Detection of Mild Traumatic Brain Injury by Machine Learning Classification Using Resting State Functional Network Connectivity and Fractional Anisotropy

Functional Connectivity Patterns Following Mild Traumatic Brain Injury and the Association With Longitudinal Cognitive Function

Functional magnetic resonance imaging (fMRI) has revealed subtle neuroplastic changes in brain networks following mild traumatic brain injury (mTBI), even when standard clinical imaging fails to detect abnormalities. However, prior findings have been inconsistent, in part due to methodological differences and high researcher degrees of freedom in region-based analyses, which often rely on predefined hypotheses and overlook complex, distributed connectivity patterns. Here, we apply an unbiased, data-driven multi-voxel pattern analysis (MVPA) to examine whole-brain functional connectivity differences in a large cohort of individuals with acute mTBI. Unlike conventional statistical approaches, MVPA enables a data-driven analysis of brain-wide connectivity patterns without requiring prior assumptions about the location or nature of abnormalities, allowing for the identification of the most informative features. This approach provides an exploratory characterization of whole-brain functional connectivity patterns and their relationship with cognitive recovery, offering new insights into the neural mechanisms underlying post-injury outcomes. A total of 265 adults (87 women) between 18 and 83 years old with Glasgow Coma Scale (GCS) scores of 13-15 were included in this analysis. Two replicate samples (n = 165, n = 155), with similar demographic characteristics, were also included. Data were collected as part of the prospective multi-center Transforming Research and Clinical Knowledge in TBI (TRACK-TBI). The goal of this study was to assess whole-brain functional connectivity patterns using fc-MVPA and post hoc seed-to-voxel analyses in a large, well-characterized sample to determine if changes in functional connectivity can differentiate subacute mTBI (within 2 weeks of injury) from a matched group of orthopedic control subjects (n = 49). Additionally, we aimed to investigate whether these connectivity patterns were linked to cognitive performance at 2 weeks, 6 months, and 12 months post-injury to better understand cognitive trajectories and recovery over time in individuals with mTBI. Voxel-to-voxel functional connectivity across the entire connectome revealed significant differences between TBI and no TBI in the functional connectivity patterns of 8 clusters (p-voxel < 0.001, FEW cluster-level p < 0.05) (k > 40, Fmax = 15.36), including right occipital cortex, anterior cingulate gyrus, inferior and middle temporal gyrus, right thalamus, left cerebellum, and the bilateral frontal pole. These clusters belong mainly to the visual network (VIS), frontoparietal network (FPN), default mode network (DMN) and limbic network (LIM). Post hoc characterization of each significant cluster revealed by MVPA using seed-to-voxel analysis showed a mixed pattern of connectivity between relevant networks and subcortico-cortical connections. After connectivity characterization, visual-motor skills assessed with Trail Making Test (TMT) A were significantly associated with the increased anticorrelation between the inferior temporal cortex and the bilateral occipital pole (FPN-VIS connectivity), along with decreased anticorrelations between the cerebellum and extensive areas of the somatomotor network (SMN) over 12 months post injury. Additionally, hypoconnectivity between the frontal pole (LIM) and anterior cingulate gyrus (salience network [SAL]) was associated with better executive functions performance measured by TMT-B over 12 months post-mTBI. In our study examining neuroplastic changes following TBI across the entire voxel-to-voxel functional connectome, we identified significant differences in the functional connectivity patterns of several regions known to be particularly vulnerable to injury mechanisms. Our findings highlight the complex and compensatory nature of brain network alterations after mTBI, suggesting both detrimental and adaptive changes in connectivity that affect cognitive functions. Consequently, our study provides novel evidence that specific brain networks and regions are particularly susceptible to functional connectivity changes during the acute stages of mTBI, which are related to cognitive recovery post-injury.

An iterative ROC procedure identifies white matter tracts diagnostic for traumatic brain injury: an exploratory analysis in U.S. Veterans

OBJECTIVE

Understanding the pathophysiology of traumatic brain injury (TBI) is crucial for effectively managing care. Diffusion tensor imaging (DTI) is an MRI technology that evaluates TBI pathology in brain white matter. However, DTI analysis generates numerous measures. Choosing between them remains an obstacle to clinical translation. In this study, we leveraged an iterative receiver operating characteristic (ROC) analysis to examine white matter tracts in a group of 380 Veterans, consisting of TBI (n = 243) and non-TBI patients (n = 137).

METHOD

For each participant, we obtained a whole brain tractography and extracted DTI measures from 50 tracts. The ROC analyzed these variables and produced decision trees of tracts diagnostic for TBI. We expanded our findings by applying jackknife resampling. This procedure removed potential outliers and yielded tracts not observed in the initial ROCs. Finally, we used logistic regression to confirm the tracts predicted TBI status.

RESULTS

Our results indicate ROC can identify tracts diagnostic for TBI. We also found that groups of tracts are more predictive than any single one.

CONCLUSIONS

These analyses show that ROC is a useful tool for exploring large, multivariate datasets and may inform the design of clinical algorithms.

The Value of Clinical Decision Support in Healthcare: A Focus on Screening and Early Detection

In a rapidly changing technology landscape, "Clinical Decision Support" (CDS) has become an important tool to improve patient management. CDS systems offer medical professionals new insights to improve diagnostic accuracy, therapy planning, and personalized treatment. In addition, CDS systems provide cost-effective options to augment conventional screening for secondary prevention. This review aims to (i) describe the purpose and mechanisms of CDS systems, (ii) discuss different entities of algorithms, (iii) highlight quality features, and (iv) discuss challenges and limitations of CDS in clinical practice. Furthermore, we (v) describe contemporary algorithms in oncology, acute care, cardiology, and nephrology. In particular, we consolidate research on algorithms across diseases that imply a significant disease and economic burden, such as lung cancer, colorectal cancer, hepatocellular cancer, coronary artery disease, traumatic brain injury, sepsis, and chronic kidney disease.

Relationship Between Structural and Functional Network Connectivity Changes for Patients With Traumatic Brain Injury and Chronic Health Symptoms

Combination of structural and functional brain connectivity methods provides a more complete and effective avenue into the investigation of cortical network responses to traumatic brain injury (TBI) and subtle alterations in brain connectivity associated with TBI. Structural connectivity (SC) can be measured using diffusion tensor imaging to evaluate white matter integrity, whereas functional connectivity (FC) can be studied by examining functional correlations within or between functional networks. In this study, the alterations of SC and FC were assessed for TBI patients, with and without chronic symptoms (TBIcs/TBIncs), compared with a healthy control group (CG). The correlation between global SC and FC was significantly increased for both TBI groups compared with CG. SC was significantly lower in the TBIcs group compared with CG, and FC changes were seen in the TBIncs group compared with CG. When comparing TBI groups, FC differences were observed in the TBIcs group compared with the TBIncs group. These observations show that the presence of chronic symptoms is associated with a distinct pattern of SC and FC changes including the atrophy of the SC and a mixture of functional hypoconnectivity and hyperconnectivity, as well as loss of segregation of functional networks.

Clinical Insights Into Default Mode Network Abnormalities in Mild Traumatic Brain Injury: Unraveling Axonal Injury Through Functional, Structural, and Molecular Analyses

BACKGROUND

Mild traumatic brain injury (mTBI) frequently results in persistent cognitive, emotional, and functional impairments, closely linked to disruptions in the default mode network (DMN). Understanding the mechanisms driving these network abnormalities is critical for advancing diagnostic and therapeutic strategies.

METHODS

This study adopted a multimodal approach, combining functional connectivity (FC) analysis, diffusion tensor imaging (DTI), and gene expression profiling to investigate DMN disruptions in mTBI. A primary focus was placed on the middle cingulate cortex (MCC), a region consistently identified with increased connectivity. We explored the structural and molecular changes underlying this phenomenon. Receiver operating characteristic (ROC) curve analysis was utilized to assess the diagnostic potential of DTI-derived metrics, while white matter tractography was employed to explore structural connectivity between the MCC and the dorsolateral prefrontal cortex (DLPFC).

RESULTS

Our findings revealed significant disruptions in DMN connectivity, with the MCC prominently involved in mTBI pathology. DTI analyses identified pronounced axonal injury in the MCC, characterized by decreased fractional anisotropy (FA) and axial diffusivity (AD), alongside increased isotropy (ISO), indicating compromised white matter integrity and diffuse axonal injury. Gene expression profiling revealed the upregulation of pathways related to synaptic transmission, ion channel regulation, and axonal injury response. ROC analysis demonstrated that ISO serves as a particularly effective biomarker for mTBI, showing high diagnostic accuracy (AUC = 0.871). White matter tractography further confirmed strong structural connectivity between the MCC and the DLPFC, identifying potential therapeutic targets for neuromodulation.

