Abstract: A method, computer program, and computer system is provided for predicting data from an image. An image is divided into one or more patch images. Spatial features corresponding to the one or more patch images are compressed. Output data corresponding to the compressed spatial features is predicted. The output data is predicted based on minimizing one or more loss functions corresponding to the compressed spatial features.

Abstract: A method, apparatus, and non-transitory computer-readable medium are provided for image processing of medical images. The method may include receiving one or more medical images. Tissue regions may be identified in each of the one or more medical images. The method may include generating patches associated with respective identified tissue regions. A patch tumor burden (TMB) score may be predicted for each of the patches based on a neural network model, wherein the neural network model is trained using a distribution smoother and a weighted loss function. The method may also include a tumor burden score being generated for each of the one or more medical images based on the patch tumor burden score for patches associated with the respective identified tissue regions in respective medical images from the one or more medical images.

Abstract: A method, computer program, and computer system is provided for predicting data from an image. An image is divided into one or more patch images. Spatial features corresponding to the one or more patch images are compressed. Output data corresponding to the compressed spatial features is predicted. The output data is predicted based on minimizing one or more loss functions corresponding to the compressed spatial features.