Scientists from China and Spain have developed a system based on artificial intelligence that can distinguish cancerous cells from healthy ones, according to *Natural* magazine.
The magazine reported that the new system can also detect viral infections within cells at a very early stage.
The researchers behind the system explained that the machine learning algorithm they developed, when used with high-resolution microscopy, is a highly precise tool for detecting cellular heterogeneity, which allows for the distinction between cancerous cells and other cells.
The magazine noted that this technique has some limitations that must be overcome before clinical trials and testing can begin.
According to the magazine, academician Alexander Gintsburg, General Director of the Gamaleya Research Center for Epidemiology and Microbiology in Russia, previously announced an anti-cancer vaccine that artificial intelligence will help develop, tailored for specific individuals according to their condition and body characteristics.
It is worth mentioning that Mayo Clinic in the United States announced last year the development of an AI-based model to detect cancer.
The clinic stated that researchers at the Mayo Clinic Comprehensive Cancer Center used the most comprehensive imaging dataset in the world to build a multi-capability AI model that demonstrated the potential for independent detection of pancreatic cancer using standard computed tomography (CT) imaging at a stage where surgical intervention offers promising treatment.
Ajit H. Goenka, a Doctor of Medicine and radiologist at Mayo Clinic, as well as the principal investigator and author of the study, mentioned that "this is where the study emerges as a beacon of hope."
He added, "It addresses the challenge of the final step, which is detecting cancer at a stage beyond the expertise of specialists."
The group developed the AI model using data from more than three thousand patients.
According to the magazine, the most significant aspect of the model is its ability to detect visually undetectable pancreatic cancer from a normal-appearing pancreas on CT scans, well before clinical diagnosis, and at a very early stage, with an average of 438 days before clinical diagnosis.
Dr. Goenka explained: "These findings suggest that AI has the potential to detect hidden cancers in individuals who show no symptoms, allowing for surgical treatment at a stage where cure is still possible."
0 Comments