CONCLUSION

This study provides robust evidence that diffuse axonal injury plays a pivotal role in DMN abnormalities observed in mTBI. The integration of FC, DTI, and gene expression profiling offers a comprehensive framework for understanding mTBI's impact on brain networks. Our findings also highlight the DLPFC as a promising target for therapeutic interventions aimed at addressing cognitive and emotional deficits associated with mTBI.

Drivers of resting-state fMRI heterogeneity in traumatic brain injury across injury characteristics and imaging methods: a systematic review and semiquantitative analysis

Traumatic brain injury (TBI) is common and costly. Although neuroimaging modalities such as resting-state functional MRI (rsfMRI) promise to differentiate injured from healthy brains and prognosticate long-term outcomes, the field suffers from heterogeneous findings. To assess whether this heterogeneity stems from variability in the TBI populations studied or the imaging methods used, and to determine whether a consensus exists in this literature, we performed the first systematic review of studies comparing rsfMRI functional connectivity (FC) in patients with TBI to matched controls for seven canonical brain networks across injury severity, age, chronicity, population type, and various imaging methods. Searching PubMed, Web of Science, Google Scholar, and ScienceDirect, 1,105 manuscripts were identified, 50 fulfilling our criteria. Across these manuscripts, 179 comparisons were reported between a total of 1,397 patients with TBI and 1,179 matched controls. Collapsing across injury characteristics, imaging methods, and networks, there were roughly equal significant to null findings and increased to decreased connectivity differences reported. Whereas most factors did not explain these mixed findings, stratifying across severity and chronicity, separately, showed a trend of increased connectivity at higher severities and greater chronicities of TBI. Among methodological factors, studies were more likely to find connectivity differences when scans were longer than 360 s, custom image processing pipelines were used, and when patients kept their eyes open versus closed during scans. We offer guidelines to address this variability, focusing on aspects of study design and rsfMRI acquisition to move the field toward reproducible results with greater potential for clinical translation.

Aberrant dynamic functional network connectivity in patients with diffuse axonal injury

Diffuse axonal injury (DAI) results in aberrant functional connectivity and is significantly linked to cognitive impairment. Nevertheless, the network mechanisms influencing neurocognitive function following DAI remain unclear. This study aimed to examine the characteristics of static and dynamic functional network connectivity (FNC) in patients with DAI. Resting-state functional magnetic resonance imaging data were collected from 26 patients with DAI and 27 healthy controls. Resting-state networks were extracted using independent component analysis. We evaluated the connectivity strength through spatial maps and static FNC, and then further dynamic properties were identified using a sliding time-window approach and k-means clustering, and investigated their associations with clinical variables. Patients with DAI showed stronger intra-network spatial maps in the default mode network and subcortical network than healthy controls, but static inter-network functional connectivity remained stable. Furthermore, three recurring states for dynamic connectivity were identified in all participants, and state 1 occurred most frequently in patients with DAI and exhibited higher fractional time, and as well as longer mean dwell time, which was positively associated with MMSE scores. Meanwhile, patients with DAI exhibited mostly increased functional connectivity strength of dynamic FNC in all states, particularly within the default mode network and visual network. These findings suggest that patients with DAI are characterized by altered dynamic FNC and temporal properties, which provide distinct complementary information different from static functional connectivity, and new insights into the neural pathophysiology of DAI associated with cognitive impairment.

Safety and efficacy of simple training protocol in patients after mild traumatic brain injury

AIMS

Mild Traumatic Brain Injury (mTBI) is the most common type of craniocerebral injury. Proper management appears to be a key factor in preventing post-concussion syndrome. The aim of this prospective study was to evaluate the effect and safety of selected training protocol in patients after mTBI.

METHODS

This was a prospective study that included 25 patients with mTBI and 25 matched healthy controls. Assessments were performed in two sessions and included a post-concussion symptoms questionnaire, battery of neurocognitive tests, and magnetic resonance with tractography. Participants were divided into two groups: a passive subgroup with no specific recommendations and an active subgroup with simple physical and cognitive training.

RESULTS

The training program with slightly higher initial physical and cognitive loads was well tolerated and was harmless according to the noninferiority test. The tractography showed overall temporal posttraumatic changes in the brain. The predictive model was able to distinguish between patients and controls in the first (AUC=0.807) and second (AUC=0.652) sessions. In general, tractography had an overall predictive dominance of measures.

CONCLUSION

The results from our study objectively point to the safety of our chosen training protocol, simultaneously with the signs of slight benefits in specific cognitive domains. The study also showed the capability of machine learning and predictive models in mTBI patient recognition.

Potential Applications and Ethical Considerations for Artificial Intelligence in Traumatic Brain Injury Management

Artificial intelligence (AI) systems have emerged as promising tools for rapidly identifying patterns in large amounts of healthcare data to help guide clinical decision making, as well as to assist with medical education and the planning of research studies. Accumulating evidence suggests AI techniques may be particularly useful for aiding the diagnosis and clinical management of traumatic brain injury (TBI)-a considerably heterogeneous neurologic condition that can be challenging to detect and treat. However, important methodological and ethical concerns with the use of AI in medicine necessitate close monitoring and regulation of these techniques as advancements continue. The purpose of this narrative review is to provide an overview of common AI techniques in medical research and describe recent studies on the possible clinical applications of AI in the context of TBI. Finally, the review describes the ethical challenges with the use of AI in medicine, as well as guidelines from the White House, the Department of Defense (DOD), the National Academies of Sciences, Engineering, and Medicine (NASEM), and other organizations on the appropriate uses of AI in research.

Changes in cerebral vascular reactivity following mild repetitive head injury in awake rats: modeling the human experience

The changes in brain function in response to mild head injury are usually subtle and go undetected. Physiological biomarkers would aid in the early diagnosis of mild head injury. In this study we used hypercapnia to follow changes in cerebral vascular reactivity after repetitive mild head injury. We hypothesized head injury would reduce vascular reactivity. Rats were maintained on a reverse light-dark cycle and head impacted daily at 24 h intervals over three days. All head impacts were delivered while rats were fully awake under red light illumination. There was no neuroradiological evidence of brain damage. After the 3rd impact rats were exposed to 5% CO2 and imaged for changes in BOLD signal. All imaging was done while rats were awake without the confound of anesthesia. The data were registered to a 3D MRI rat atlas with 171 segmented brain areas providing site specific information on vascular reactivity. The changes in vascular reactivity were not uniform across the brain. The prefrontal cortex, somatosensory cortex and basal ganglia showed the hypothesized decrease in vascular reactivity while the cerebellum, thalamus, brainstem, and olfactory system showed an increase in BOLD signal to hypercapnia.

Translational Connectomics: overview of machine learning in macroscale Connectomics for clinical insights

Connectomics is a neuroscience paradigm focused on noninvasively mapping highly intricate and organized networks of neurons. The advent of neuroimaging has led to extensive mapping of the brain functional and structural connectome on a macroscale level through modalities such as functional and diffusion MRI. In parallel, the healthcare field has witnessed a surge in the application of machine learning and artificial intelligence for diagnostics, especially in imaging. While reviews covering machine learn ing and macroscale connectomics exist for specific disorders, none provide an overview that captures their evolving role, especially through the lens of clinical application and translation. The applications include understanding disorders, classification, identifying neuroimaging biomarkers, assessing severity, predicting outcomes and intervention response, identifying potential targets for brain stimulation, and evaluating the effects of stimulation intervention on the brain and connectome mapping in patients before neurosurgery. The covered studies span neurodegenerative, neurodevelopmental, neuropsychiatric, and neurological disorders. Along with applications, the review provides a brief of common ML methods to set context. Conjointly, limitations in ML studies within connectomics and strategies to mitigate them have been covered.

Identification and Connectomic Profiling of Concussion Using Bayesian Machine Learning

Accurate early diagnosis of concussion is useful to prevent sequelae and improve neurocognitive outcomes. Early after head impact, concussion diagnosis may be doubtful in persons whose neurological, neuroradiological, and/or neurocognitive examinations are equivocal. Such individuals can benefit from novel accurate assessments that complement clinical diagnostics. We introduce a Bayesian machine learning classifier to identify concussion through cortico-cortical connectome mapping from magnetic resonance imaging in persons with quasi-normal cognition and without neuroradiological findings. Classifier features are generated from connectivity matrices specifying the mean fractional anisotropy of white matter connections linking brain structures. Each connection's saliency to classification was quantified by training individual classifier instantiations using a single feature type. The classifier was tested on a discovery sample of 92 healthy controls (HCs; 26 females, age μ ± σ: 39.8 ± 15.5 years) and 471 adult mTBI patients (158 females, age μ ± σ: 38.4 ± 5.9 years). Results were replicated in an independent validation sample of 256 HCs (149 females, age μ ± σ: 55.3 ± 12.1 years) and 126 patients with concussion (46 females, age μ ± σ: 39.0 ± 17.7 years). Classifier accuracy exceeds 99% in both samples, suggesting robust generalizability to new samples. Notably, 13 bilateral cortico-cortical connection pairs predict diagnostic status with accuracy exceeding 99% in both discovery and validation samples. Many such connection pairs are between prefrontal cortex structures, fronto-limbic and fronto-subcortical structures, and occipito-temporal structures in the ventral ("what") visual stream. This and related connectivity form a highly salient network of brain connections that is particularly vulnerable to concussion. Because these connections are important in mediating cognitive control, memory, and attention, our findings explain the high frequency of cognitive disturbances after concussion. Our classifier was trained and validated on concussed participants with cognitive profiles very similar to those of HCs. This suggests that the classifier can complement current diagnostics by providing independent information in clinical contexts where patients have quasi-normal cognition but where concussion diagnosis stands to benefit from additional evidence.

