![]() ![]() However, measurements of DVA can under-represent the visual dysfunction present in diabetes, with observational studies suggesting that contrast sensitivity, microperimetry, dark adaptation, matrix perimetry and colour vision are more sensitive outcome measures. Hence, in diabetic retinopathy, less progress has been made in the adoption of robust functional outcomes beyond best corrected distance visual acuity (DVA). However, the structural and spatial heterogeneity of diabetic retinopathy lesions makes identification of direct structure–function relationships challenging when compared with conditions such as glaucoma, in which mathematical models have been found to be useful. Psychophysical tests have been used in observational studies and clinical trials to track deteriorating function with disease progression. Many studies now suggest that neural dysfunction precedes the visible vascular signs typically used to diagnose the onset of diabetic retinopathy. While diabetic retinopathy is often described primarily in terms of vascular dysfunction, it has been recognised that the disturbed metabolic environment has a profound impact on neural cells. Anatomical signs of diabetic retinopathy may take years to develop, but early detection and treatment are crucial to prevent vision loss. Graphical Abstractĭiabetes mellitus is a major cause of vision loss through development of diabetic retinopathy and diabetic macular oedema (DMO). Together, these two techniques have great potential for developing prediction models using untidy real-world clinical data. Interpretable ML methods enabled us to identify profiles of visual function associated with different stages of diabetic eye disease, and to disentangle associations from artefacts of the data collection process. Conclusions/interpretationĮnsemble ML models predicted status of diabetic eye disease with high accuracy using just age, sex and measures of visual function. Microperimetry variables ranked highly for all three tasks, but this was partly due to a data artefact (a large proportion of missing values). Reading index was highly ranked for tasks A and B, whereas near visual acuity and Moorfields chart acuity were important for task C. The performance of the ensemble ML models was good across all three classification tasks, with accuracies of 0.92, 1.00 and 0.84, respectively, for tasks A–C, substantially exceeding the accuracies for logistic regression (0.84, 0.61 and 0.80, respectively). An interpretable ML technique was used to rank the contribution of visual function variables to predictions and to disentangle associations between diabetic eye disease and visual function from artefacts of the data collection process. More conventional multiple logistic regression models were also fitted for comparison. Ensemble ML models were fitted to classify group membership for three tasks, distinguishing (A) the DM no DR group from the no DM group (B) the DR no DMO group from the DM no DR group and (C) the DR with DMO group from the DR no DMO group. Participants were divided into four groups using clinical measures and grading of ophthalmic images: no diabetes mellitus (no DM), diabetes but no diabetic retinopathy (DM no DR), diabetic retinopathy without diabetic macular oedema (DR no DMO) and diabetic retinopathy with DMO (DR with DMO). Missing values were imputed using chained equations. We measured the visual function of 1901 eyes from 1032 participants in the Northern Ireland Sensory Ageing Study, deriving 12 variables from nine visual function tests. To determine the extent to which diabetic retinopathy severity stage may be classified using machine learning (ML) and commonly used clinical measures of visual function together with age and sex.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |