Automated Detection of Schizophrenia from Brain MRI Slices using Optimized Deep-Features

In humans, the incident rate of mental illness is gradually rising due to various causes. Schizophrenia is one of the chronic mental illness and its happing rate also rising in the current era. The patient with Schizophrenia will experience a confused mental condition and a timely recognition and treatment is essential to reduce the risk. The proposed work aims to implement a methodology to support the automated detection of Schizophrenia from the brain MRI slices of T1 modality (T1W). The assessment of brain MRI is executed using the pre-trained VGG16 system and the deep-features extracted are optimized with the Slime-Mould-Algorithm (SMA) and the reduced features are then considered to train, test and validate the binary classifiers employed in this work. This research is implemented using 500 images of each case (healthy/abnormal) and the attained result with the SVM-Cubic is superior compared to other classifiers considered in the automated disease detection system.