Usefulness of Artificial Intelligence in Traumatic Brain Injury: A Bibliometric Analysis and Mini-review

BACKGROUND

Traumatic brain injury (TBI) has become a major source of disability worldwide, increasing the interest in algorithms that use artificial intelligence (AI) to optimize the interpretation of imaging studies, prognosis estimation, and critical care issues. In this study we present a bibliometric analysis and mini-review on the main uses that have been developed for TBI in AI.

METHODS

The results informing this review come from a Scopus database search as of April 15, 2023. The bibliometric analysis was carried out via the mapping bibliographic metrics method. Knowledge mapping was made in the VOSviewer software (V1.6.18), analyzing the "link strength" of networks based on co-occurrence of key words, countries co-authorship, and co-cited authors. In the mini-review section, we highlight the main findings and contributions of the studies.

RESULTS

A total of 495 scientific publications were identified from 2000 to 2023, with 9262 citations published since 2013. Among the 160 journals identified, The Journal of Neurotrauma, Frontiers in Neurology, and PLOS ONE were those with the greatest number of publications. The most frequently co-occurring key words were: "machine learning", "deep learning", "magnetic resonance imaging", and "intracranial pressure". The United States accounted for more collaborations than any other country, followed by United Kingdom and China. Four co-citation author clusters were found, and the top 20 papers were divided into reviews and original articles.

CONCLUSIONS

AI has become a relevant research field in TBI during the last 20 years, demonstrating great potential in imaging, but a more modest performance for prognostic estimation and neuromonitoring.

Dynamic Functional Network Connectivity Clustering and Harmonization Evaluation Metric

Neuroscience studies have improved significantly through multi-site datasets. Brain dynamic functional network connectivity (dFNC) often uses clustering to represent various patterns of functional connectivity that repeat with time. Data collected from different sources combined may be affected by site effects. This can be mitigated by harmonizing the dataset before doing additional analysis. Here, we explore the impact of ComBat harmonization on dFNC data collected from two different mild traumatic brain injury (mTBI) studies. Our research showed that site effects significantly influence the formation of dFNC states. We developed a new method to measure the distribution of samples among the clusters based on site effects in clustering called an 'Inclusivity' model. With this inclusivity model, we evaluate how site effects influence the clustering of dFNC states before and after data harmonization.

Functional Connectivity Changes on Resting-State fMRI after Mild Traumatic Brain Injury: A Systematic Review

BACKGROUND

Mild traumatic brain injury is theorized to cause widespread functional changes to the brain. Resting-state fMRI may be able to measure functional connectivity changes after traumatic brain injury, but resting-state fMRI studies are heterogeneous, using numerous techniques to study ROIs across various resting-state networks.

PURPOSE

We systematically reviewed the literature to ascertain whether adult patients who have experienced mild traumatic brain injury show consistent functional connectivity changes on resting-state -fMRI, compared with healthy patients.

DATA SOURCES

We used 5 databases (PubMed, EMBASE, Cochrane Central, Scopus, Web of Science).

STUDY SELECTION

Five databases (PubMed, EMBASE, Cochrane Central, Scopus, and Web of Science) were searched for research published since 2010. Search strategies used keywords of "functional MR imaging" and "mild traumatic brain injury" as well as related terms. All results were screened at the abstract and title levels by 4 reviewers according to predefined inclusion and exclusion criteria. For full-text inclusion, each study was evaluated independently by 2 reviewers, with discordant screening settled by consensus.

DATA ANALYSIS

Data regarding article characteristics, cohort demographics, fMRI scan parameters, data analysis processing software, atlas used, data characteristics, and statistical analysis information were extracted.

DATA SYNTHESIS

Across 66 studies, 80 areas were analyzed 239 times for at least 1 time point, most commonly using independent component analysis. The most analyzed areas and networks were the whole brain, the default mode network, and the salience network. Reported functional connectivity changes varied, though there may be a slight trend toward decreased whole-brain functional connectivity within 1 month of traumatic brain injury and there may be differences based on the time since injury.

LIMITATIONS

Studies of military, sports-related traumatic brain injury, and pediatric patients were excluded. Due to the high number of relevant studies and data heterogeneity, we could not be as granular in the analysis as we would have liked.

CONCLUSIONS

Reported functional connectivity changes varied, even within the same region and network, at least partially reflecting differences in technical parameters, preprocessing software, and analysis methods as well as probable differences in individual injury. There is a need for novel rs-fMRI techniques that better capture subject-specific functional connectivity changes.

A Comprehensive and Broad Approach to Resting-State Functional Connectivity in Adult Patients with Mild Traumatic Brain Injury

BACKGROUND AND PURPOSE

Several recent works using resting-state fMRI suggest possible alterations of resting-state functional connectivity after mild traumatic brain injury. However, the literature is plagued by various analysis approaches and small study cohorts, resulting in an inconsistent array of reported findings. In this study, we aimed to investigate differences in whole-brain resting-state functional connectivity between adult patients with mild traumatic brain injury within 1 month of injury and healthy control subjects using several comprehensive resting-state functional connectivity measurement methods and analyses.

MATERIALS AND METHODS

A total of 123 subjects (72 patients with mild traumatic brain injury and 51 healthy controls) were included. A standard fMRI preprocessing pipeline was used. ROI/seed-based analyses were conducted using 4 standard brain parcellation methods, and the independent component analysis method was applied to measure resting-state functional connectivity. The fractional amplitude of low-frequency fluctuations was also measured. Group comparisons were performed on all measurements with appropriate whole-brain multilevel statistical analysis and correction.

RESULTS

There were no significant differences in age, sex, education, and hand preference between groups as well as no significant correlation between all measurements and these potential confounders. We found that each resting-state functional connectivity measurement revealed various regions or connections that were different between groups. However, after we corrected for multiple comparisons, the results showed no statistically significant differences between groups in terms of resting-state functional connectivity across methods and analyses.

CONCLUSIONS

Although previous studies point to multiple regions and networks as possible mild traumatic brain injury biomarkers, this study shows that the effect of mild injury on brain resting-state functional connectivity has not survived after rigorous statistical correction. A further study using subject-level connectivity analyses may be necessary due to both subtle and variable effects of mild traumatic brain injury on brain functional connectivity across individuals.

A Siamese Convolutional Neural Network for Identifying Mild Traumatic Brain Injury and Predicting Recovery

Timely diagnosis of mild traumatic brain injury (mTBI) remains challenging due to the rapid recovery of acute symptoms and the absence of evidence of injury in static neuroimaging scans. Furthermore, while longitudinal tracking of mTBI is essential in understanding how the diseases progresses/regresses over time for enhancing personalized patient care, a standardized approach for this purpose is not yet available. Recent functional neuroimaging studies have provided evidence of brain function alterations following mTBI, suggesting mTBI-detection models can be built based on these changes. Most of these models, however, rely on manual feature engineering, but the optimal set of features for detecting mTBI may be unknown. Data-driven approaches, on the other hand, may uncover hidden relationships in an automated manner, making them suitable for the problem of mTBI detection. This paper presents a data-driven framework based on Siamese Convolutional Neural Network (SCNN) to detect mTBI and to monitor the recovery state from mTBI over time. The proposed framework is tested on the cortical images of Thy1-GCaMP6s mice, obtained via widefield calcium imaging, acquired in a longitudinal study. Results show that the proposed model achieves a classification accuracy of 96.5%. To track the state of the injured brain over time, a reference distance map is constructed, which together with the SCNN model, are employed to assess the recovery state in subsequent sessions after injury, revealing that the recovery progress varies among subjects. The promising results of this work suggest that a similar approach could be potentially applicable for monitoring recovery from mTBI, in humans.

