< Back to previous page

Project

A New Traumatic Axonal Injury Classification Scheme based in Clinical and Improved MR Imaging Biomarkers (TAI-MRI)

Traumatic Axonal Injury (TAI) is now considered to be a frequent and important injury in all severities of traumatic brain injury (TBI). The global aim of TAI-MRI is to develop a novel classification for TAI using data from multimodal MRI and to determine its clinical value for the characterization of injury severity and prediction of outcome.This project, involving 4 partners, will use MRI datasets obtained early after injury (including clinical and advanced MRI) from two local studies (The Trondheim and Cambridge TBI studies: ~580 patients) and the EU-funded multicenter CENTER-TBI study (~800 patients). TAI-MRI will thus be the largest MRI study worldwide. These datasets comprise a comprehensive collection of acute phase variables reflecting the severity of injury with the possibility to adjust for confounding variables and outcome measures at multiple time points during the first year. Several training sets will be used for model selection. Automated methods involving deep learning techniques will be developed and used for lesion mapping in combination with manual assessments. Methods for computer aided diagnosis (CAD) will be refined and validated, and analyses will determine which aspects of CAD based evaluation could replace expert clinical evaluation by radiologists. Finally, this novel MRI classification system will be validated in the large CENTER-TBI dataset.An improved MRI-based classification system of TAI will provide both a better assessment of injury severity in the acute phase and better outcome prediction. Recent advances in CAD provide a unique opportunity to develop a classification with great clinical applicability. Hence, we will provide a timely new tool for neuroradiologists, clinicians and researchers to facilitate TBI diagnosis, thus improving the treatment and rehabilitation of TBI patients. Finally, TAI-MRI will bring the field forward by increasing our understanding of the pathophysiology of TBI, and how reduced consciousness can be linked to injury type and location and outcome.
Date:1 Sep 2017 →  31 Aug 2019
Keywords:MEDICAL IMAGING
Disciplines:Multimedia processing, Biological system engineering, Signal processing