The detection of cerebral aneurysms, particularly among professional divers and swimmers, is crucial due to the high physical demands and pressure changes experienced in these activities. Currently, identifying these "intracranial bombs" is challenging, often leading to subarachnoid hemorrhage with high mortality and disability rates. Clipping surgery and endovascular embolization are the primary treatments, but early detection is vital for effective intervention. This study introduces the Shark Smell Optimization and Deep Learning-Enabled Automated Intracranial Aneurysms (SSODLE-AIA) model, specifically tailored for the aquatic sports community. The SSODLE-AIA model innovatively partitions cerebral aneurysms into uniform blocks, employing an EfficientNet-based feature extractor for generating feature vectors. It uniquely integrates Shark Smell Optimization (SSO) for optimal hyperparameter tuning, enhancing the model's relevance to the diving and swimming domains where sensory acuity is paramount. Furthermore, a Bidirectional Gated Recurrent Unit (BiGRU) model classifies these blocks into two types: smooth and structured. This classification is crucial for divers and swimmers, whose cerebral structures may adapt to their aquatic environments. The identification process includes mean and patch matching for these regions, ensuring high precision in detecting subtle aneurysm-related changes. The SSODLE-AIA model's effectiveness is evaluated using a cerebral aneurysm dataset. Our experimental results show that this model outperforms existing techniques, offering a promising tool for early aneurysm detection in athletes exposed to unique aquatic pressures and environments. This advancement not only aids in timely medical intervention but also contributes to the safety and longevity of careers in professional diving and swimming.