Predicting shock-induced cavitation using machine learning: implications for blast-injury models

While cavitation has been suspected as a mechanism of blast-induced traumatic brain injury (bTBI) for a number of years, this phenomenon remains difficult to study due to the current inability to measure cavitation in vivo. Therefore, numerical simulations are often implemented to study cavitation in the brain and surrounding fluids after blast exposure. However, these simulations need to be validated with the results from cavitation experiments. Machine learning algorithms have not generally been applied to study blast injury or biological cavitation models. However, such algorithms have concrete measures for optimization using fewer parameters than those of finite element or fluid dynamics models. Thus, machine learning algorithms are a viable option for predicting cavitation behavior from experiments and numerical simulations. This paper compares the ability of two machine learning algorithms, k-nearest neighbor (kNN) and support vector machine (SVM), to predict shock-induced cavitation behavior. The machine learning models were trained and validated with experimental data from a three-dimensional shock tube model, and it has been shown that the algorithms could predict the number of cavitation bubbles produced at a given temperature with good accuracy. This study demonstrates the potential utility of machine learning in studying shock-induced cavitation for applications in blast injury research.

Deep learning-based multimodality classification of chronic mild traumatic brain injury using resting-state functional MRI and PET imaging

Mild traumatic brain injury (mTBI) is a public health concern. The present study aimed to develop an automatic classifier to distinguish between patients with chronic mTBI (n = 83) and healthy controls (HCs) (n = 40). Resting-state functional MRI (rs-fMRI) and positron emission tomography (PET) imaging were acquired from the subjects. We proposed a novel deep-learning-based framework, including an autoencoder (AE), to extract high-level latent and rectified linear unit (ReLU) and sigmoid activation functions. Single and multimodality algorithms integrating multiple rs-fMRI metrics and PET data were developed. We hypothesized that combining different imaging modalities provides complementary information and improves classification performance. Additionally, a novel data interpretation approach was utilized to identify top-performing features learned by the AEs. Our method delivered a classification accuracy within the range of 79-91.67% for single neuroimaging modalities. However, the performance of classification improved to 95.83%, thereby employing the multimodality model. The models have identified several brain regions located in the default mode network, sensorimotor network, visual cortex, cerebellum, and limbic system as the most discriminative features. We suggest that this approach could be extended to the objective biomarkers predicting mTBI in clinical settings.

Dynamic Functional Connectivity in Pediatric Mild Traumatic Brain Injury

Resting-state fMRI can be used to identify recurrent oscillatory patterns of functional connectivity within the human brain, also known as dynamic brain states. Alterations in dynamic brain states are highly likely to occur following pediatric mild traumatic brain injury (pmTBI) due to the active developmental changes. The current study used resting-state fMRI to investigate dynamic brain states in 200 patients with pmTBI (ages 8-18 years, median = 14 years) at the subacute (∼1-week post-injury) and early chronic (∼ 4 months post-injury) stages, and in 179 age- and sex-matched healthy controls (HC). A k-means clustering analysis was applied to the dominant time-varying phase coherence patterns to obtain dynamic brain states. In addition, correlations between brain signals were computed as measures of static functional connectivity. Dynamic connectivity analyses showed that patients with pmTBI spend less time in a frontotemporal default mode/limbic brain state, with no evidence of change as a function of recovery post-injury. Consistent with models showing traumatic strain convergence in deep grey matter and midline regions, static interhemispheric connectivity was affected between the left and right precuneus and thalamus, and between the right supplementary motor area and contralateral cerebellum. Changes in static or dynamic connectivity were not related to symptom burden or injury severity measures, such as loss of consciousness and post-traumatic amnesia. In aggregate, our study shows that brain dynamics are altered up to 4 months after pmTBI, in brain areas that are known to be vulnerable to TBI. Future longitudinal studies are warranted to examine the significance of our findings in terms of long-term neurodevelopment.

Clinical performance of a multiparametric MRI-based post concussive syndrome index

Introduction

Diffusion Tensor Imaging (DTI) has revealed measurable changes in the brains of patients with persistent post-concussive syndrome (PCS). Because of inconsistent results in univariate DTI metrics among patients with mild traumatic brain injury (mTBI), there is currently no single objective and reliable MRI index for clinical decision-making in patients with PCS.

Purpose

This study aimed to evaluate the performance of a newly developed PCS Index (PCSI) derived from machine learning of multiparametric magnetic resonance imaging (MRI) data to classify and differentiate subjects with mTBI and PCS history from those without a history of mTBI.

Materials and methods

Data were retrospectively extracted from 139 patients aged between 18 and 60 years with PCS who underwent MRI examinations at 2 weeks to 1-year post-mTBI, as well as from 336 subjects without a history of head trauma. The performance of the PCS Index was assessed by comparing 69 patients with a clinical diagnosis of PCS with 264 control subjects. The PCSI values for patients with PCS were compared based on the mechanism of injury, time interval from injury to MRI examination, sex, history of prior concussion, loss of consciousness, and reported symptoms.

Results

Injured patients had a mean PCSI value of 0.57, compared to the control group, which had a mean PCSI value of 0.12 (p = 8.42e-23) with accuracy of 88%, sensitivity of 64%, and specificity of 95%, respectively. No statistically significant differences were found in the PCSI values when comparing the mechanism of injury, sex, or loss of consciousness.

Conclusion

The PCSI for individuals aged between 18 and 60 years was able to accurately identify patients with post-concussive injuries from 2 weeks to 1-year post-mTBI and differentiate them from the controls. The results of this study suggest that multiparametric MRI-based PCSI has great potential as an objective clinical tool to support the diagnosis, treatment, and follow-up care of patients with post-concussive syndrome. Further research is required to investigate the replicability of this method using other types of clinical MRI scanners.

Advances in neuroproteomics for neurotrauma: unraveling insights for personalized medicine and future prospects

Neuroproteomics, an emerging field at the intersection of neuroscience and proteomics, has garnered significant attention in the context of neurotrauma research. Neuroproteomics involves the quantitative and qualitative analysis of nervous system components, essential for understanding the dynamic events involved in the vast areas of neuroscience, including, but not limited to, neuropsychiatric disorders, neurodegenerative disorders, mental illness, traumatic brain injury, chronic traumatic encephalopathy, and other neurodegenerative diseases. With advancements in mass spectrometry coupled with bioinformatics and systems biology, neuroproteomics has led to the development of innovative techniques such as microproteomics, single-cell proteomics, and imaging mass spectrometry, which have significantly impacted neuronal biomarker research. By analyzing the complex protein interactions and alterations that occur in the injured brain, neuroproteomics provides valuable insights into the pathophysiological mechanisms underlying neurotrauma. This review explores how such insights can be harnessed to advance personalized medicine (PM) approaches, tailoring treatments based on individual patient profiles. Additionally, we highlight the potential future prospects of neuroproteomics, such as identifying novel biomarkers and developing targeted therapies by employing artificial intelligence (AI) and machine learning (ML). By shedding light on neurotrauma's current state and future directions, this review aims to stimulate further research and collaboration in this promising and transformative field.

Sensor-level MEG combined with machine learning yields robust classification of mild traumatic brain injury patients

OBJECTIVE

Diagnosis of mild traumatic brain injury (mTBI) is challenging despite its high incidence, due to the unspecificity and variety of symptoms and the frequent lack of structural imaging findings. There is a need for reliable and simple-to-use diagnostic tools that would be feasible across sites and patient populations.

METHODS

We evaluated linear machine learning (ML) methods' ability to separate mTBI patients from healthy controls, based on their sensor-level magnetoencephalographic (MEG) power spectra in the subacute phase (<2 months) after a head trauma. We recorded resting-state MEG data from 25 patients and 25 age-sex matched controls and utilized a previously collected data set of 20 patients and 20 controls from a different site. The data sets were analyzed separately with three ML methods.

RESULTS

The median classification accuracies varied between 80 and 95%, without significant differences between the applied ML methods or data sets. The classification accuracies were significantly higher with ML than with traditional sensor-level MEG analysis based on detecting pathological low-frequency activity.

CONCLUSIONS

Easily applicable linear ML methods provide reliable and replicable classification of mTBI patients using sensor-level MEG data.

SIGNIFICANCE

Power spectral estimates combined with ML can classify mTBI patients with high accuracy and have high promise for clinical use.

Machine learning-based classification of chronic traumatic brain injury using hybrid diffusion imaging

Background and purpose

Traumatic brain injury (TBI) can cause progressive neuropathology that leads to chronic impairments, creating a need for biomarkers to detect and monitor this condition to improve outcomes. This study aimed to analyze the ability of data-driven analysis of diffusion tensor imaging (DTI) and neurite orientation dispersion imaging (NODDI) to develop biomarkers to infer symptom severity and determine whether they outperform conventional T1-weighted imaging.

Materials and methods

A machine learning-based model was developed using a dataset of hybrid diffusion imaging of patients with chronic traumatic brain injury. We first extracted the useful features from the hybrid diffusion imaging (HYDI) data and then used supervised learning algorithms to classify the outcome of TBI. We developed three models based on DTI, NODDI, and T1-weighted imaging, and we compared the accuracy results across different models.

