Healthcare-AI

Two Deep Learning CNN (Convolutional Neural Network) models are proposed :-

1. ResNet (Residual Network) Classifier Model – To confirm the possibility of a brain tumor.
2. ResUNet Segmentation Model – To detect the exact location of the tumor.

The methodology of Transfer Learning is applied to re-train these Models in order to perform the classification and localization of tumors through the Brain MRI scans. Transfer Learning is used since starting from a pretrained model can drastically reduce the computational time required if the training is performed from scratch.

Dataset

https://www.kaggle.com/mateuszbuda/lgg-mri-segmentation

Sample Outputs

ResNet Classifier Model (Tumor Detection)

---

ResUNet Segmentation Model (Tumor Localization)