From severity scoring to predictive analytics: the emerging role of AI in neurosurgery

The rapid integration of artificial intelligence (AI) into neurosurgical practice is transforming every phase of patient care from diagnostic imaging and preoperative planning to intraoperative decision-making and postoperative management. This narrative review traces the evolution of data-driven neurosurgery, beginning with traditional severity scoring systems and advancing toward predictive analytics and intelligent automation. By examining structured data (such as electronic health records and laboratory values) alongside complex unstructured inputs (including neuroimaging, surgical videos, and free-text notes), can extract clinically meaningful patterns, with reported performance metrics such as Dice scores of 0.82-0.84 for tumor segmentation and AUC values of 0.80-0.90 for molecular prediction and outcome forecasting. Applications in lesion detection, surgical navigation, prognostication, and rehabilitation are discussed, along with critical challenges in interpretability, data harmonization, bias mitigation, and regulatory approval. Emerging paradigms such as federated learning, generative AI, and continuous learning ecosystems are also explored as future pathways toward ethical, adaptive, and globally connected neurosurgical intelligence. As a narrative review, this work synthesizes key developments qualitatively; specific performance metrics and limitations regarding systematic selection, quantitative synthesis, and variable model validation are addressed. Ultimately, AI serves not as a replacement for the neurosurgeon but as a cognitive collaborator, augmenting precision, efficiency, and patient-centered outcomes in modern neurosurgery.