Dataset Information

Classification algorithms using multiple MRI features in mild traumatic brain injury.

ABSTRACT:

Objective

The purpose of this study was to develop an algorithm incorporating MRI metrics to classify patients with mild traumatic brain injury (mTBI) and controls.

Methods

This was an institutional review board-approved, Health Insurance Portability and Accountability Act-compliant prospective study. We recruited patients with mTBI and healthy controls through the emergency department and general population. We acquired data on a 3.0T Siemens Trio magnet including conventional brain imaging, resting-state fMRI, diffusion-weighted imaging, and magnetic field correlation (MFC), and performed multifeature analysis using the following MRI metrics: mean kurtosis (MK) of thalamus, MFC of thalamus and frontal white matter, thalamocortical resting-state networks, and 5 regional gray matter and white matter volumes including the anterior cingulum and left frontal and temporal poles. Feature selection was performed using minimal-redundancy maximal-relevance. We used classifiers including support vector machine, naive Bayesian, Bayesian network, radial basis network, and multilayer perceptron to test maximal accuracy.

Results

We studied 24 patients with mTBI and 26 controls. Best single-feature classification uses thalamic MK yielding 74% accuracy. Multifeature analysis yields 80% accuracy using the full feature set, and up to 86% accuracy using minimal-redundancy maximal-relevance feature selection (MK thalamus, right anterior cingulate volume, thalamic thickness, thalamocortical resting-state network, thalamic microscopic MFC, and sex).

Conclusion

Multifeature analysis using diffusion-weighted imaging, MFC, fMRI, and volumetrics may aid in the classification of patients with mTBI compared with controls based on optimal feature selection and classification methods.

Classification of evidence

This study provides Class III evidence that classification algorithms using multiple MRI features accurately identifies patients with mTBI as defined by American Congress of Rehabilitation Medicine criteria compared with healthy controls.

SUBMITTER: Lui YW

PROVIDER: S-EPMC4180485 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BACKGROUND AND PURPOSE:Mild traumatic brain injury is a leading cause of death and disability worldwide with 42 million cases reported annually, increasing the need to understand the underlying pathophysiology because this could help guide the development of targeted therapy. White matter, particularly the corpus callosum, is susceptible to injury. Animal models suggest stretch-induced mechanoporation of the axonal membrane resulting in ionic shifts and altered sodium ion distribution. The purpose of this study was to compare the distribution of total sodium concentration in the corpus callosum between patients with mild traumatic brain injury and controls using sodium (23Na) MR imaging. MATERIALS AND METHODS:Eleven patients with a history of mild traumatic brain injury and 10 age- and sex-matched controls underwent sodium (23Na) MR imaging using a 3T scanner. Total sodium concentration was measured in the genu, body, and splenium of the corpus callosum with 5-mm ROIs; total sodium concentration of the genu-to-splenium ratio was calculated and compared between patients and controls. RESULTS:Higher total sodium concentration in the genu (49.28 versus 43.29 mmol/L, P = .01) and lower total sodium concentration in the splenium (which was not statistically significant; 38.35 versus 44.06 mmol/L, P = .08) was seen in patients with mild traumatic brain injury compared with controls. The ratio of genu total sodium concentration to splenium total sodium concentration was also higher in patients with mild traumatic brain injury (1.3 versus 1.01, P = .001). CONCLUSIONS:Complex differences are seen in callosal total sodium concentration in symptomatic patients with mild traumatic brain injury, supporting the notion of ionic dysfunction in the pathogenesis of mild traumatic brain injury. The total sodium concentration appears to be altered beyond the immediate postinjury phase, and further work is needed to understand the relationship to persistent symptoms and outcome.

Project description:Mild traumatic brain injury (mTBI) is a major public health concern that can result in a broad spectrum of short-term and long-term symptoms. Recently, machine learning (ML) algorithms have been used in neuroscience research for diagnostics and prognostic assessment of brain disorders. The present study aimed to develop an automatic classifier to distinguish patients suffering from chronic mTBI from healthy controls (HCs) utilizing multilevel metrics of resting-state functional magnetic resonance imaging (rs-fMRI). Sixty mTBI patients and forty HCs were enrolled and allocated to training and testing datasets with a ratio of 80:20. Several rs-fMRI metrics including fractional amplitude of low-frequency fluctuation (fALFF), regional homogeneity (ReHo), degree centrality (DC), voxel-mirrored homotopic connectivity (VMHC), functional connectivity strength (FCS), and seed-based FC were generated from two main analytical categories: local measures and network measures. Statistical two-sample t-test was employed comparing between mTBI and HCs groups. Then, for each rs-fMRI metric the features were selected extracting the mean values from the clusters showing significant differences. Finally, the support vector machine (SVM) models based on separate and multilevel metrics were built and the performance of the classifiers were assessed using five-fold cross-validation and via the area under the receiver operating characteristic curve (AUC). Feature importance was estimated using Shapley additive explanation (SHAP) values. Among local measures, the range of AUC was 86.67-100% and the optimal SVM model was obtained based on combined multilevel rs-fMRI metrics and DC as a separate model with AUC of 100%. Among network measures, the range of AUC was 80.42-93.33% and the optimal SVM model was obtained based on the combined multilevel seed-based FC metrics. The SHAP analysis revealed the DC value in the left postcentral and seed-based FC value between the motor ventral network and right superior temporal as the most important local and network features with the greatest contribution to the classification models. Our findings demonstrated that different rs-fMRI metrics can provide complementary information for classifying patients suffering from chronic mTBI. Moreover, we showed that ML approach is a promising tool for detecting patients with mTBI and might serve as potential imaging biomarker to identify patients at individual level.Clinical trial registration[clinicaltrials.gov], identifier [NCT03241732].

