Forecasting future Humphrey Visual Fields using deep learning.

PURPOSE: To determine if deep learning networks could be trained to forecast future 24-2 Humphrey Visual Fields (HVFs). METHODS: All data points from consecutive 24-2 HVFs from 1998 to 2018 were extracted from a university database. Ten-fold cross validation with a held out test set was used to develop the three main phases of model development: model architecture selection, dataset combination selection, and time-interval model training with transfer learning, to train a deep learning artificial neural network capable of generating a point-wise visual field prediction. The point-wise mean absolute error (PMAE) and difference in Mean Deviation (MD) between predicted and actual future HVF were calculated. RESULTS: More than 1.7 million perimetry points were extracted to the hundredth decibel from 32,443 24-2 HVFs. The best performing model with 20 million trainable parameters, CascadeNet-5, was selected. The overall point-wise PMAE for the test set was 2.47 dB (95% CI: 2.45 dB to 2.48 dB), and deep learning showed a statistically significant improvement over linear models. The 100 fully trained models successfully predicted future HVFs in glaucomatous eyes up to 5.5 years in the future with a correlation of 0.92 between the MD of predicted and actual future HVF and an average difference of 0.41 dB. CONCLUSIONS: Using unfiltered real-world datasets, deep learning networks show the ability to not only learn spatio-temporal HVF changes but also to generate predictions for future HVFs up to 5.5 years, given only a single HVF.

Generating retinal flow maps from structural optical coherence tomography with artificial intelligence.

Despite advances in artificial intelligence (AI), its application in medical imaging has been burdened and limited by expert-generated labels. We used images from optical coherence tomography angiography (OCTA), a relatively new imaging modality that measures retinal blood flow, to train an AI algorithm to generate flow maps from standard optical coherence tomography (OCT) images, exceeding the ability and bypassing the need for expert labeling. Deep learning was able to infer flow from single structural OCT images with similar fidelity to OCTA and significantly better than expert clinicians (P < 0.00001). Our model allows generating flow maps from large volumes of previously collected OCT data in existing clinical trials and clinical practice. This finding demonstrates a novel application of AI to medical imaging, whereby subtle regularities between different modalities are used to image the same body part and AI is used to generate detailed inferences of tissue function from structure imaging.

Diminished orientation-specific surround suppression of visual processing in schizophrenia.

Visual perception of a stimulus is a function of the visual context in which it is displayed. Surround suppression is a specific form of contextual modulation whereby the perceived contrast of a center stimulus is decreased by a high-contrast surround. Recent studies have demonstrated that individuals with schizophrenia are less prone to visual contextual effects, suggesting impairments in cortical lateral connectivity. We tested whether altered contextual modulation in schizophrenia is stimulus orientation selective. Participants viewed an annulus consisting of contrast-reversing sinusoidal gratings and determined if any one segment of the annulus had lower contrast relative to the other segments. Three stimulus configurations were tested: no surround (NS), parallel surround (PS), and orthogonal surround (OS). In the PS condition, the annulus was embedded in a 100% contrast grating parallel to the annulus gratings. In the OS condition, the surround grating was rotated 90 degrees relative to the orientation of the annulus gratings. The main dependent measure was the suppression index-the change in contrast threshold in the OS and PS conditions relative to the NS condition. There was a group x condition interaction such that patients had significantly lower PS suppression index than controls, but there were no group differences in the OS suppression index. We conclude that individuals with schizophrenia possess an abnormality in surround suppression that is specific for stimulus orientation. In conjunction with physiological and anatomical evidence from basic and postmortem studies, our results suggest a deficit of inhibition in primary visual cortex in schizophrenia.