Automated Midline Shift Detection in Head CTs Using Localization and Symmetry Techniques Based on User Selected Slice

BACKGROUND AND PURPOSE: Midline shift is an intracranial pathology characterized by the displacement of brain parenchyma across the skull’s midsagittal axis, typically caused by mass effect from space-occupying lesions or traumatic brain injuries. Prompt detection of midline shift is crucial, as delays in identification and intervention can negatively impact patient outcomes. The gap we have addressed in this work is the development of a deep learning algorithm that encompasses the full severity range from mild to severe cases of midline shift. Notably, in more severe cases, the mass effect often effaces the septum pellucidum, rendering it unusable as a fiducial point of reference.

MATERIALS AND METHODS: We sought to enable rapid and accurate detection of midline shift by leveraging advances in artificial intelligence. Using a cohort of 981 patient CT scans with a breadth of cerebral pathologies from our institution, we manually chose an individual slice from each CT scan primarily based on the presence of the lateral ventricles and annotated 400 of these scans for the lateral ventricles and skull-axis midline using Roboflow. Finally, we trained an artificial intelligence model based on the You Only Look Once object detection system to identify midline shifts in the individual slices of the remaining 581 CT scans.

RESULTS: When comparing normal and mild cases to moderate and severe cases of midline shift detection, our model yielded an AUC of 0.79 with a sensitivity of 0.73 and specificity of 0.72 indicating our model is sensitive enough to capture moderate and severe midline shifts and specific enough to differentiate them from mild and normal cases.

CONCLUSIONS: We developed an artificial intelligence model that reliably identifies the lateral ventricles and the cerebral midline across various pathologies in patient CT scans. Most importantly, our model accurately identifies and stratifies clinically significant and emergent midline shifts from non-emergent cases. This could serve as a foundational element for a future clinically integrated approach that flags urgent studies for expedited review, potentially facilitating more timely treatment when necessary.

ABBREVIATIONS: CT = Computed Tomography; AUC = Area Under the Curve; MLS = Midline Shift; IML = Ideal Midline.