Abstract: Brain hemorrhage is a critical medical condition that requires accurate and timely diagnosis for effective treatment. Deep learning techniques have shown promising results in medical image analysis tasks, including the classification of brain hemorrhages. This paper proposes a brain hemorrhage classification system using the VGG16, ResNet18, ResNet50 convolutional neural network (CNN) architecture. The VGG16, ResNet18, ResNet50 architecture is known for its ability to capture intricate image features through the use of multiple parallel convolutional layers. The proposed system utilizes a dataset of brain images, including both normal and hemorrhage cases, to train the VGG16, ResNet18, ResNet50 model. The images are preprocessed and fed into the network, which learns to extract relevant features and classify the images into different hemorrhage categories, such as intracerebral hemorrhage, subarachnoid hemorrhage, and epidural hemorrhage. The training process involves optimizing the model parameters using backpropagation and gradient descent techniques. To evaluate the performance of the proposed system, extensive experiments are conducted on a separate test set of brain images. Various evaluation metrics, such as accuracy, precision, recall, and F1-score, are used to assess the classification results. The results demonstrate the effectiveness of the deep learning-based approach for brain hemorrhage classification, with the VGG16, ResNet18, ResNet50 model achieving high accuracy and reliable performance compared to traditional methods.

Keywords: Brain Hemorrhage, Deep Learning, VGG16, ResNet18, ResNet50, Convolutional Neural Network (CNN).

| DOI: 10.17148/IARJSET.2023.10593