Computer assisted diagnosis of bipolar disorder using invariant features

Abstract

In the study, a supervised learning framework is focused to identify the bipolar disorder (BD) using structural magnetic resonance imaging is focused. The work is based on the newly developed 3D SIFT and 3D SURF feature vectors with pattern recognition technique. The overall hypothesis is to deduct BD results from dysfunctional cellular metabolism within specific brain systems (i.e., anterior limbic brain network) as reflected in abnormalities in brain activation patterns and in specific neurochemical measures. The proposed method is used to integrate neuroimaging in exploring the biomarkers of BD to reveal the mechanism. In the method, two newly developed feature vectors and kernel PCA are combined or connected to project the feature vectors. Diagnosis process is done by random forest. The results reveal that the method has high potential to identify the BD than earlier works, and an average accuracy of 77.77% is reached. This research reveals that neuroimaging studies will help to differentiate BD from healthy controls.

Detection of bipolar disorder by means of ensemble machine learning classifier

“…In an article (9) promising results have been found in predicting tasks while researching machine learning algorithms including Support Vector Machines, Random Forest, Naive Bayes, and Multilayer Perceptron. Some articles (10,11) presents the proposed approach, which is a blend of varoius deep learning architectures: bidirectional long-term memory (biLSTMs) and convolutional neural networks (CNNs).…”