CPC G16H 50/20 (2018.01) [G06Q 10/1095 (2013.01); G06T 7/0012 (2013.01); G06V 10/454 (2022.01); G06V 10/82 (2022.01); G06V 20/70 (2022.01); G06V 40/18 (2022.01); G16H 15/00 (2018.01); G16H 50/30 (2018.01); G06V 2201/03 (2022.01)] | 19 Claims |
1. A method of detecting glaucoma, the method comprising: pre-training at least one neural network model of a plurality of neural network models using a dataset, the pre-training including: generating a plurality of vectors based on eye image variations within the dataset, wherein each vector includes a plurality of labels designating eye conditions; assigning a class label to each vector based on the plurality of labels; and generating a glaucoma-specific pre-trained model based on the assigned class labels; training the plurality of neural network models based on a plurality of indications of glaucoma based on retinal data; simultaneously generating a risk score associated with each of the plurality of indications based on the trained plurality of neural network models; combining the risk score associated with each of the plurality of indications, a socio-demographic parameter, and a presence of diabetes as a feature vector, based on a classification model, generating a probability score of glaucoma based on the feature vector, using the trained plurality of neural network models; producing based on the probability score a likelihood of glaucoma; and determining whether glaucoma is present based on the likelihood of glaucoma.
|