The National Health Service (NHS) in the UK is using artificial intelligence to detect cancer in patients. The NHS has collaborated with Intel to create a digital repository containing many thousands of tumor and immune cells.
The aim is to use algorithms to compare new tissue samples to known cancerous cells in the database to help doctors identify cancer more quickly and with greater accuracy.
Fast diagnosis of cancer means treatment can start more quickly. The sooner treatment can be started, the greater the probability of success. It is therefore hoped that by using artificial intelligence to detect cancer in patients will result in better patient outcomes, and may help to improve survival rates.
Professor Nasir Rajpoot of the University of Warwick Tissue Image Analytics (TIA) lab is heading the project, with the University Hospitals Coventry & Warwickshire NHS Trust (UHCW) and the Alan Turing Institute also collaborating.
The AI system being used for the project is based on Google’s TensorFlow, with the computing power coming from Intel Xeon processors.
The initial phase of the project will focus on creating a model of cellular distinctions associated with lung cancer. Prof. Rajpoot said one of the main aims of the project is “optimizing our digital pathology image analysis software pipeline and deploying some of the latest cutting-edge technologies developed in our lab for computer-assisted diagnosis and grading of cancer.”
Artificial Intelligence systems are not intended to replace oncologists and physicians, instead they will allow them to make diagnoses much more quickly. AI systems are able to analyze tissue samples much more quickly than humans and can compare the features of those tissues with hundreds of thousands of tissue samples in the AI database. Ultimately, AI systems will be able to perform much of the time-consuming work which will allow physicians and oncologists to use their time more efficiently.