Novel and accurate non-linear index for the automated detection of haemorrhagic brain stroke using CT images

Abstract Brain stroke is an emergency medical condition which occurs mainly due to insufficient blood flow the brain. It results in permanent cellular-level damage. There are two main types of brain stroke, ischemic and hemorrhagic. Ischemic caused by a lack flow, haemorrhagic form internal bleeding. The affected part will not function properly after this attack. Hence, early detection important for more efficacious treatment. Computer-aided diagnosis type non-invasive diagnostic tool can help detecting life-threatening disease its stage utilizing image processing soft computing techniques. In paper, we have developed one such model assess intracerebral haemorrhage employing non-linear features combined with probabilistic neural network classifier computed tomography (CT) images. Our achieved maximum accuracy 97.37% discerning normal versus subjects. An index also using only three significant features. clinical statistical validation confirms suitability providing improved treatment planning making strategic decisions.