New AI Model To Enhance Patient Response to Cancer Therapy

Australian researchers have made a significant breakthrough in cancer treatment by developing an advanced artificial intelligence (AI) tool designed to identify the most effective therapies for cancer patients.

The innovative tool, named DeepPT, was created by scientists at The Australian National University (ANU). It works by predicting a patient’s messenger RNA (mRNA) profile, which is crucial for protein production and serves as vital molecular information for personalized cancer treatment.

When used in conjunction with another tool called ENLIGHT, DeepPT has demonstrated a high success rate in predicting patient responses to various cancer therapies. This was confirmed by lead researcher Dr. Danh-Tai Hoang from ANU, who noted that DeepPT was trained using data from over 5,500 patients across 16 common cancer types, including breast, lung, head and neck, cervical, and pancreatic cancers.

The findings, published in the journal Nature Cancer, indicate that the AI tool significantly improves patient response rates. DeepPT utilizes histopathology images—microscopic pictures of patient tissue—which offer a substantial advantage by reducing the time required to process complex molecular data. This is particularly beneficial for patients with aggressive tumors who need immediate treatment.

Dr. Hoang emphasized that histopathology images are routinely available, cost-effective, and provide timely results, thereby minimizing delays that could adversely affect patient outcomes.

Historical Context

The development of AI in medical applications is part of a broader trend that began in the mid-20th century with the advent of computer technology. The use of AI in healthcare has evolved significantly, from early diagnostic systems to sophisticated tools like DeepPT that can analyze complex biological data. This progress reflects ongoing efforts to personalize medicine, a concept that gained traction in the 21st century with advancements in genomics and biotechnology.

Summary

• Australian researchers have developed an AI tool named DeepPT to improve cancer treatment selection.
• DeepPT predicts a patient’s mRNA profile, crucial for personalized cancer therapy.
• Combined with the ENLIGHT tool, DeepPT successfully predicts responses to various cancer therapies.
• The tool was trained on data from over 5,500 patients across 16 common cancer types.
• Published in Nature Cancer, the tool improves patient response rates by using histopathology images.
• Histopathology images are cost-effective, timely, and reduce delays in treatment.
• The development is part of a broader trend in personalized medicine and AI in healthcare.