Project description:To investigate longitudinal changes in global and regional brain volume in patients 1 year after mild traumatic brain injury (MTBI) and to correlate such changes with clinical and neurocognitive metrics.This institutional review board-approved study was HIPAA compliant. Twenty-eight patients with MTBI (with 19 followed up at 1 year) with posttraumatic symptoms after injury and 22 matched control subjects (with 12 followed up at 1 year) were enrolled. Automated segmentation of brain regions to compute regional gray matter (GM) and white matter (WM) volumes was performed by using three-dimensional T1-weighted 3.0-T magnetic resonance imaging, and results were correlated with clinical metrics. Pearson and Spearman rank correlation coefficients were computed between longitudinal brain volume and neurocognitive scores, as well as clinical metrics, over the course of the follow-up period.One year after MTBI, there was measurable global brain atrophy, larger than that in control subjects. The anterior cingulate WM bilaterally and the left cingulate gyrus isthmus WM, as well as the right precuneal GM, showed significant decreases in regional volume in patients with MTBI over the 1st year after injury (corrected P < .05); this was confirmed by means of cross-sectional comparison with data in control subjects (corrected P < .05). Left and right rostral anterior cingulum WM volume loss correlated with changes in neurocognitive measures of memory (r = 0.65, P = .005) and attention (r = 0.60, P = .01). At 1-year follow-up, WM volume in the left cingulate gyrus isthmus correlated with clinical scores of anxiety (Spearman rank correlation r = -0.68, P = .007) and postconcussive symptoms (Spearman rank correlation r = -0.65, P = .01).These observations demonstrate structural changes to the brain 1 year after injury after a single concussive episode. Regional brain atrophy is not exclusive to moderate and severe traumatic brain injury but may be seen after mild injury. In particular, the anterior part of the cingulum and the cingulate gyrus isthmus, as well as the precuneal GM, may be distinctively vulnerable 1 year after MTBI.

Project description:ImportanceTraumatic brain injury (TBI) affects millions of people in the US each year. Most patients with TBI seen in emergency departments (EDs) have a Glasgow Coma Scale (GCS) score of 15 and a head computed tomography (CT) scan showing no acute intracranial traumatic injury (negative head CT scan), yet the short-term and long-term functional outcomes of this subset of patients remain unclear.ObjectiveTo describe the 2-week and 6-month recovery outcomes in a cohort of patients with mild TBI with a GCS score of 15 and a negative head CT scan.Design, setting, and participantsThis cohort study analyzed participants who were enrolled from January 1, 2014, to December 31, 2018, in the Transforming Research and Clinical Knowledge in Traumatic Brain Injury (TRACK-TBI) study, a prospective, observational cohort study of patients with TBI that was conducted in EDs of 18 level I trauma centers in urban areas. Of the total 2697 participants in the TRACK-TBI study, 991 had a GCS score of 15 and negative head CT scan and were eligible for inclusion in this analysis. Data were analyzed from September 1, 2021, to May 30, 2022.Main outcomes and measuresThe primary outcome was the Glasgow Outcome Scale-Extended (GOS-E) score, which was stratified according to functional recovery (GOS-E score, 8) vs incomplete recovery (GOS-E score, <8), at 2 weeks and 6 months after the injury. The secondary outcome was severity of mild TBI-related symptoms assessed by the Rivermead Post Concussion Symptoms Questionnaire (RPQ) total score.ResultsA total of 991 participants (mean [SD] age, 38.5 [15.8] years; 631 male individuals [64%]) were included. Of these participants, 751 (76%) were followed up at 2 weeks after the injury: 204 (27%) had a GOS-E score of 8 (functional recovery), and 547 (73%) had a GOS-E scores less than 8 (incomplete recovery). Of 659 participants (66%) followed up at 6 months after the injury, 287 (44%) had functional recovery and 372 (56%) had incomplete recovery. Most participants with incomplete recovery reported that they had not returned to baseline or preinjury life (88% [479 of 546]; 95% CI, 85%-90%). Mean RPQ score was 16 (95% CI, 14-18; P < .001) points lower at 2 weeks (7 vs 23) and 18 (95% CI, 16-20; P < .001) points lower at 6 months (4 vs 22) in participants with a GOS-E score of 8 compared with those with a GOS-E score less than 8.Conclusions and relevanceThis study found that most participants with a GCS score of 15 and negative head CT scan reported incomplete recovery at 2 weeks and 6 months after their injury. The findings suggest that emergency department clinicians should recommend 2-week follow-up visits for these patients to identify those with incomplete recovery and to facilitate their rehabilitation.