Researchers have taken photographs of children’s retinas and screened them using a deep learning AI algorithm to diagnose autism with 100% accuracy. The findings support using AI as an objective screening tool for early diagnosis, especially when access to a specialist child psychiatrist is limited.
AI-screened eye pics diagnose childhood autism with 100% accuracy::undefined
I think that’s just a case of bad journalism, the author probably did not understand the paper properly. To quote:
To screen for symptom severity measured with ADOS-2 calibrated severity scores, 305 retinal photographs were used (154 for scores ≥8 and 151 for scores <8). The 10 models differentiated severe ASD from mild to moderate ASD measured with the ADOS-2 at the participant level, with a mean AUROC of 0.74 (95% CI, 0.67-0.80), sensitivity of 0.58 (95% CI, 0.49-0.66), specificity of 0.74 (95% CI, 0.67-0.82), and accuracy of 0.66 (95% CI, 0.60-0.73) for the test set
and
The models failed to screen for SRS-2–based symptom severity, with a mean AUROC of 0.44 (95% CI, 0.38-0.50), sensitivity of 0.52 (95% CI, 0.46-0.59), specificity of 0.44 (95% CI, 0.38-0.51), and accuracy of 0.48 (95% CI, 0.44-0.53) for the test set
Those numbers are far more reasonable and less sensationalistic. Eventually, all the original authors claim is that the method should be evaluated further and augmented by 3d data acquired via optical coherence tomography.
I think that’s just a case of bad journalism, the author probably did not understand the paper properly. To quote:
and
Those numbers are far more reasonable and less sensationalistic. Eventually, all the original authors claim is that the method should be evaluated further and augmented by 3d data acquired via optical coherence tomography.