3D-CNN based discrimination of schizophrenia using resting-state fMRI
作者：
Highlights：
• Very optimistic results for the automated discrimination of schizophrenia using state-of-the-art 3D deep learning architecture.
• For the classification, we have used 3D convolutional neural networks architectures.
• We achieve very high diagnostic accuracy with an area under the curve of 0.9982 and accuracy of 98.09% (p < 0.001).
• With this accuracy this research may be translated into an excellent tool to assist clinicians.
• 3D ICA based functional connectivity networks were used as the input features of the classifier.
摘要
•Very optimistic results for the automated discrimination of schizophrenia using state-of-the-art 3D deep learning architecture.•For the classification, we have used 3D convolutional neural networks architectures.•We achieve very high diagnostic accuracy with an area under the curve of 0.9982 and accuracy of 98.09% (p < 0.001).•With this accuracy this research may be translated into an excellent tool to assist clinicians.•3D ICA based functional connectivity networks were used as the input features of the classifier.
论文关键词：Neuroimaging,Resting-state fMRI,3D-group ICA,3D-CNN,Classification,TensorFlow,Schizophrenia
论文评审过程：Received 28 February 2019, Revised 23 May 2019, Accepted 21 June 2019, Available online 22 June 2019, Version of Record 26 June 2019.
论文官网地址：https://doi.org/10.1016/j.artmed.2019.06.003
