Tag: Medical education

Posted on

AtlasGPT: a language model grounded in neurosurgery with domain-specific data and document retrieval

J Neurosurg 143:560–567, 2025

AtlasGPT, a neurosurgery-specific large language model grounded in expert-verified sources and retrieval-augmented generation, outperformed GPT-4 and Gemini Advanced on a neurosurgery board exam, showed greater resistance to medical misinformation, and generated more comprehensive, relevant, and well-referenced answer explanations than standard preparation materials.

• AtlasGPT is a neurosurgery-specific large language model (LLM) built on GPT-4 with retrieval-augmented generation (RAG) from trusted neurosurgical sources.

• AtlasGPT outperformed GPT-4 and Gemini Advanced on a 149-question neurosurgery board exam (accuracy: 90.6% vs 80.5%).

• AtlasGPT showed the highest accuracy on spine and imaging-based questions, even without access to image data.

• In adversarial testing, AtlasGPT was more robust to misinformation, being fooled only 14% of the time, compared to 44% for GPT-4 and 68% for Gemini Advanced.

• Expert neurosurgeons rated AtlasGPT’s explanations as more comprehensive, relevant, and better referenced than official board prep materials.

• AtlasGPT did not produce hallucinations or harmful content in its responses.

• The study suggests domain-specific LLMs like AtlasGPT can enhance medical education, decision-making, and exam preparation in complex fields.

• Limitations include use of a single question bank and need for broader source material in future work.

Posted on

Review of Cerebrospinal Fluid Physiology and Dynamics: A Call for Medical Education Reform

Neurosurgery 91:1–7, 2022

The flow of cerebrospinal fluid (CSF) has been described as a unidirectional system with the choroid plexus serving as the primary secretor of CSF and the arachnoid granulations as primary reabsorption site. This theory of neurosurgical forefathers has been universally adopted and taught as dogma. Many neuroscientists have found difficulty reconciling this theory with common pathologies, and recent studies have found that this “classic” hypothesis may not represent the full picture.

OBJECTIVE: To review modern CSF dynamic theories and to call formedical education reform.

METHODS: We reviewed the literature from January 1990 to December 2020. We searched the PubMed database using key terms “cerebrospinal fluid circulation,” “cerebrospinal fluid dynamics,” “cerebrospinal fluid physiology,” “glymphatic system,” and “glymphatic pathway.” We selected articles with a primary aim to discuss either CSF dynamics and/or the glymphatic system.

RESULTS: The Bulat–Klarica–Oreˇskovi´c hypothesis purports that CSF is secreted and reabsorbed throughout the craniospinal axis. CSF demonstrates similar physiology to that of water elsewhere in the body. CSF “circulates” throughout the subarachnoid space in a pulsatile to-and-fro fashion. Osmolarity plays a critical role in CSF dynamics. Aquaporin-4 and the glymphatic system contribute to CSF volume and flow by establishing osmolarity gradients and facilitating CSF movement. Multiple studies demonstrate that the choroid plexus does not play any significant role in CSF circulation.

CONCLUSION: We have highlighted major studies to illustrate modern principles of CSF dynamics. Despite these, the medical education system has been slow to reform curricula and update learning resources.

Posted on

Utility of postmortem imaging system for anatomical education in skull base surgery

Utility of postmortem imaging system for anatomical education in skull base surgery

Neurosurg Rev (2015) 38:165–170

Although cadaver dissections are important for skull base surgeons to acquire anatomical knowledge and techniques, their opportunities are limited in Japan. The Autopsy Imaging Center of the University of Fukui Hospital has both a CT scanner and an MR unit solely for deceased patients.

The authors applied the postmortemimaging to cadaver dissections and evaluated its usefulness in surgical education. Ten sides of five formalin-fixed cadaver heads were dissected by ten neurosurgeons. Five neurosurgeons were young, three were moderately experienced, and two were experts in skull base surgery. They performed orbitozygomatic, anterior transpetrosal, posterior transpetrosal, and transcondylar approaches. CT bone images were taken before and after dissections, and MR images were taken before dissection to merge with the CT bone images. The usefulness of the images for each neurosurgeon and for each skull base approach was evaluated.

The postmortem imaging system was useful for all neurosurgeons, especially in anterior transpetrosal, posterior transpetrosal, and transcondylar approaches. They could find the insufficiency or excessiveness of their drilling of specific bony structures with the images. Even the experts in skull base surgery could identify regions in which they could add drilling safely to widen the surgical field more.

The postmortem imaging system was useful for skull base cadaver dissections. This system is expected to be utilized for education and research on surgical anatomy.

Proudly powered by WordPress