AI Creates Cancer Treatment in 30 Days and Predicts Patient Survival Rates Using Doctors’ Notes

AI has emerged as a potent tool in the fight against fatal illnesses. This has shown that it could process and interpret extensive data sets, reveal correlations and insights, and forecast treatment outcomes.

Canadian medical experts leveraged AI to devise a therapy for a specific form of cancer in 30 days. Also, a group of researchers collaborated with cancer patients to create an AI bot akin to ChatGPT. This AI bot predicts their probability of survival and life expectancy.

Different AI systems accomplished these remarkable achievements. This underscores that this technology’s application extends beyond text and image generation.

Pharma

Collaborating with Insilico Medicine, researchers from the University of Toronto utilized an AI-based drug discovery platform called Pharma. Pharma created a prospective therapy for hepatocellular carcinoma (HCC). According to Cleveland Clinic, HCC is the most prevalent form of liver cancer in which a tumor grows on the liver.

The achievement was attained in just one month after selecting the target and after synthesizing seven compounds. A subsequent round of AI-assisted compound generation identified a more potent hit molecule; nevertheless, any potential drug would need to undergo clinical trials.

AI System for Predicting Cancer Patients’ Survival Rate

Researchers from BC. Cancer and the University of British Columbia created an AI system. This AI system exhibited the ability to forecast the survival rates of cancer patients utilizing doctors’ notes. A separate research paper released in the JAMA Network Open contained information about this discovery.

The model successfully predicted survival rates for six-month, 36-month, and 60-month periods, with a precision level exceeding 80%. The previous models could only estimate survival rates for specific cancer types. This model could identify rates for all cancer types.

The system employed natural language processing (NLP) to estimate life expectancy. NLP is a field of AI that comprehends complex human language to analyze oncologists’ notes following a patient’s first consultation. The model distinguished specific attributes that are unique to each patient. The attributes are utilized to make survival predictions with an accuracy rate exceeding 80% for six-month, 36-month, and 60-month periods.

Cancer survival rates have been retrospectively determined and categorized in the past based on only a few standard factors. These factors include cancer site and tissue type. However, the model can detect unique hints within a patient’s initial consultation document, providing a more refined evaluation.

The AI system was trained and validated using data from 47,625 patients across all six BC. Cancer facilities in British Columbia.

The featured image is from hpnonline.com