promise for enhancing patient outcomes and advancing the field of oncology. As this technology continues to evolve, it may become an indispensable tool for pathologists, aiding in the precise and timely diagnosis of various cancer types.
The study detailing DEPLOY's development and validation was published in the peer reviewed journal Nature Medicine.
https://www.nature.com/articles/s41591-024-02995-8
For the latest on
AI In Medicine, keep on logging to Thailand Medical News
Share
Tweet
Share