Results

Compared with the conventional T1-weighted imaging-based classification with an accuracy of 51.7-56.8%, our machine learning-based models achieved significantly better results with DTI-based models at 58.7-73.0% accuracy and NODDI with an accuracy of 64.0-72.3%.

Conclusion

The machine learning-based feature selection and classification algorithm based on hybrid diffusion features significantly outperform conventional T1-weighted imaging. The results suggest that advanced algorithms can be developed for inferring symptoms of chronic brain injury using feature selection and diffusion-weighted imaging.

Extracting the most discriminating functional connections in mild traumatic brain injury based on machine learning

BACKGROUND

Mild traumatic brain injury (mTBI) is characterized as brain microstructural damage, which may cause a wide range of brain functional disturbances and emotional problems. Brain network analysis based on machine learning is an important means of neuroimaging research. Obtaining the most discriminating functional connection is of great significance to analyze the pathological mechanism of mTBI.

METHODS

To better obtain the most discriminating features of functional connection networks, this study proposes a hierarchical feature selection pipeline (HFSP) composed of Variance Filtering (VF), Lasso, and Principal Component Analysis (PCA). Ablation experiments indicate that each module plays a positive role in classification, validating the robustness and reliability of the HFSP. Furthermore, the HFSP is compared with recursive feature elimination (RFE), elastic net (EN), and locally linear embedding (LLE), verifying its superiority. In addition, this study also utilizes random forest (RF), SVM, Bayesian, linear discriminant analysis (LDA), and logistic regression (LR) as classifiers to evaluate the generalizability of HFSP.

RESULTS

The results show that the indexes obtained from RF are the highest, with accuracy = 89.74%, precision = 91.26%, recall = 89.74%, and F1 score = 89.42%. The HFSP selects 25 pairs of the most discriminating functional connections, mainly distributed in the frontal lobe, occipital lobe, and cerebellum. Nine brain regions show the largest node degree.

LIMITATIONS

The number of samples is small. This study only includes acute mTBI.

CONCLUSIONS

The HFSP is a useful tool for extracting discriminating functional connections and may contribute to the diagnostic processes.

Estimation of Discriminative Multimodal Brain Network Connectivity Using Message-Passing-Based Nonlinear Network Fusion

Effective estimation of brain network connectivity enables better unraveling of the extraordinary complexity interactions of brain regions and helps in auxiliary diagnosis of psychiatric disorders. Considering different modalities can provide comprehensive characterizations of brain connectivity, we propose the message-passing-based nonlinear network fusion (MP-NNF) algorithm to estimate multimodal brain network connectivity. In the proposed method, the initial functional and structural networks were computed from fMRI and DTI separately. Then, we update every unimodal network iteratively, making it more similar to the others in every iteration, and finally converge to one unified network. The estimated brain connectivities integrate complementary information from multiple modalities while preserving their original structure, by adding the strong connectivities present in unimodal brain networks and eliminating the weak connectivities. The effectiveness of the method was evaluated by applying the learned brain connectivity for the classification of major depressive disorder (MDD). Specifically, 82.18% classification accuracy was achieved even with the simple feature selection and classification pipeline, which significantly outperforms the competing methods. Exploration of brain connectivity contributed to MDD identification suggests that the proposed method not only improves the classification performance but also was sensitive to critical disease-related neuroimaging biomarkers.

Multi-Mechanistic Approaches to the Treatment of Traumatic Brain Injury: A Review

Traumatic brain injury (TBI) is a leading cause of death and disability worldwide. Despite extensive research efforts, the majority of trialed monotherapies to date have failed to demonstrate significant benefit. It has been suggested that this is due to the complex pathophysiology of TBI, which may possibly be addressed by a combination of therapeutic interventions. In this article, we have reviewed combinations of different pharmacologic treatments, combinations of non-pharmacologic interventions, and combined pharmacologic and non-pharmacologic interventions for TBI. Both preclinical and clinical studies have been included. While promising results have been found in animal models, clinical trials of combination therapies have not yet shown clear benefit. This may possibly be due to their application without consideration of the evolving pathophysiology of TBI. Improvements of this paradigm may come from novel interventions guided by multimodal neuromonitoring and multimodal imaging techniques, as well as the application of multi-targeted non-pharmacologic and endogenous therapies. There also needs to be a greater representation of female subjects in preclinical and clinical studies.

Machine learning classification of chronic traumatic brain injury using diffusion tensor imaging and NODDI: A replication and extension study

Individuals with acute and chronic traumatic brain injury (TBI) are associated with unique white matter (WM) structural abnormalities, including fractional anisotropy (FA) differences. Our research group previously used FA as a feature in a linear support vector machine (SVM) pattern classifier, observing high classification between individuals with and without acute TBI (i.e., an area under the curve [AUC] value of 75.50%). However, it is not known whether FA could similarly classify between individuals with and without history of chronic TBI. Here, we attempted to replicate our previous work with a new sample, investigating whether FA could similarly classify between incarcerated men with (n = 80) and without (n = 80) self-reported history of chronic TBI. Additionally, given limitations associated with FA, including underestimation of FA values in WM tracts containing crossing fibers, we extended upon our previous study by incorporating neurite orientation dispersion and density imaging (NODDI) metrics, including orientation dispersion (ODI) and isotropic volume (Viso). A linear SVM based classification approach, similar to our previous study, was incorporated here to classify between individuals with and without self-reported chronic TBI using FA and NODDI metrics as separate features. Overall classification rates were similar when incorporating FA and NODDI ODI metrics as features (AUC: 82.50%). Additionally, NODDI-based metrics provided the highest sensitivity (ODI: 85.00%) and specificity (Viso: 82.50%) rates. The current study serves as a replication and extension of our previous study, observing that multiple diffusion MRI metrics can reliably classify between individuals with and without self-reported history of chronic TBI.

Identification of chronic mild traumatic brain injury using resting state functional MRI and machine learning techniques

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].

Chronic Mild Traumatic Brain Injury: Aberrant Static and Dynamic Connectomic Features Identified Through Machine Learning Model Fusion

Traumatic Brain Injury (TBI) is a frequently occurring condition and approximately 90% of TBI cases are classified as mild (mTBI). However, conventional MRI has limited diagnostic and prognostic value, thus warranting the utilization of additional imaging modalities and analysis procedures. The functional connectomic approach using resting-state functional MRI (rs-fMRI) has shown great potential and promising diagnostic capabilities across multiple clinical scenarios, including mTBI. Additionally, there is increasing recognition of a fundamental role of brain dynamics in healthy and pathological cognition. Here, we undertake an in-depth investigation of mTBI-related connectomic disturbances and their emotional and cognitive correlates. We leveraged machine learning and graph theory to combine static and dynamic functional connectivity (FC) with regional entropy values, achieving classification accuracy up to 75% (77, 74 and 76% precision, sensitivity and specificity, respectively). As compared to healthy controls, the mTBI group displayed hypoconnectivity in the temporal poles, which correlated positively with semantic (r = 0.43, p < 0.008) and phonemic verbal fluency (r = 0.46, p < 0.004), while hypoconnectivity in the right dorsal posterior cingulate correlated positively with depression symptom severity (r = 0.54, p < 0.0006). These results highlight the importance of residual FC in these regions for preserved cognitive and emotional function in mTBI. Conversely, hyperconnectivity was observed in the right precentral and supramarginal gyri, which correlated negatively with semantic verbal fluency (r=-0.47, p < 0.003), indicating a potential ineffective compensatory mechanism. These novel results are promising toward understanding the pathophysiology of mTBI and explaining some of its most lingering emotional and cognitive symptoms.

A comparison of the functional connectome in mild traumatic brain injury and post-traumatic stress disorder

Post-traumatic stress disorder (PTSD) and mild traumatic brain injury (mTBI) often co-occur in the context of threat to one's life. These conditions also have an overlapping symptomatology and include symptoms of anxiety, poor concentration and memory problems. A major challenge has been articulating the underlying neurobiology of these overlapping conditions. The primary aim of this study was to compare intrinsic functional connectivity between mTBI (without PTSD) and PTSD (without mTBI). The study included functional MRI data from 176 participants: 42 participants with mTBI, 67 with PTSD and a comparison group of 66 age and sex-matched healthy controls. We used network-based statistical analyses for connectome-wide comparisons of intrinsic functional connectivity between mTBI relative to PTSD and controls. Our results showed no connectivity differences between mTBI and PTSD groups. However, we did find that mTBI had significantly reduced connectivity relative to healthy controls within an extensive network of regions including default mode, executive control, visual and auditory networks. The mTBI group also displayed hyperconnectivity between dorsal and ventral attention networks and perceptual regions. The PTSD group also demonstrated abnormal connectivity within these networks relative to controls. Connectivity alterations were not associated with severity of PTSD or post-concussive symptoms in either clinical group. Taken together, the similar profiles of intrinsic connectivity alterations in these two conditions provide neural evidence that can explain, in part, the overlapping symptomatology between mTBI and PTSD.

