Differentiating schizophrenic patients from healthy control; application of machine learning to resting state fmri

In recent years, one analysis approach that has grown in popularity is the use of machine learning algorithms to train classifiers to decode stimuli, mental states, behaviors and other variables of interest from fMRI data. Most of these studies focus on fMRI low frequency oscillations. This study focuses on the amplitude of low-frequency fluctuations (ALFF) and fractional amplitude of low-frequency fluctuations (fALFF). A Voxel-wise analysis is performed on the whole brain for two groups of subjects. A machine learning algorithm is applied to two independent groups of subjects (a total of 160 healthy control and schizophrenic subjects) to classify Schizophrenia subjects from healthy control. Kendall tau rank correlation coefficient is also used to dominate most important voxels (features). This study is done on three datasets: a) fALFF b) mALFF dataset and c) combination of mALFF and fALFF. The results show that using the combination dataset improves the classification and demonstrates that machine learning algorithms can extract new information from a resting state image of schizophrenia which can help in diagnosing and treating schizophrenic patients in the future. Future studies can focus on testing these algorithms on different modalities and moreover on different physiological disorders.

Name thesis:degree_name

Master of Science in Biomedical Engineering - (M.S.)

Discipline thesis:degree_discipline

Biomedical Engineering

Year

2014

Author dc:creator

• Ebrahimi Nezhad, Hossein

Contributors dc:contributor

• Bharat Biswal
• Xin Di
• Zhi Wei

• Machine learning
• Amplitude of low-frequency fluctuations (ALFF)
• Fractional amplitude of low-frequency fluctuations (fALFF)
• Biomedical Engineering and Bioengineering

Repository record dc:identifier

https://digitalcommons.njit.edu/theses/184

OAI identifier oai:identifier

oai:digitalcommons.njit.edu:theses-1183