Acute Neurological Condition Forecasting And Progression

Triage for stroke patients is time critical. CT is the standard imaging ordered for patients presenting to emergency departments (typically within 15 minutes of arrival at ED).  However, MRI is more sensitive to the haemorrhage but is rarely available immediately. The diagnosis of acute neurological emergencies, like stroke and traumatic brain injury (TBI), presents significant challenges, with the sensitivity of initial CT imaging often insufficient for early lesion detection, necessitating the more sensitive MRI that may not be promptly available. Prior research has indicated the potential of AI in enhancing diagnostic imaging, yet there remains a significant gap in its predictive application for stroke and TBI progression.

Leveraging recent advances in machine learning, our research study aims to develop an AI model that predicts the evolution of these conditions from initial CT scans, potentially transforming emergency neuroimaging. This study aims to refine these models, with the hypothesis that AI can bridge the gap between CT’s availability and MRI’s sensitivity, thereby improving diagnostic accuracy and patient outcomes. This research study is a collaborative effort between radiology and computer science, aiming to advance the field of emergency imaging and patient care.