Combining Multiple Indices of Diffusion Tensor Imaging Can Better Differentiate Patients with Traumatic Brain Injury from Healthy Subjects

Aim

Diffuse axonal injury (DAI) is one of the most common pathological features of traumatic brain injury (TBI). Diffusion tensor imaging (DTI) indices can be used to identify and quantify white matter microstructural changes following DAI. Recently, many studies have used DTI with various machine learning approaches to predict white matter microstructural changes following TBI. The current study sought to examine whether our classification approach using multiple DTI indices in conjunction with machine learning is a useful tool for diagnosing/classifying TBI patients and healthy controls.

Methods

Participants were adult patients with chronic TBI (n = 26) with DAI pathology, and age- and sex-matched healthy controls (n = 26). DTI images were obtained from all participants. Tract-based spatial statistics analyses were applied to DTI images. Classification models were built using principal component analysis and support vector machines. Receiver operator characteristic curve analysis and area under the curve were used to assess the classification performance of the different classifiers.

Results

Tract-based spatial statistics revealed significantly decreased fractional anisotropy, as well as increased mean diffusivity, axial diffusivity, and radial diffusivity in patients with TBI compared with healthy controls (all p-values < 0.01). The principal component analysis and support vector machine-based machine learning classification using combined DTI indices classified patients with TBI and healthy controls with an accuracy of 90.5% with an area under the curve of 93 ± 0.09.

Conclusion

These results highlight the potential of our approach combining multiple DTI measures to identify patients with TBI.

Multi-Site Mild Traumatic Brain Injury Classification with Machine Learning and Harmonization

Traumatic brain injury (TBI) can drastically affect an individual's cognition, physical, emotional wellbeing, and behavior. Even patients with mild TBI (mTBI) may suffer from a variety of long-lasting symptoms, which motivates researchers to find better biomarkers. Machine learning algorithms have shown promising results in detecting mTBI from resting-state functional network connectivity (rsFNC) data. However, data collected at multiple sites introduces additional noise called site-effects, resulting in erroneous conclusions. Site errors are controlled through a process called harmonization, but its use in classifying neuroimaging data has been addressed lightly. With the ongoing need to improve mTBI detection, this study shows that harmonization should be integrated into the machine learning process when working with multi-site neuroimaging datasets.

Identifying Alcohol Use Disorder With Resting State Functional Magnetic Resonance Imaging Data: A Comparison Among Machine Learning Classifiers

Alcohol use disorder (AUD) is a burden to society creating social and health problems. Detection of AUD and its effects on the brain are difficult to assess. This problem is enhanced by the comorbid use of other substances such as nicotine that has been present in previous studies. Recent machine learning algorithms have raised the attention of researchers as a useful tool in studying and detecting AUD. This work uses AUD and controls samples free of any other substance use to assess the performance of a set of commonly used machine learning classifiers detecting AUD from resting state functional network connectivity (rsFNC) derived from independent component analysis. The cohort used included 51 alcohol dependent subjects and 51 control subjects. Despite alcohol, none of the 102 subjects reported use of nicotine, cannabis or any other dependence or habit formation substance. Classification features consisted of whole brain rsFNC estimates undergoing a feature selection process using a random forest approach. Features were then fed to 10 different machine learning classifiers to be evaluated based on their classification performance. A neural network classifier showed the highest performance with an area under the curve (AUC) of 0.79. Other good performers with similar AUC scores were logistic regression, nearest neighbor, and support vector machine classifiers. The worst results were obtained with Gaussian process and quadratic discriminant analysis. The feature selection outcome pointed to functional connections between visual, sensorimotor, executive control, reward, and salience networks as the most relevant for classification. We conclude that AUD can be identified using machine learning classifiers in the absence of nicotine comorbidity.

Combining MRI and cognitive evaluation to classify concussion in university athletes

Current methods of concussion assessment lack the objectivity and reliability to detect neurological injury. This multi-site study uses combinations of neuroimaging (diffusion tensor imaging and resting state functional MRI) and cognitive measures to train algorithms to detect the presence of concussion in university athletes. Athletes (29 concussed, 48 controls) completed symptom reports, brief cognitive evaluation, and MRI within 72 h of injury. Hierarchical linear regression compared groups on cognitive and neuroimaging measures while controlling for sex and data collection site. Logistic regression and support vector machine models were trained using cognitive and neuroimaging measures and evaluated for overall accuracy, sensitivity, and specificity. Concussed athletes reported greater symptoms than controls (∆R² = 0.32, p < .001), and performed worse on tests of concentration (∆R² = 0.07, p < .05) and delayed memory (∆R² = 0.17, p < .001). Concussed athletes showed lower functional connectivity within the frontoparietal and primary visual networks (p < .05), but did not differ on mean diffusivity and fractional anisotropy. Of the cognitive measures, classifiers trained using delayed memory yielded the best performance with overall accuracy of 71%, though sensitivity was poor at 46%. Of the neuroimaging measures, classifiers trained using mean diffusivity yielded similar accuracy. Combining cognitive measures with mean diffusivity increased overall accuracy to 74% and sensitivity to 64%, comparable to the sensitivity of symptom report. Trained algorithms incorporating both MRI and cognitive performance variables can reliably detect common neurobiological sequelae of acute concussion. The integration of multi-modal data can serve as an objective, reliable tool in the assessment and diagnosis of concussion.

Functional Connectome Dynamics After Mild Traumatic Brain Injury According to Age and Sex

Neural and cognitive deficits after mild traumatic brain injury (mTBI) are paralleled by changes in resting state functional correlation (FC) networks that mirror post-traumatic pathophysiology effects on functional outcomes. Using functional magnetic resonance images acquired both acutely and chronically after injury (∼1 week and ∼6 months post-injury, respectively), we map post-traumatic FC changes across 136 participants aged 19-79 (52 females), both within and between the brain's seven canonical FC networks: default mode, dorsal attention, frontoparietal, limbic, somatomotor, ventral attention, and visual. Significant sex-dependent FC changes are identified between (A) visual and limbic, and between (B) default mode and somatomotor networks. These changes are significantly associated with specific functional recovery patterns across all cognitive domains (p < 0.05, corrected). Changes in FC between default mode, somatomotor, and ventral attention networks, on the one hand, and both temporal and occipital regions, on the other hand, differ significantly by age group (p < 0.05, corrected), and are paralleled by significant sex differences in cognitive recovery independently of age at injury (p < 0.05, corrected). Whereas females' networks typically feature both significant (p < 0.036, corrected) and insignificant FC changes, males more often exhibit significant FC decreases between networks (e.g., between dorsal attention and limbic, visual and limbic, default-mode and somatomotor networks, p < 0.0001, corrected), all such changes being accompanied by significantly weaker recovery of cognitive function in males, particularly older ones (p < 0.05, corrected). No significant FC changes were found across 35 healthy controls aged 66-92 (20 females). Thus, male sex and older age at injury are risk factors for significant FC alterations whose patterns underlie post-traumatic cognitive deficits. This is the first study to map, systematically, how mTBI impacts FC between major human functional networks.

Integrated Features for Optimizing Machine Learning Classifiers of Pediatric and Young Adults With a Post-Traumatic Headache From Healthy Controls

Post-traumatic headache (PTH) is a challenging clinical condition to identify and treat as it integrates multiple subjectively defined symptoms with underlying physiological processes. The precise mechanisms underlying PTH are unclear, and it remains to be understood how to integrate the patient experience with underlying biology when attempting to classify persons with PTH, particularly in the pediatric setting where patient self-report may be highly variable. The objective of this investigation was to evaluate the use of different machine learning (ML) classifiers to differentiate pediatric and young adult subjects with PTH from healthy controls using behavioral data from self-report questionnaires that reflect concussion symptoms, mental health, pain experience of the participants, and structural brain imaging from cortical and sub-cortical locations. Behavioral data, alongside brain imaging, survived data reduction methods and both contributed toward final models. Behavioral data that contributed towards the final model included both the child and parent perspective of the pain-experience. Brain imaging features produced two unique clusters that reflect regions that were previously found in mild traumatic brain injury (mTBI) and PTH. Affinity-based propagation analysis demonstrated that behavioral data remained independent relative to neuroimaging data that suggest there is a role for both behavioral and brain imaging data when attempting to classify children with PTH.

