Researchers at The Australian National University (ANU) have developed a new artificial intelligence (AI) tool to help physicians choose the most effective treatment for cancer patients. Named DeepPT, this AI model was created in collaboration with scientists at the National Cancer Institute in America and pharmaceutical company Pangea Biomed. It predicts a patient’s messenger RNA (mRNA) profile, which is essential for protein production and personalized cancer medicine.
Key Insights from the Study
– **Publication**: The research appears in *Nature Cancer*. – **Collaboration**: ANU, National Cancer Institute, Pangea Biomed. – **Prediction Capabilities**: Deep PT predicts mRNA profiles, key for personalized cancer treatment.
Improved Patient Outcomes
Lead author Dr. Danh-Tai Hoang from ANU highlighted that, when used with a second tool called ENLIGHT, Deep PT successfully predicted patient responses to various cancer therapies.
– **Training**: Deep PT was trained on data from over 5,500 patients across 16 common cancer types, including breast, lung, head and neck, cervical, and pancreatic cancers. – **Results**: The patient response rate improved from 33.3% without using the model to 46.5% with the model.
Previous Work and Additional Benefits
Deep PT builds on prior work by the same ANU researchers who developed a tool to classify brain tumors. Both AI tools utilize histopathology images—microscopic pictures of patient tissue—which offer significant benefits:
– **Reduced Delays**: Cuts down on the processing time for complex molecular data, which can take weeks. – **Accessibility**: Histopathology images are routinely available, cost-effective, and timely.
Dr. Hoang emphasized the importance of this advancement, especially for patients with high-grade tumors requiring immediate treatment.
