Citations

Introduction To assess and compare the accuracy, readability, and overall performance of large language models (LLMs) in answering questions about functional hypothalamic amenorrhea (FHA) for patients and healthcare professionals. Methods A total of 11 patient-level and 15 clinician-level FHA-related questions were entered separately into four LLMs: ChatGPT 3.5 (free version), ChatGPT 4.0 (updated, paid subscription), Gemini, and OpenEvidence. OpenEvidence was used only for clinician-based questions. Responses were evaluated by three expert reviewers blinded to the LLM used who rated them as accurate and complete, accurate but incomplete, or inaccurate. A fourth reviewer resolved discordant scores. Readability for patient-level questions was assessed using the Flesch Reading Ease Score (FRES) and word count. Lower FRES scores indicate more difficult reading. Accuracy and completeness were compared using odds ratios (95% CI) with ChatGPT 3.5 as the reference model, and differences in readability were analyzed using Friedman’s test. Results LLM performance varied across question types. For patient-level questions, ChatGPT 4.0 achieved the highest accuracy (9 of 11; 82%), followed by ChatGPT 3.5 and Gemini (each 8 of 11; 73%), with no statistically significant differences. Among clinician-level questions, OpenEvidence demonstrated perfect accuracy (15 of 15; 100%), compared with 93% for and 80% for ChatGPT 4.0 and Gemini. Completeness followed similar patterns, with OpenEvidence providing the most complete clinician responses (93%) and ChatGPT 4.0 the most complete patient-level responses (89%). Readability differed significantly among models ( p = 0.012), with Gemini producing the most readable patient-level content (median FRES 43.5 [IQR 36.8–53.4]) compared with ChatGPT 3.5 (30.6 [16.8–48.4]) and ChatGPT 4.0 (28.8 [22.1–37.6]). Word counts did not differ significantly ( p = 0.39). Discussion LLMs demonstrated good overall performance in answering FHA-related questions but often provided incorrect or incomplete information. Fine tuning field-specific data, engineered prompts, and obtaining human-in-the-loop feedback may help improve the accuracy of these models.

Abstract Objective Invasive presurgical evaluation plays a key role in pediatric epilepsy surgery, particularly in magnetic resonance imaging (MRI)‐negative cases, by guiding resective, disconnective, or ablative procedures. This International League Against Epilepsy (ILAE) Pediatric Epilepsy Surgery Taskforce study provides an updated global overview of current invasive evaluation practices. Methods Group‐level data were collected from 61 epilepsy surgery programs (49 pediatric‐only) in 29 countries across six continents. Included were children and adolescents who underwent presurgical evaluation and epilepsy surgery in 2023. The study was designed to enable comparison with the similar ILAE survey conducted in 2004. Results A total of 2427 patients were included. Invasive evaluations were performed in 21.1% of cases, most frequently in North America (33.7%, higher than Europe: 18.0%, p = .003). Among invasive cases, 32.3% had no detectable MRI abnormalities. The main indication for invasive evaluation was seizure onset localization (88.1%), followed by motor or sensory mapping (17.2%) and language mapping (14.2%). Stereoelectroencephalography (SEEG) was the predominant technique (19.8% overall, 93.6% of invasive cases), more common in North America (30.0%, p = .021) and less common in South America (7.4%, p < .001). Subdural electrodes were used in only 3.2% of invasive cases, and combined depth and subdural approaches in 3.2%. SEEG‐guided radiofrequency thermocoagulation (RF‐TC) was performed in 40.9% of SEEG cases, most commonly in Asia (63.8%). In 16.2% of invasive evaluations, patients did not proceed to resection, disconnection, or ablation, with the highest rate in Europe (28.5%). Significance This global survey provides the first broad overview of invasive evaluation practices in pediatric epilepsy surgery across participating centers worldwide. It highlights the widespread adoption of SEEG, declining use of subdural electrodes, and increasing application of SEEG‐guided RF‐TC. The high proportion of MRI‐negative cases and the considerable proportion of patients not proceeding to resection, disconnection, or ablation underscore the complexity of contemporary surgical candidates and the need for further refinement of selection strategies.