Enhanced Performance by Interpretable Low-Frequency Electroencephalogram Oscillations in the Machine Learning-Based Diagnosis of Post-traumatic Stress Disorder

Electroencephalography (EEG)-based diagnosis of psychiatric diseases using machine-learning approaches has made possible the objective diagnosis of various psychiatric diseases. The objective of this study was to improve the performance of a resting-state EEG-based computer-aided diagnosis (CAD) system to diagnose post-traumatic stress disorder (PTSD), by optimizing the frequency bands used to extract EEG features. We used eyes-closed resting-state EEG data recorded from 77 PTSD patients and 58 healthy controls (HC). Source-level power spectrum densities (PSDs) of the resting-state EEG data were extracted from 6 frequency bands (delta, theta, alpha, low-beta, high-beta, and gamma), and the PSD features of each frequency band and their combinations were independently used to discriminate PTSD and HC. The classification performance was evaluated using support vector machine with leave-one-out cross validation. The PSD features extracted from slower-frequency bands (delta and theta) showed significantly higher classification performance than those of relatively higher-frequency bands. The best classification performance was achieved when using delta PSD features (86.61%), which was significantly higher than that reported in a recent study by about 13%. The PSD features selected to obtain better classification performances could be explained from a neurophysiological point of view, demonstrating the promising potential to develop a clinically reliable EEG-based CAD system for PTSD diagnosis.

Detection of Chronic Blast-Related Mild Traumatic Brain Injury with Diffusion Tensor Imaging and Support Vector Machines

Blast-related mild traumatic brain injury (bmTBI) often leads to long-term sequalae, but diagnostic approaches are lacking due to insufficient knowledge about the predominant pathophysiology. This study aimed to build a diagnostic model for future verification by applying machine-learning based support vector machine (SVM) modeling to diffusion tensor imaging (DTI) datasets to elucidate white-matter features that distinguish bmTBI from healthy controls (HC). Twenty subacute/chronic bmTBI and 19 HC combat-deployed personnel underwent DTI. Clinically relevant features for modeling were selected using tract-based analyses that identified group differences throughout white-matter tracts in five DTI metrics to elucidate the pathogenesis of injury. These features were then analyzed using SVM modeling with cross validation. Tract-based analyses revealed abnormally decreased radial diffusivity (RD), increased fractional anisotropy (FA) and axial/radial diffusivity ratio (AD/RD) in the bmTBI group, mostly in anterior tracts (29 features). SVM models showed that FA of the anterior/superior corona radiata and AD/RD of the corpus callosum and anterior limbs of the internal capsule (5 features) best distinguished bmTBI from HCs with 89% accuracy. This is the first application of SVM to identify prominent features of bmTBI solely based on DTI metrics in well-defined tracts, which if successfully validated could promote targeted treatment interventions.

A Decentralized ComBat Algorithm and Applications to Functional Network Connectivity

Recent studies showed that working with neuroimage data collected from different research facilities or locations may incur additional source dependency, affecting the overall statistical power. This problem can be mitigated with data harmonization approaches. Recently, the ComBat method has become commonly adopted for various neuroimage modalities. While open neuroimaging datasets are becoming more common, a substantial amount of data is still unable to be shared for various reasons. In addition, current approaches require moving all the data to a central location, which requires additional resources and creates redundant copies of the same datasets. To address these issues, we propose a decentralized harmonization approach that does not create redundant copies of the original datasets and performs remote operations on the datasets separately without sharing any individual subject data, ensuring a certain level of privacy and reducing regulatory hurdles. We proposed a novel approach called "Decentralized ComBat" which can harmonize datasets separately without combining the datasets. We tested our model by harmonizing functional network connectivity datasets from two traumatic brain injury studies in a decentralized way. Also, we used simulations to analyze the performance and scalability of our model when the number of data collection sites increases. We compare the output with centralized ComBat and show that the proposed approach produces similar results, increasing the sensitivity of the functional network connectivity analysis and validating our approach. Simulations show that our model can be easily scaled to many more datasets based on the requirement. In sum, we believe this provides a powerful tool, further complementing open data and allowing for integrating public and private datasets.

Mixture Model Framework for Traumatic Brain Injury Prognosis Using Heterogeneous Clinical and Outcome Data

Prognoses of Traumatic Brain Injury (TBI) outcomes are neither easily nor accurately determined from clinical indicators. This is due in part to the heterogeneity of damage inflicted to the brain, ultimately resulting in diverse and complex outcomes. Using a data-driven approach on many distinct data elements may be necessary to describe this large set of outcomes and thereby robustly depict the nuanced differences among TBI patients' recovery. In this work, we develop a method for modeling large heterogeneous data types relevant to TBI. Our approach is geared toward the probabilistic representation of mixed continuous and discrete variables with missing values. The model is trained on a dataset encompassing a variety of data types, including demographics, blood-based biomarkers, and imaging findings. In addition, it includes a set of clinical outcome assessments at 3, 6, and 12 months post-injury. The model is used to stratify patients into distinct groups in an unsupervised learning setting. We use the model to infer outcomes using input data, and show that the collection of input data reduces uncertainty of outcomes over a baseline approach. In addition, we quantify the performance of a likelihood scoring technique that can be used to self-evaluate the extrapolation risk of prognosis on unseen patients.

Cingulo-Opercular and Frontoparietal Network Control of Effort and Fatigue in Mild Traumatic Brain Injury

Neural substrates of fatigue in traumatic brain injury (TBI) are not well understood despite the considerable burden of fatigue on return to productivity. Fatigue is associated with diminishing performance under conditions of high cognitive demand, sense of effort, or need for motivation, all of which are associated with cognitive control brain network integrity. We hypothesize that the pathophysiology of TBI results in damage to diffuse cognitive control networks, disrupting coordination of moment-to-moment monitoring, prediction, and regulation of behavior. We investigate the cingulo-opercular (CO) and frontoparietal (FP) networks, which are engaged to sustain attention for task and maintain performance. A total of 61 individuals with mild TBI and 42 orthopedic control subjects participated in functional MRI during performance of a constant effort task requiring altering the amount of effort (25, 50, or 75% of maximum effort) utilized to manually squeeze a pneumostatic bulb across six 30-s trials. Network-based statistics assessed within-network organization and fluctuation with task manipulations by group. Results demonstrate small group differences in network organization, but considerable group differences in the evolution of task-related modulation of connectivity. The mild TBI group demonstrated elevated CO connectivity throughout the task with little variation in effort level or time on task (TOT), while CO connectivity diminished over time in controls. Several interregional CO connections were predictive of fatigue in the TBI group. In contrast, FP connectivity fluctuated with task manipulations and predicted fatigue in the controls, but connectivity fluctuations were delayed in the mild traumatic brain injury (mTBI) group and did not relate to fatigue. Thus, the mTBI group's hyper-connectivity of the CO irrespective of task demands, along with hypo-connectivity and delayed peak connectivity of the FP, may allow for attainment of task goals, but also contributes to fatigue. Findings are discussed in relation to performance monitoring of prediction error that relies on internal cues from sensorimotor feedback during task performance. Delay or inability to detect and respond to prediction errors in TBI, particularly evident in bilateral insula-temporal CO connectivity, corresponds to day-to-day fatigue and fatigue during task performance.

Development of a machine learning algorithm for predicting in-hospital and 1-year mortality after traumatic spinal cord injury

BACKGROUND CONTEXT

Current prognostic tools such as the Injury Severity Score (ISS) that predict mortality following trauma do not adequately consider the unique characteristics of traumatic spinal cord injury (tSCI).

PURPOSE

Our aim was to develop and validate a prognostic tool that can predict mortality following tSCI.

STUDY DESIGN

Retrospective review of a prospective cohort study.

PATIENT SAMPLE

Data was collected from 1245 persons with acute tSCI who were enrolled in the Rick Hansen Spinal Cord Injury Registry between 2004 and 2016.

OUTCOME MEASURES

In-hospital and 1-year mortality following tSCI.

METHODS

Machine learning techniques were used on patient-level data (n=849) to develop the Spinal Cord Injury Risk Score (SCIRS) that can predict mortality based on age, neurological level and completeness of injury, AOSpine classification of spinal column injury morphology, and Abbreviated Injury Scale scores. Validation of the SCIRS was performed by testing its accuracy in an independent validation cohort (n=396) and comparing its performance to the ISS, a measure which is used to predict mortality following general trauma.

RESULTS

For 1-year mortality prediction, the values for the Area Under the Receiver Operating Characteristic Curve (AUC) for the development cohort were 0.84 (standard deviation=0.029) for the SCIRS and 0.55 (0.041) for the ISS. For the validation cohort, AUC values were 0.86 (0.051) for the SCIRS and 0.71 (0.074) for the ISS. For in-hospital mortality, AUC values for the development cohort were 0.87 (0.028) and 0.60 (0.050) for the SCIRS and ISS, respectively. For the validation cohort, AUC values were 0.85 (0.054) for the SCIRS and 0.70 (0.079) for the ISS.

CONCLUSIONS

The SCIRS can predict in-hospital and 1-year mortality following tSCI more accurately than the ISS. The SCIRS can be used in research to reduce bias in estimating parameters and can help adjust for coefficients during model development. Further validation using larger sample sizes and independent datasets is needed to assess its reliability and to evaluate using it as an assessment tool to guide clinical decision-making and discussions with patients and families.

