Artificial Intelligence Based Techniques for Identification of Neonatal Brain Hemorrhage: A Review

The medical imaging field needs the development of machine learning and deep learning methods for meeting inbuilt radiological image processing complexities and to enhance the treatment method using Artificial Intelligence (AI) techniques. To implement these techniques large dataset required can be obtained with—augmentation, segmentation, and deep ensemble approaches like Generative Adversarial Network (GAN), U-Net and deep ensemble neural networks respectively for Neonatal Brain Hemorrhage (NBH). The contribution of Deep Learning (DL) is found to be more significant in the Magnetic Resonance Image (MRI) processing of neonates. The review article emphasized the role of AI in clinical outcomes and the early diagnosis of NBH with newly developed algorithms. Several pre-processing, segmentation, feature extraction and classification approaches are discussed to provide a clear information to future researchers.