Search

To develop an artificial intelligence (AI) model that predicts anti-vascular endothelial growth factor (VEGF) agent-specific anatomical treatment outcomes in neovascular age-related macular degeneration (AMD), thereby assisting clinicians in selecting the most suitable anti-VEGF agent for each patient. This retrospective study included patients diagnosed with neovascular AMD who received three loading injections of either ranibizumab or aflibercept. Training was performed using optical coherence tomography (OCT) images with an attention generative adversarial network (GAN) model. To test the performance of the AI model, the sensitivity and specificity to predict the presence of retinal fluid after treatment were calculated for the AI model, an experienced (Examiner 1), and a less experienced (Examiner 2) human examiners. A total of 1684 OCT images from 842 patients (419 treated with ranibizumab and 423 treated with aflibercept) were used as the training set. Testing was performed using images from 98 patients. In patients treated with ranibizumab, the sensitivity and specificity, respectively, were 0.615 and 0.667 for the AI model, 0.385 and 0.861 for Examiner 1, and 0.231 and 0.806 for Examiner 2. In patients treated with aflibercept, the sensitivity and specificity, respectively, were 0.857 and 0.881 for the AI model, 0.429 and 0.976 for Examiner 1, and 0.429 and 0.857 for Examiner 2. In 18.5% of cases, the fluid status of synthetic posttreatment images differed between ranibizumab and aflibercept. The AI model using GAN might predict anti-VEGF agent-specific short-term treatment outcomes with relatively higher sensitivity than human examiners. Additionally, there was a difference in the efficacy in fluid resolution between the anti-VEGF agents. These results suggest the potential of AI in personalized medicine for patients with neovascular AMD., (© 2023. The Author(s).)

Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is thought to be a pure genetic form of subcortical vascular cognitive impairment (SVCI). The aim of this study was to compare white matter integrity and cortical thickness between typical CADASIL, a genetic form, and two sporadic forms of SVCI (with NOTCH3 and without NOTCH3 variants). We enrolled typical CADASIL patients (N = 11) and SVCI patients [with NOTCH3 variants (N = 15), without NOTCH3 variants (N = 101)]. To adjust the age difference, which reflects the known difference in clinical and radiologic courses between typical CADASIL patients and SVCI patients, we constructed a W-score of measurement for diffusion tensor image and cortical thickness. Typical CADASIL patients showed more frequent white matter hyperintensities in the bilateral posterior temporal region compared to SVCI patients (p < 0.001, uncorrected). We found that SVCI patients, regardless of the presence of NOTCH3 variants, showed significantly greater microstructural alterations (W-score, p < 0.05, FWE-corrected) and cortical thinning (W-score, p < 0.05, FDR-corrected) than typical CADASIL patients. In this study, typical CADASIL and SVCI showed distinct anatomic vulnerabilities in the cortical and subcortical structures. However, there was no difference between SVCI with NOTCH3 variants and SVCI without NOTCH3 variants.