Artificial intelligence (AI) is a valuable support in multiple areas of health. New Tool May Allow "Near-Real-Time" Analysis of Brain Tumor During Patient Operation, Helping Surgeons Operate "Safer and More Accurately", Study Finds published this Monday in the journal Nature Medicine .
AI, tested by a team of American researchers can tell in less than 2 and a half minutes if the cells removed are cancerous or not, compared to an average of 20 to 30 minutes for a conventional analysis.
Over 2.5 million images analyzed
For their work, the researchers combined an innovative laser imaging technique and an algorithm driven by the analysis of more than 2.5 million biopsy images. With the tool, "we are better equipped to keep healthy tissue and remove only tissue infiltrated by cancer cells, which results in fewer complications and better results for cancer patients", explains Daniel Orringer, from New York University.
"In neurosurgery and in many other areas of cancer surgery, detecting and diagnosing tumors during the operation is essential for performing the most appropriate surgical procedure," said the neurosurgeon.
Establish a correct diagnosis
From the sample taken, the program also allows us to tell what type of tumor it is, among the 10 most common types of brain cancer. And this, with an efficiency comparable to that of pathologists. In a clinical trial including 278 patients with a brain tumor, artificial intelligence established a correct diagnosis in 94.6% of cases, compared to 93.9% for human analysis.
"Surprisingly, in all cases where the pathologists got it wrong, our algorithm was right, and in all cases where the algorithm got it wrong, the pathologists got it right," observes Dr. Orringer. According to the authors, such a tool could improve the relevance of human analysis and overcome the shortage of experts in places where there are not enough of them.
Health
Breast cancer: Artificial intelligence more precise than humans in making a diagnosis
Economy
CES Las Vegas 2020: A connected medical chair to make life easier for doctors
- Brain
- Health
- Cancer
- Tumor