Integrating AI and Semi-Immersive AR Technology for Brain Tumor Diagnosis

Brain tumours, characterised by abnormal cell development inside the brain, pose a significant risk to individuals due to their high mortality rate. Early identification is essential for accurate diagnosis and successful treatment to enhance patient outcomes. However, diagnosing small tumours can be difficult and error-prone, as it largely depends on the experience and qualifications of medical personnel. Therefore, an automated diagnostic approach that reduces diagnostic time while increasing accuracy is desperately needed. This study addresses the challenges of manual diagnosis by applying deep learning algorithms to brain tumour segmentation. Additional challenges in applying deep learning for tumour segmentation include handling low-quality scans, variations in tumour features, and insufficient data. We present an emerging concept combining Artificial Intelligence and Augmented Reality as a valuable tool, providing deep learning model deployment, immersive visualisation, and interactive functionalities to support diagnosis and surgical interventions for brain tumours. The achieved performance is competitive with the literature, and feedback from medical professionals after deployment evaluation is satisfactory.