Mario Serrano-Sosa1, Jared Van Snellenberg2, and Chuan Huang2,3
1Biomedical Engineering, Stony Brook University, Stony Brook, NY, United States, 2Psychiatry, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, United States, 3Radiology, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, United States
1Biomedical Engineering, Stony Brook University, Stony Brook, NY, United States, 2Psychiatry, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, United States, 3Radiology, Renaissance School of Medicine at Stony Brook University, Stony Brook, NY, United States
We have developed an interpretable deep
learning algorithm to predict working memory scores from 2-back fMRI data that was able to create averaged saliency maps
highlighting regions most predictive of working memory scores.
Figure 3: Averaged
saliency maps obtained after training and optimizing the CNN to predict WM
subconstruct scores.
Figure 2: Network
outputs for both CNN and KRR vs ground truth WM score. Blue dots are CNN
outputs and black triangles are KRR outputs.