Scientists at Harvard Medical School have unveiled an innovative AI tool that can rapidly decode a brain tumor's DNA during surgery.
We learn in a report shared by ScienceDaily that this technological breakthrough provides neurosurgeons with critical information about the tumor's molecular identity within minutes, which previously took days or weeks.
The tool, known as CHARM (Cryosection Histopathology Assessment and Review Machine), empowers surgeons to make informed decisions regarding the extent of brain tissue removal and the administration of tumor-killing drugs directly into the brain, all while the patient is still on the operating table.
Importance of Molecular Diagnosis
Accurate molecular diagnosis during surgery is of utmost importance in neurosurgery.
It allows surgeons to strike a delicate balance, ensuring the removal of enough brain tissue in aggressive tumors while preserving vital functions in less aggressive cases.
The traditional approach of freezing brain tissue and examining it under a microscope often leads to altered cell appearance, compromising clinical evaluation accuracy.
Additionally, subtle genomic variations can be challenging for human pathologists to detect.
However, CHARM's AI-driven approach overcomes these limitations, revolutionizing how brain tumor molecular diagnosis is performed.
Read Also: AI-Powered Tool Shows Promise in Detecting Dementia and Alzheimer's
Real-Time Precision Oncology
The ability to determine a tumor's molecular identity in realtime during surgery is a game-changer for precision oncology.
Dr. Kun-Hsing Yu, the study's senior author and an assistant professor of biomedical informatics at HMS, highlighted the potential of this advancement.
He claims that even state-of-the-art clinical practice cannot profile tumors molecularly during surgery. The tool overcomes this limitation by retrieving previously untapped biomedical signals from frozen pathology slides.
By identifying a tumor's molecular type during surgery, doctors can tailor treatments to the individual patient, including administering drug-coated wafers placed directly into the brain.
This improves patient outcomes and paves the way for real-time precision oncology, where treatments are dynamically adjusted based on molecular insights.
Training and Validation of CHARM
The development of the CHARM tool involved training it on a comprehensive dataset of 2,334 brain tumor samples from 1,524 glioma patients.
Remarkably, the AI model achieved an accuracy of 93% in distinguishing tumors with specific molecular mutations.
It successfully classified major gliomas with distinct molecular features, offering insights into their growth patterns, spread, and treatment responses. Gliomas, the most common form of brain cancer, are known for their molecular complexity and variations in cell appearance.
Furthermore, the tool captured crucial visual characteristics of malignant cells' surrounding tissue, signaling more aggressive glioma types.
What's Next?
Although CHARM holds immense promise, it is essential to undergo clinical validation and obtain clearance from the FDA before deploying it in hospitals.
However, the potential impact of this technology on glioma diagnosis and treatment is significant. By incorporating molecular information and providing real-time clinical decision support, CHARM improves diagnostic accuracy and accessibility.
It aligns with the World Health Organization's updated classification system for brain tumors, which emphasizes genomic profiling in diagnoses.
Stay posted here at Tech Times.
Related Article: First-Fully AI-Generated Drug to Reach Human Clinical Trials This Week
