The system comprises an input device for acquiring H&E WSIS of cervical cancer, a data selection device for selecting H&E WSI of CC; a data pre- processing device for extracting tile thereby dividing images into reduced tiles, and normalizing color to remove the color variations, a central processor for training a convolution neural network based deep learning model to predict cervical cancer subtypes from histopathology images and guiding treatment decisions and determining the risk for developing the disease, wherein the CNN-based deep learning model comprises: a feature extractor for obtaining a feature vector representing the input tile and resizing each tile thereby normalizing by the mean and standard deviation of the ImageNet dataset before feeding to the pre-tramed ResNet-101 and a predictor for predicting subtypes of cervical cancer preferably selected from CADC and CSCC using feature vector.