Khanna, Cheryl L

Background Despite, the potential clinical utility of 60–4 visual fields, they are not frequently used in clinical practice partly, due to the purported impact of facial contour on field defects. The purpose of this study was to design and test an artificial intelligence-driven platform to predict facial structure-dependent visual field defects on 60–4 visual field tests. Methods Subjects with no ocular pathology were included. Participants were subject to optical coherence tomography, 60–4 Swedish interactive thresholding algorithm visual field tests and photography. The predicted visual field was compared with observed 60–4 visual field results in subjects. Average and point-specific sensitivity, specificity, precision, negative predictive value, accuracy, and F1-scores were primary outcome measures. Results 30 healthy were enrolled. Three-dimensional facial reconstruction using a convolution neural network (CNN) was able to predict facial contour-dependent 60–4 visual field defects in 30 subjects without ocular pathology. Overall model accuracy was 97%±3% and 96%±3% and the F1-score, dependent on precision and sensitivity, was 58%±19% and 55%±15% for the right eye and left eye, respectively. Spatial-dependent model performance was observed with increased sensitivity and precision within the far inferior nasal field reflected by an average F1-score of 76%±20% and 70%±29% for the right eye and left eye, respectively. Conclusions This pilot study reports the development of a CNN-enhanced platform capable of predicting 60–4 visual field defects in healthy controls based on facial contour. Further study with this platform may enhance understanding of the influence of facial contour on 60–4 visual field testing.

Purpose: New-onset persistent diplopia has become a common complication after glaucoma drainage device (GDD) placement. Understanding the orbital anatomy of such patients may provide information regarding risk of diplopia, GDD selection, and post-operative management. The purpose of this study was to examine the orbital anatomic differences in diplopic and non-diplopic patients after GDD implantation using high-resolution MRI. Methods: Seven eyes ( N = 4 with diplopia and N = 3 without diplopia after GDD placement) of seven patients that had undergone placement of Baerveldt 250 (B250), Baerveldt 350 (B350), or Ahmed FP7 (FP7) GDD were prospectively enrolled at a single institution. All patients underwent a 3.0T orbital MRI with 3D volumetric T1 and T2 weighted sequence. Images were analyzed for orbital volume, axial length, orbital distances, presence of superior rectus-lateral rectus (SR-LR) band, position of GDD, and SR-LR angles. Results: Patients with diplopia had smaller mean ± SD orbital axial (911.5 ± 111.8 mm 3 vs 931.7 ± 79.7 mm 3 ) and coronal volumes (1162.5 ± 145.5 mm 3 vs 1180 ± 34.6 mm 3 ) compared to non-diplopic patients. Average orbital rim distances were larger for the diplopic group. The SR-LR displacement angle for diplopic patients was larger (101.6° ± 8.1 vs 94.7° ± 17.6) while the SR-LR quadrantic angle (86.6° ± 4.2 vs 89.1° ± 4.3) was smaller. SR-LR band was present and intact in all patients. GDD malpositioning was not evident in any patient. Conclusion: The decreased orbital axial and coronal volumes as well as increased orbital rim distances in diplopic patients suggests the need for further studies to understand the role of orbital anatomy in occurrence of diplopia. Dynamic MRI imaging may be helpful in identifying differences in extraocular muscle function that reveal an etiology of diplopia in patients with GDD implantation.