To most people artificial intelligence is related to automatic cars and robots at manufacturing units but now a group of researchers are using this technology to improve screening facilities of lung cancer to enable early detection and treatment that can improve survival rates of patients. But current screening methods for this type of cancer have very high false positive rates so a new method has been developed to identify actual cases and reduce false alarms. Current screening method of CT scans throw up nodules in the lung leading to anxiety and increased health costs though it may turn out in the long run that only 4% of them actually have cancer.
Co-author of the study Panayiotis Benos who is also VC of computational and systems biology at Pittsburgh University stated that under most circumstances the follow-up us too expensive and not risk free either. The new method will help doctors identify which nodules on lungs are benign and which are risky to help as around 96% have benign nodules. To confirm their theory Benos and his team used medical data of 218 high risk people into an AI based program for creating a model that can calculate cancer probability.
After this researchers compared results of their model against actual diagnosis of all these patients. The team stated that this model has the capability to spare 30% of benign lung nodule patients from undergoing any more unnecessary tests and did not miss a single actual cancer cases. As they were able to rule out cancer in one third patients these people had to just come back after a year for similar test and would not need detailed tests like biopsies, PET Scans or any other reports. Details of the study and research method have been published in scientific journal Thorax. Lung cancer is the leading cause of death in US and there are around 228,000 cases of the disease.