Post-concussive mTBI in Student Athletes: MRI Features and Machine Learning

Purpose: To determine and characterize the radiomics features from structural MRI (MPRAGE) and Diffusion Tensor Imaging (DTI) associated with the presence of mild traumatic brain injuries on student athletes with post-concussive syndrome (PCS).

Material and Methods: 122 student athletes (65 M, 57 F), median (IQR) age 18.8 (15–20) years, with a mixed level of play and sports activities, with a known history of concussion and clinical PCS, and 27 (15 M, 12 F), median (IQR) age 20 (19, 21) years, concussion free athlete subjects were MRI imaged in a clinical MR machine. MPRAGE and DTI-FA and DTI-ADC images were used to extract radiomic features from white and gray matter regions within the entire brain (2 ROI) and the eight main lobes of the brain (16 ROI) for a total of 18 analyzed regions. Radiomic features were divided into five different data sets used to train and cross-validate five different filter-based Support Vector Machines. The top selected features of the top model were described. Furthermore, the test predictions of the top four models were ensembled into a single average prediction. The average prediction was evaluated for the association to the number of concussions and time from injury.

Results: Ninety-one PCS subjects passed inclusion criteria (91 Cases, 27 controls). The average prediction of the top four models had a sensitivity of 0.80, 95% CI: [0.71, 0.88] and specificity of 0.74 95%CI [0.54, 0.89] for distinguishing subjects from controls. The white matter features were strongly associated with mTBI, while the whole-brain analysis of gray matter showed the worst association. The predictive index was significantly associated with the number of concussions (p < 0.0001) and associated with the time from injury (p < 0.01).

Conclusion: MRI Radiomic features are associated with a history of mTBI and they were successfully used to build a predictive machine learning model for mTBI for subjects with PCS associated with a history of one or more concussions.

Differential Tractography and Correlation Tractography Findings on Patients With Mild Traumatic Brain Injury: A Pilot Study

Differential tractography and correlation tractography are new tractography modalities to study neuronal changes in brain diseases, but their performances in detecting neuronal injuries are yet to be investigated in patients with mild traumatic brain injury (mTBI). Here we investigated the white matter injury in mTBI patients using differential and correlation tractography. The diffusion MRI was acquired at 33 mTBI patients and 31 health controls. 7 of the mTBI patients had one-year follow-up scans, and differential tractography was used to evaluate injured fiber bundles on these 7 patients. All subjects were evaluated using digital symbol substitution test (DSST) and trail making test A (TMT-A), and the correlation tractography was performed to explore the exact pathways related to the cognitive performance. Our results showed that differential tractography revealed neuronal changes in the corpus callosum in all 7 follow-up mTBI patients with FDR between 0.007 and 0.17. Further, the correlation tractography showed that the splenium of the corpus callosum, combined with the right superior longitudinal fasciculus and right cingulum, were correlated with DSST (FDR = 0.001669) in the acute mTBI patients. The cognitive impairment findings in the acute stage and the longitudinal findings in the corpus callosum in the chronic stage of mTBI patients suggest that differential tractography and correlation tractography are valuable tools in the diagnostic and prognostic evaluation of neuronal injuries in mTBI patients.

Deep Learning Recurrent Neural Network for Concussion Classification in Adolescents Using Raw Electroencephalography Signals: Toward a Minimal Number of Sensors

Artificial neural networks (ANNs) are showing increasing promise as decision support tools in medicine and particularly in neuroscience and neuroimaging. Recently, there has been increasing work on using neural networks to classify individuals with concussion using electroencephalography (EEG) data. However, to date the need for research grade equipment has limited the applications to clinical environments. We recently developed a deep learning long short-term memory (LSTM) based recurrent neural network to classify concussion using raw, resting state data using 64 EEG channels and achieved high accuracy in classifying concussion. Here, we report on our efforts to develop a clinically practical system using a minimal subset of EEG sensors. EEG data from 23 athletes who had suffered a sport-related concussion and 35 non-concussed, control athletes were used for this study. We tested and ranked each of the original 64 channels based on its contribution toward the concussion classification performed by the original LSTM network. The top scoring channels were used to train and test a network with the same architecture as the previously trained network. We found that with only six of the top scoring channels the classifier identified concussions with an accuracy of 94%. These results show that it is possible to classify concussion using raw, resting state data from a small number of EEG sensors, constituting a first step toward developing portable, easy to use EEG systems that can be used in a clinical setting.

Discriminating mild traumatic brain injury using sparse dictionary learning of functional network dynamics

Mild traumatic brain injury (mTBI) is usually caused by a bump, blow, or jolt to the head or penetrating head injury, and carries the risk of inducing cognitive disorders. However, identifying the biomarkers for the diagnosis of mTBI is challenging as evident abnormalities in brain anatomy are rarely found in patients with mTBI. In this study, we tested whether the alteration of functional network dynamics could be used as potential biomarkers to better diagnose mTBI. We propose a sparse dictionary learning framework to delineate spontaneous fluctuation of functional connectivity into the subject-specific time-varying evolution of a set of overlapping group-level sparse connectivity components (SCCs) based on the resting-state functional magnetic resonance imaging (fMRI) data from 31 mTBI patients in the early acute phase (<3 days postinjury) and 31 healthy controls (HCs). The identified SCCs were consistently distributed in the cohort of subjects without significant inter-group differences in connectivity patterns. Nevertheless, subject-specific temporal expression of these SCCs could be used to discriminate patients with mTBI from HCs with a classification accuracy of 74.2% (specificity 64.5% and sensitivity 83.9%) using leave-one-out cross-validation. Taken together, our findings indicate neuroimaging biomarkers for mTBI individual diagnosis based on the temporal expression of SCCs underlying time-resolved functional connectivity.

Phybrata Sensors and Machine Learning for Enhanced Neurophysiological Diagnosis and Treatment

Concussion injuries remain a significant public health challenge. A significant unmet clinical need remains for tools that allow related physiological impairments and longer-term health risks to be identified earlier, better quantified, and more easily monitored over time. We address this challenge by combining a head-mounted wearable inertial motion unit (IMU)-based physiological vibration acceleration (“phybrata”) sensor and several candidate machine learning (ML) models. The performance of this solution is assessed for both binary classification of concussion patients and multiclass predictions of specific concussion-related neurophysiological impairments. Results are compared with previously reported approaches to ML-based concussion diagnostics. Using phybrata data from a previously reported concussion study population, four different machine learning models (Support Vector Machine, Random Forest Classifier, Extreme Gradient Boost, and Convolutional Neural Network) are first investigated for binary classification of the test population as healthy vs. concussion (Use Case 1). Results are compared for two different data preprocessing pipelines, Time-Series Averaging (TSA) and Non-Time-Series Feature Extraction (NTS). Next, the three best-performing NTS models are compared in terms of their multiclass prediction performance for specific concussion-related impairments: vestibular, neurological, both (Use Case 2). For Use Case 1, the NTS model approach outperformed the TSA approach, with the two best algorithms achieving an F1 score of 0.94. For Use Case 2, the NTS Random Forest model achieved the best performance in the testing set, with an F1 score of 0.90, and identified a wider range of relevant phybrata signal features that contributed to impairment classification compared with manual feature inspection and statistical data analysis. The overall classification performance achieved in the present work exceeds previously reported approaches to ML-based concussion diagnostics using other data sources and ML models. This study also demonstrates the first combination of a wearable IMU-based sensor and ML model that enables both binary classification of concussion patients and multiclass predictions of specific concussion-related neurophysiological impairments.

Integrative Neuroinformatics for Precision Prognostication and Personalized Therapeutics in Moderate and Severe Traumatic Brain Injury

Despite changes in guideline-based management of moderate/severe traumatic brain injury (TBI) over the preceding decades, little impact on mortality and morbidity have been seen. This argues against the "one-treatment fits all" approach to such management strategies. With this, some preliminary advances in the area of personalized medicine in TBI care have displayed promising results. However, to continue transitioning toward individually-tailored care, we require integration of complex "-omics" data sets. The past few decades have seen dramatic increases in the volume of complex multi-modal data in moderate and severe TBI care. Such data includes serial high-fidelity multi-modal characterization of the cerebral physiome, serum/cerebrospinal fluid proteomics, admission genetic profiles, and serial advanced neuroimaging modalities. Integrating these complex and serially obtained data sets, with patient baseline demographics, treatment information and clinical outcomes over time, can be a daunting task for the treating clinician. Within this review, we highlight the current status of such multi-modal omics data sets in moderate/severe TBI, current limitations to the utilization of such data, and a potential path forward through employing integrative neuroinformatic approaches, which are applied in other neuropathologies. Such advances are positioned to facilitate the transition to precision prognostication and inform a top-down approach to the development of personalized therapeutics in moderate/severe TBI.