Objective: Schizophrenia (SCH) is primarily diagnosed based on specific clinical symptoms, with the lack of any objective SCH-related biomarkers often resulting in patient misdiagnosis and the underdiagnosis of this condition. This study was developed to assess the utility of amplitude of low-frequency fluctuation (ALFF) values analyzed via support vector machine (SVM) methods as a means of diagnosing SCH.
Methods: In total, 131 SCH patients and 128 age- and gender-matched healthy control (HC) individuals underwent resting-state functional magnetic resonance imaging (rs-fMRI), with the resultant data then being analyzed using ALFF values and SVM methods.
Results: Relative to HC individuals, patients with SCH exhibited ALFF reductions in the left angular gyrus (AG), fusiform gyrus, anterior cingulate cortex (ACC), right cerebellum, bilateral middle temporal gyrus (MTG), and precuneus (PCu) regions. No SCH patient brain regions exhibited significant increases in ALFF relative to HC individuals. SVM results indicated that reductions in ALFF values in the bilateral PCu can be used to effectively differentiate between SCH patients and HCs with respective accuracy, sensitivity, and specificity values of 73.36, 91.60, and 54.69%.
Conclusion: These data indicate that SCH patients may exhibit characteristic reductions in regional brain activity, with decreased ALFF values of the bilateral PCu potentially offering value as a candidate biomarker capable of distinguishing between SCH patients and HCs.
Keywords: amplitude of low-frequency fluctuation; imaging biomarker; resting-state fMRI; schizophrenia; support vector machine.
Copyright © 2022 Gao, Tong, Hu, Huang, Guo, Wang, Li and Wang.