Researchers at Yonsei University School of Medicine (South Korea) have developed an AI model that can detect childhood autism with near 100% accuracy.
According to the study, “Retinal photographs of individuals with ASD were prospectively collected between April and October 2022, and those of age- and sex-matched individuals with TD (typical development) were retrospectively collected between December 2007 and February 2023. Deep ensembles of 5 models were built with 10-fold cross-validation using the pretrained ResNeXt-50 (32×4d) network. Score-weighted visual explanations for convolutional neural networks, with a progressive erasing technique, were used for model visualization and quantitative validation. Data analysis was performed between December 2022 and October 2023.
“This study included 1890 eyes of 958 participants. The ASD and TD groups each included 479 participants (945 eyes), had a mean (SD) age of 7.8 (3.2) years, and comprised mostly boys (392 [81.8%]). For ASD screening, the models had a mean AUROC, sensitivity, and specificity of 1.00 (95% CI, 1.00-1.00) on the test set. These models retained a mean AUROC of 1.00 using only 10% of the image containing the optic disc. For symptom severity screening, the models had a mean AUROC of 0.74 (95% CI, 0.67-0.80), sensitivity of 0.58 (95% CI, 0.49-0.66), and specificity of 0.74 (95% CI, 0.67-0.82) on the test set.”
With a confidence interval of 95%, this is a strong indication that the team of researchers at Yonsei University have come close to solving this detection test. This is a magnificent breakthrough for the treatment and early detection prognosis for ASD. Computer vision and machine learning once again prove to be a formidable pair for combating complex diseases and disorders.
Subscribe to stay in touch with us and get the latest news and articles delivered straight to your inbox!