Background/Objectives: Large language models (LLMs) are revolutionizing the world and the field of medicine while constantly improving themselves. With recent advancements in image interpretation, evaluating the reasoning capabilities of these models and benchmarking their performance on brain MRI tasks has become crucial, as they may be utilized—albeit off-label—for patient care by both neurosurgeons and non-neurosurgeons. Methods: ChatGPT-4o, Grok, and Gemini were presented with 35,711 slices of brain MRI, including various pathologies and normal MRIs. Models were asked to identify the MRI sequence and determine the presence of pathology. Their individual performances were measured and compared with one another. Results: GPT refused to answer 28.02% of the slices despite three attempts, whereas Grok and Gemini provided responses on the first attempt for every slice. Gemini achieved 74.54% pathology prediction and 46.38% sequence prediction accuracy. GPT-4o achieved 74.33% pathology prediction and 85.98% sequence prediction accuracy for questions that it had answered (53.50% and 61.67% in total, respectively). Grok achieved 65.64% pathology prediction and 66.23% sequence prediction accuracy. Conclusions: The image interpretation capabilities of the investigated LLMs are limited for now and require further refinement before competing with specifically trained and fine-tuned dedicated applications. Amongst them, Gemini outperforms the others in pathology prediction while Grok outperforms others in sequence prediction. These limitations should be kept in mind if use during patient care is planned.

Background: Craniopharyngiomas are rare tumors that originate from the squamous epithelium in Rathke's sac residues. There is a limited body of literature on the effect of using intraoperative magnetic resonance imaging (IO-MRI) in resection of craniopharyngiomas via endoscopic endonasal surgery (EES). The present study aims to assess the outcomes of EES for resection of craniopharyngiomas using IO-MRI. Methods: This is a retrospective analysis of the data collected on the point of care, consisting of 27 patients who underwent EES for histologically confirmed craniopharingiomas. All procedures were performed by a single neurosurgeon in between 2018 and 2020. Results: IO-MRI showed gross total resections were achieved in 25 (92.6 %) patients, and subtotal resections were achieved in 2 (7.4 %) patients. Pre-anticipated resection rate was achieved in the intraoperative scans in all patients and none of the patients required further resection. IO-MRI prolonged the anesthesia duration by an average of 32 +/- 4 min. In the follow-up, 24 (88.9 %) cases of permanent diabetes insipidus and 10 (37 %) cases of panhypopituitarism were observed. Cerebrospinal fluid rhinorrhea complicated 6 (22.2 %) cases. Conclusion: This present study is the largest study evaluating the efficiency of IO-MRI in craniopharyngioma. In addition, unlike other studies, only the patients who underwent EES were included. Although it has been known that the EES for craniopharyngioma is an effective and less invasive technique, this presented early results would suggest that combining this method with IO-MRI did not make significant contributions to surgical outcome. (c) 2024 Asian Surgical Association and Taiwan Society of Coloproctology. Publishing services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/ by-nc-nd/4.0/).

Gamma Knife radiosurgery (GKRS) is a widely used treatment for intracranial meningiomas; however, its effects on endoplasmic reticulum (ER) stress–related molecular pathways remain unclear. This study aimed to evaluate changes in circulating ER stress–associated chaperone proteins before and three months after GKRS and to explore their relationship with radiological and dosimetric findings. A total of 27 patients with meningioma treated with GKRS and 20 healthy controls were included. Peripheral blood samples were collected prior to treatment and three months post-treatment. Serum levels of glucose-regulated protein 78 (GRP78), calreticulin (CALR), calnexin (CNX), mesencephalic astrocyte-derived neurotrophic factor (MANF), protein disulfide isomerase (PDI), and protein disulfide isomerase A3 (PDIA3) were measured using enzyme-linked immunosorbent assay (ELISA). Magnetic resonance imaging was used to evaluate tumor volume and peritumoral edema. Dosimetric parameters were obtained from the treatment planning system. Pre-treatment levels of PDIA3, CALR, CNX, MANF, and PDI were higher in patients with meningioma than in healthy controls, with statistically significant differences observed for PDIA3, CALR, CNX, MANF, and PDI. Following GKRS, all six investigated proteins showed numerical decreases. Nominally significant pre- to post-treatment reductions were observed for PDIA3, CALR, MANF, and PDI; however, after Holm correction for multiple comparisons, only the reduction in CALR remained statistically significant. MRI assessment demonstrated tumor volume reduction in 5 of 27 patients (18.5%), while tumor volume remained stable in 22 patients (81.5%). GKRS was associated with measurable changes in circulating ER stress–related proteins. These findings indicate a temporal association rather than a causal effect and should be interpreted with caution. Although these markers are not suitable as standalone biomarkers, they may provide complementary information when evaluated together with radiological findings. Further longitudinal and mechanistic studies are needed to determine their biological significance and potential clinical relevance.