Researchers at The Institute of Cancer Research (ICR) in London have developed an artificial intelligence (AI) tool named MorphoMIL, designed to expedite the drug discovery process for cancer treatments. Traditionally, developing a single drug can take between 10 to 12 years. MorphoMIL utilizes deep learning to analyze the three-dimensional shapes of drug-treated cancer cells, providing insights into how these cells respond to various treatments. By focusing on 3D cell structures, this approach offers a more accurate representation of cellular behavior within the human body compared to previous 2D analyses.
In training sessions involving 95,000 melanoma cells treated with different drugs, MorphoMIL achieved a prediction accuracy of 99.3%. This high level of precision enables researchers to identify potential drug targets more efficiently. The ICR anticipates that MorphoMIL could reduce the preclinical research phase from three years to three months and potentially shorten the overall drug trial process by up to six years. Professor Chris Bakal, who led the development, emphasized that this tool could significantly streamline drug discovery, ultimately delivering new treatment options to cancer patients more rapidly. Click for More Details
