Intraoperative ultrasound monitoring of spinal cord swelling and parenchymal changes in a porcine model of thoracic spinal cord injury

Abstract

Tools for monitoring traumatic spinal cord injury (SCI) severity and progression are limited. Ultrasound (US) imaging may be suitable but requires improved understanding of post-SCI spinal cord changes observed on US. Using serial US images in a clinically relevant porcine model of SCI, this study: (1) determined temporal changes to the spinal cord and subarachnoid space over 24-h post-SCI; and, (2) quantitatively compared US to magnetic resonance imaging (MRI) and histology at 24-h post-SCI. Ten anesthetised females pigs received a thoracic contusion SCI across three injury groups, and midsagittal B-mode US images were obtained hourly from baseline over 24 h. At 24-h, T2-weighted MRI was performed and spinal cord tissue was harvested for histology. Spinal cord and dura diameters were extracted from US. Greyscale distribution parameters of spinal cord parenchyma were assessed on US and MRI. The following US-based parameters were used to assess injury progression: spinal cord swelling, subarachnoid occlusion, dural distension, and parenchymal echogenicity (median and interquartile range [IQR] of the greyscale distribution). On US, for all animals, parenchymal echogenicity (median and IQR) increased rapidly within 2-h, subarachnoid occlusion occurred by 13-h, and maximal spinal cord swelling occurred by 23-h, accompanied by dura distension up to 10 %. Increased US median echogenicity was correlated with histologically measured intraparenchymal haemorrhage and tissue loss, but not with MRI signal changes. These findings support the use of intraoperative US as an objective, real-time, tool for assessing SCI progression.