Dual Success at World Congress on Neurosurgery

Deakin Applied Artificial Intelligence Initiative has achieved a significant milestone with the acceptance of two research papers to the prestigious 19th World Congress on Neurosurgery. These papers, developed in collaboration with Alfred Health, showcase the transformative potential of artificial intelligence (AI) in trauma care and clinical decision support.

“These projects reflect the Applied Artificial Intelligence Initiative’s commitment to advancing AI research in healthcare and the initiative’s strong partnerships with clinical institutions like Alfred Health.” – Dr. Rena Logothetis.

AI for CT Scan Analysis in Traumatic Brain Injury

The first abstract, “A Systematic Review of AI-Based Approaches for Automated CT Scan Analysis in Traumatic Brain Injury”, explores how AI can revolutionise the diagnosis and triaging of traumatic brain injuries (TBI).

With non-contrast CT scans being the primary imaging tool for acute TBI assessment, delays in manual interpretation can be critical. This study systematically reviews AI models that can automatically detect and measure TBI, alerting radiologists to severe cases.

Automating Trauma Checklist Extraction Using Large Language Models

The second paper, “Automating Trauma Checklist Extraction from Clinical Documents: A Study of Large Language Models and Prompting Techniques”, investigates how effective large language models (LLMs) can be in extracting checklists from clinical documents.

Traditionally, trauma checklists are manually extracted from clinical documents in a process prone to human error and inefficiency. Accurate checklists are a key component required to engineer clinical decision support systems (CDSS).

This study evaluates the use of LLMs to extract items from key neurotrauma checklists.

“Evaluating the effectiveness of AI in clinical settings with expert clinicians underscores the impact of interdisciplinary collaboration in shaping the future of medical diagnosis and decision-making.” – Deakin Distinguished Prof. Kon Mouzakis.

These papers will be presented at WFNS 2025 in Dubai this December.