Abstract:
Deep learning has been successful in classifying the label and severity stage of certain diseases, but few of them explain how to make predictions. We propose to use deep learning in medical diagnosis to identify pathogens, inspired by Koch’s Postulates. We can identify the symptoms a diabetic retinopathy (DR) detector predicts by isolating and visualizing neuron activation patterns. We first define novel pathological descriptors using activated DR detector neurons to encode lesion spatial and appearance information. Patho-GAN, a new network that synthesizes medically plausible retinal images, visualizes the descriptor’s symptom. These descriptors let us arbitrarily control lesions’ position, quantity, and categories. Our synthesized images also exhibit diabetic retinopathy symptoms. Our images are qualitatively and quantitatively better than previous methods. Our second-level speed can be used for data augmentation, unlike existing methods that take hours to generate an image.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
