Protocol: revolutionizing central nervous system tumour diagnosis in low- and middle-income countries: an innovative observational study on intraoperative smear and deep learning.

ABSTRACT

OBJECTIVE: The aim of this study is to assess the feasibility and implementation of a novel approach for intraoperative brain smears within the operating room, which is augmented with deep learning technology. Materials and methods: This study is designed as an observational to evaluate the feasibility and implementation of using an innovative approach to intraoperative brain smears within the operating room, augmented with deep learning technology. The study will be conducted at Aga Khan University Hospital in Karachi, Pakistan, from May 2024 to July 2026, with an estimated sample size of 258. A neurosurgical trainee, trained by the study neuropathologist, will prepare and examine the smears under a microscope in the operating room. The findings of the trainee will be documented and compared to routine intraoperative consultations (smear and/or frozen section) and final histopathology results obtained from the pathology department. Additionally, the study will incorporate artificial intelligence tools to assist with the interpretation of smear and a telepathology interface to enable consultation from an off-site neuropathologist. CONCLUSIONS: The results of this study will hold significant potential to revolutionise neurosurgery practices in lowand middle-income countries by introducing a cost-effective, efficient, and high-quality intraoperative consultation method to settings that currently lack the necessary infrastructure and expertise. The implementation of this innovative approach has the potential to improve patient outcomes and increase access to intraoperative diagnosis, thereby addressing a significant unmet need in LMICs.