Gökberk Erol

Gökberk Erol

Auto-generated profile, based on publicly available information

Gökberk Erol

Department of Neurosurgery, Yeditepe University School of Medicine, Istanbul, Turkey

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 April 2023 in Neurosurgical Review
Verified
Journal Article
Closed access
PDF file
USumut tan Sevgi
umut tan Sevgi
GEGökberk Erol
Gökberk Erol
YDYücel Doğruel
Yücel Doğruel
OSOsman Fikret Sönmez
Osman Fikret Sönmez
STShane Tubbs
Shane Tubbs
AGAbuzer Gungor
Abuzer Gungor

24 May 2025 in Diagnostics
Verified
Journal Article
Open access
PDF file
ASAlperen SOZER
Alperen SOZER
MSMustafa Sahin
Mustafa Sahin
BSBatuhan Sözer
Batuhan Sözer
GEGokberk Erol
Gokberk Erol
OTOzan Yavuz Tüfek
Ozan Yavuz Tüfek
ECEmrah Celtikci
Emrah Celtikci
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.

15 November 2024 in Neurosurgical Review
Verified
Journal Article
Closed access
PDF file
USumut tan Sevgi
umut tan Sevgi
AGAbuzer Gungor
Abuzer Gungor
GEGökberk Erol
Gökberk Erol
ÇCÇağrı Canbolat
Çağrı Canbolat
EMErik Middlebrooks
Erik Middlebrooks
UTUğur Türe
Uğur Türe

9 September 2025 in Acta Neurochirurgica
Verified
Journal Article
Open access
PDF file
MÇMahmut Çamlar
Mahmut Çamlar
USumut tan Sevgi
umut tan Sevgi
GEGökberk Erol
Gökberk Erol
FKFurkan Karakaş
Furkan Karakaş
YDYücel Doğruel
Yücel Doğruel
AGAbuzer Gungor
Abuzer Gungor
Background Recent studies suggest that large language models (LLMs) such as ChatGPT are useful tools for medical students or residents when preparing for examinations. These studies, especially those conducted with multiple-choice questions, emphasize that the level of knowledge and response consistency of the LLMs are generally acceptable; however, further optimization is needed in areas such as case discussion, interpretation, and language proficiency. Therefore, this study aimed to evaluate the performance of six distinct LLMs for Turkish and English neurosurgery multiple-choice questions and assess their accuracy and consistency in a specialized medical context. Methods A total of 599 multiple-choice questions drawn from Turkish Board examinations and an English neurosurgery question bank were presented to six LLMs (ChatGPT-o1pro, ChatGPT-4, AtlasGPT, Gemini, Copilot, and ChatGPT-3.5). Correctness rates were compared using the proportion z-test, and inter-model consistency was examined using Cohen’s kappa. Results ChatGPT-o1pro, ChatGPT-4, and AtlasGPT demonstrated relatively high accuracy for Single Best Answer–Recall of Knowledge (SBA-R), Single Best Answer–Interpretative Application of Knowledge (SBA-I), and True/False question types; however, performance notably decreased for questions with images, with some models leaving many unanswered items. Conclusion Our findings suggest that GPT-4-based models and AtlasGPT can handle specialized neurosurgery questions at a near-expert level for SBA-R, SBA-I, and True/False formats. Nevertheless, all models exhibit notable limitations in questions with images, indicating that these tools remain supplementary rather than definitive solutions for neurosurgical training and decision-making.

13 March 2026 in Operative Neurosurgery
Verified
Journal Article
Closed access
GEGökberk Erol
Gökberk Erol
AGAbuzer Gungor
Abuzer Gungor
USumut tan Sevgi
umut tan Sevgi
MGMuhammet Enes Gurses
Muhammet Enes Gurses
MŞMustafa Şahin
Mustafa Şahin
UTUğur Türe
Uğur Türe
BACKGROUND AND OBJECTIVES: The intricate anatomy of the posterior fossa comprises layered muscles, fasciae, and both extracranial and intracranial neurovascular structures. Surgical intervention in this region requires a precise and comprehensive understanding of these structures, which can be difficult to achieve with only 2-dimensional materials. Traditional cadaver dissections provide basic 3-dimensional anatomic insight but are increasingly constrained by cost, ethical concerns, and difficulties in cadaver sourcing. Advances in 3-dimensional (3D) modeling, particularly through photogrammetry, offer an innovative approach for improving visuospatial understanding for neurosurgical education. This study aimed to create realistic 3D models of the suboccipital musculature and neurovascular anatomy of the posterior fossa with photogrammetry. In addition, a middle fossa dissection was included to define cranial nerve pathways, providing a supplementary educational tool for neurosurgical training. METHODS: Five silicone-injected cadaveric head specimens were used for posterior and middle fossa dissection under a surgical microscope. Photographs were taken from multiple angles to generate 3D models through photogrammetry, which were then processed with the Scaniverse application and uploaded to the Sketchfab platform. Each model's anatomic structures were labeled and made accessible for augmented reality and virtual reality viewing on multiple devices, enhancing interactive learning experiences. Patient-specific step-by-step 3D models were generated using radiological segmentation to support preoperative planning for the posterior fossa approach. RESULTS: Sixteen 3D models were created, illustrating detailed anatomy of suboccipital and deep neck muscles, neurovascular structures, and cranial nerve pathways within the posterior and middle fossae. These models improve visualization and spatial understanding of complex anatomic relationships, making them a valuable resource for neurosurgical education. CONCLUSION: The 3D models developed in this study provide an effective tool to enhance the spatial understanding of the complex neurovascular anatomy of posterior fossa and cranial nerve routes, supporting and complementing traditional cadaver dissection in neurosurgical training.

10 January 2026 in Neurosurgical Review
Verified
Journal Article
Closed access
PDF file
ASAlperen SOZER
Alperen SOZER
MSMustafa Sahin
Mustafa Sahin
OTOzan Yavuz Tüfek
Ozan Yavuz Tüfek
FKFurkan Karakas
Furkan Karakas
GEGokberk Erol
Gokberk Erol
ECEmrah Celtikci
Emrah Celtikci