Posted on October 10, 2017

A collaborative study involving researchers from Hamad Medical Corporation (HMC) and Qatar University has developed an innovative information technology system (artificial intelligence) capable of predicting the future growth of damaged areas of the brain following an acute stroke.

“A stroke occurs when the blood supply to the brain, which supplies vital nutrients and oxygen, is impaired or cut off. When this happens areas of the brain can die. This damaged area is called an infarct. Currently, there is no accurate way to predict if the infarct will grow and by how much. The system we developed in this study is able to predict the future infarction growth for the patient, therefore providing important data to support our doctors when determining a suitable treatment plan,” explained Dr. Saadat Kamran, Senior Consultant Neurologist at HMC and one of the lead study investigators.

Stroke is one of the leading causes of mortality and the number one cause of chronic disability in the world. Due to the high prevalence of many risk factors for stroke among Qatar’s population – including diabetes, smoking, obesity, high cholesterol, hypertension, and inactivity – the incidence of stroke is high. “This study is an important milestone and the development of a predictive system could provide a key tool to assist doctors in the treatment of acute stroke patients and in clinical research,” said Dr. Kamran.

During the study, which has been published in the prestigious journal Nature Scientific Reports, the research team examined currently available models described in the literature and compared the models with the adaptive network-based fuzzy inference system (ANFIS), a method previously unused in the prediction of infarction growth. “By uploading real data on infarction growth to the ANFIS system we were able to develop an information technology system capable of predicting future infarction growth,” said Dr. Uvais Qidwai, from the KINDI Center for Computing Research at Qatar University, and one of the lead study investigators. “Early cross-correlation indicated similarity between the ANFIS predicted data and the original data of 82 percent, a very positive outcome which we expect to increase over time as we continue to add data and variables to the system,’ commented Dr. Qidwai.

The study team is continuing to develop the system to ensure it is capable of delivering the most accurate predictions possible. “We have already shown through our initial investigations that our ANFIS system is superior to any other previously established method. It is our hope that the system will provide valuable information for our clinical teams when treating acute stroke patients and enhance their decision-making ability. For example, the system will help in determining whether the best course of action is to operate quickly on a patient to prevent further brain damage, or if the best option is to wait as the damage is unlikely to increase,” explained Dr. Kamran. 

Categories: