In a recent study, conducted by the researchers from the NYU School of Medicine, which published online in Nature Medicine, found a type of artificial intelligence or a machine-learning program that can differentiate between squamous cell carcinoma and adenocarcinoma (type of lung cancer) with 97% of accuracy. Even the experienced pathologists fail to detect these two lung cancers types without confirmatory tests.
With this research, a computer program that can specify cancer types, examine pictures of patients’ lung tumors, and even it can detect altered genes driving abnormal cell development. The AI tool is also capable of determining that whether unusual versions of six genes related to lung cancer such as KRAS, EGFR, and TP53 exists in the cells. Depending upon the gene, the tool is capable of determining cancerous genes up to 73% to 86% of accuracy. Such genetic mutations or changes repeatedly cause the abnormal growth seen in cancer; however, they can also change a cell’s interaction with its surroundings and its shape, giving visual clues for automated analysis.
Researchers, states that focused therapies that work specially to eliminate cancer cells with specific mutations has played a vital role in determining which genes are altered in each tumor. For example, 20% of patients suffering from adenocarcinoma are expected to experience mutations in EGFR or the gene epidermal growth factor receptor that can now be managed with help of approved drugs.
However, these days the genetic tests used to approve the presence of mutations can take weeks to return results, according to the lead author of the study.