BIFOLD Colloquium 14/2025
November
18,
2025 
16:00
- 17:00
AI in Neurosurgery – Current State and Beyond
Abstract: Artificial intelligence (AI) is emerging as a transformative paradigm in neurosurgery, offering data-driven tools to enhance diagnostic accuracy, prognostic modeling, and intraoperative decision-making. This presentation delineates contemporary developments and future directions in the field, emphasizing the role of supervised, unsupervised, reinforcement, and generative learning approaches. Recent advances in multimodal data integration, deep learning–based image segmentation, outcome prediction, and synthetic vascular modeling underscore AI’s capacity to augment clinical reasoning and workflow efficiency. Nevertheless, critical challenges persist, including data heterogeneity, model interpretability, external validation, and clinical trustworthiness. Research conducted at the Charité Lab for AI in Medicine (CLAIM) demonstrates progress in automated neuroimaging pipelines, privacy-preserving model training, and machine learning–assisted supramarginal resection (MARS). Collectively, these efforts illustrate AI’s potential to drive precision neurosurgery through robust, reproducible, and ethically grounded methodologies that integrate computational intelligence with clinical expertise.
Bio: PD Dr. med. Dietmar Frey is a board-certified neurosurgeon and founding director of CLAIM - the Charité Lab for Artificial Intelligence in Medicine. After completing law school and passing both state examinations in law, he studied medicine at the Charité and graduated in 2007 with a grade of 1.0 and joined the neurosurgery department at Charité. In 2014, he completed his specialist training in neurosurgery. Since 2016, he has been building the Charité Lab for AI in Medicine at the Charité. Funded by several federal and European grants he has assembled a strong interdisciplinary team of physicians, machine learning engineers and computer scientists to develop an AI-based decision support system that provides risk scoring and treatment planning in stroke. Currently CLAIM is involved in many activities across the medical spectrum combining data science, machine learning, GenAI and medical decision-making. Dietmar Frey is the coordinator of the Horizon Europe project VALIDATE (Validation of a Trustworthy AI-based Prognostic Tool for Predicting Patient Outcome in Acute Stroke, #101057263, 2022-2026) to further develop, test, and - in a multi-center trial - validate a clinical decision support for the treatment stratification of acute stroke patients to improve patient outcome. He also coordinates CYLCOMED, an EU-funded project to improve cybersecurity on connected medical devices (2022-25) and is partnering in the EU Horizon Europe project STRATIF-AI. In addition, he is partnering in the BMBF-funded project ANONY-MED in which secure and safe anonymization procedures for medical data are developed, validated and evaluated.