Investigation of Machine Learning and Deep Learning Approaches for Detection of Mild Traumatic Brain Injury from Human Sleep Electroencephalogram

2021 ◽  
Author(s):  
Manoj Vishwanath ◽  
Salar Jafarlou ◽  
Ikhwan Shin ◽  
Nikil Dutt ◽  
Amir M. Rahmani ◽  
...  
Keyword(s):  
Machine Learning ◽  
Traumatic Brain Injury ◽  
Deep Learning ◽  
Brain Injury ◽  
Mild Traumatic Brain Injury ◽  
Learning Approaches ◽  
Human Sleep

scholarly journals Self-Reported Complaints as Prognostic Markers for Outcome After Mild Traumatic Brain Injury in Elderly: A Machine Learning Approach

2021 ◽  
Vol 12 ◽  
Author(s):  
Mayra Bittencourt ◽  
Sebastián A. Balart-Sánchez ◽  
Natasha M. Maurits ◽  
Joukje van der Naalt
Keyword(s):  
Machine Learning ◽  
Traumatic Brain Injury ◽  
Brain Injury ◽  
Elderly Patients ◽  
Neck Pain ◽  
Functional Outcome ◽  
Mild Traumatic Brain Injury ◽  
Prognostic Markers ◽  
The Elderly ◽  
Post Injury

Self-reported complaints are common after mild traumatic brain injury (mTBI). Particularly in the elderly with mTBI, the pre-injury status might play a relevant role in the recovery process. In most mTBI studies, however, pre-injury complaints are neither analyzed nor are the elderly included. Here, we aimed to identify which individual pre- and post-injury complaints are potential prognostic markers for incomplete recovery (IR) in elderly patients who sustained an mTBI. Since patients report many complaints across several domains that are strongly related, we used an interpretable machine learning (ML) approach to robustly deal with correlated predictors and boost classification performance. Pre- and post-injury levels of 20 individual complaints, as self-reported in the acute phase, were analyzed. We used data from two independent studies separately: UPFRONT study was used for training and validation and ReCONNECT study for independent testing. Functional outcome was assessed with the Glasgow Outcome Scale Extended (GOSE). We dichotomized functional outcome into complete recovery (CR; GOSE = 8) and IR (GOSE ≤ 7). In total 148 elderly with mTBI (median age: 67 years, interquartile range [IQR]: 9 years; UPFRONT: N = 115; ReCONNECT: N = 33) were included in this study. IR was observed in 74 (50%) patients. The classification model (IR vs. CR) achieved a good performance (the area under the receiver operating characteristic curve [ROC-AUC] = 0.80; 95% CI: 0.74–0.86) based on a subset of only 8 out of 40 pre- and post-injury complaints. We identified increased neck pain (p = 0.001) from pre- to post-injury as the strongest predictor of IR, followed by increased irritability (p = 0.011) and increased forgetfulness (p = 0.035) from pre- to post-injury. Our findings indicate that a subset of pre- and post-injury physical, emotional, and cognitive complaints has predictive value for determining long-term functional outcomes in elderly patients with mTBI. Particularly, post-injury neck pain, irritability, and forgetfulness scores were associated with IR and should be assessed early. The application of an ML approach holds promise for application in self-reported questionnaires to predict outcomes after mTBI.

scholarly journals The Evolution of White Matter Changes After Mild Traumatic Brain Injury: A DTI and NODDI Study

2018 ◽  
Author(s):  
Eva M. Palacios ◽  
Julia P Owen ◽  
Esther L. Yuh ◽  
Maxwell B. Wang ◽  
Mary J. Vassar ◽  
...  
Keyword(s):  
Machine Learning ◽  
Traumatic Brain Injury ◽  
Brain Injury ◽  
White Matter ◽  
Mild Traumatic Brain Injury ◽  
Vasogenic Edema ◽  
Free Water ◽  
Cross Sectional ◽  
Neurite Density ◽  
Learning Analysis

ABSTRACTNeuroimaging biomarkers show promise for improving precision diagnosis and prognosis after mild traumatic brain injury (mTBI), but none has yet been adopted in routine clinical practice. Biophysical modeling of multishell diffusion MRI, using the neurite orientation dispersion and density imaging (NODDI) framework, may improve upon conventional diffusion tensor imaging (DTI) in revealing subtle patterns of underlying white matter microstructural pathology, such as diffuse axonal injury (DAI) and neuroinflammation, that are important for detecting mTBI and determining patient outcome. With a cross-sectional and longitudinal design, we assessed structural MRI, DTI and NODDI in 40 mTBI patients at 2 weeks and 6 months after injury and 14 matched control participants with orthopedic trauma but not suffering from mTBI at 2 weeks. Self-reported and performance-based cognitive measures assessing postconcussive symptoms, memory, executive functions and processing speed were investigated in post-acute and chronic phase after injury for the mTBI subjects. Machine learning analysis was used to identify mTBI patients with the best neuropsychological improvement over time and relate this outcome to DTI and NODDI biomarkers. In the cross-sectional comparison with the trauma control group at 2 weeks post-injury, mTBI patients showed decreased fractional anisotropy (FA) and increased mean diffusivity (MD) on DTI mainly in anterior tracts that corresponded to white matter regions of elevated free water fraction (FISO) on NODDI, signifying vasogenic edema. Patients showed decreases from 2 weeks to 6 months in white matter neurite density on NODDI, predominantly in posterior tracts. No significant longitudinal changes in DTI metrics were observed. The machine learning analysis divided the mTBI patients into two groups based on their recovery. Voxel-wise group comparison revealed associations between white matter orientation dispersion index (ODI) and FISO with degree and trajectory of improvement within the mTBI group. In conclusion, white matter FA and MD alterations early after mTBI might reflect vasogenic edema, as shown by elevated free water on NODDI. Longer-term declines in neurite density on NODDI suggest progressive axonal degeneration due to DAI, especially in tracts known to be integral to the structural connectome. Overall, these results show that the NODDI parameters appear to be more sensitive to longitudinal changes than DTI metrics. Thus, NODDI merits further study in larger cohorts for mTBI diagnosis, prognosis and treatment monitoring.

scholarly journals Investigation of Machine Learning Approaches for Traumatic Brain Injury Classification via EEG Assessment in Mice

Sensors ◽  
2020 ◽  
Vol 20 (7) ◽  
pp. 2027 ◽  
Author(s):  
Manoj Vishwanath ◽  
Salar Jafarlou ◽  
Ikhwan Shin ◽  
Miranda M. Lim ◽  
Nikil Dutt ◽  
...  
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Traumatic Brain Injury ◽  
Brain Injury ◽  
Mouse Model ◽  
Average Power ◽  
Support Vector ◽  
Control Group ◽  
Learning Approaches ◽  
Robust Detection

Due to the difficulties and complications in the quantitative assessment of traumatic brain injury (TBI) and its increasing relevance in today’s world, robust detection of TBI has become more significant than ever. In this work, we investigate several machine learning approaches to assess their performance in classifying electroencephalogram (EEG) data of TBI in a mouse model. Algorithms such as decision trees (DT), random forest (RF), neural network (NN), support vector machine (SVM), K-nearest neighbors (KNN) and convolutional neural network (CNN) were analyzed based on their performance to classify mild TBI (mTBI) data from those of the control group in wake stages for different epoch lengths. Average power in different frequency sub-bands and alpha:theta power ratio in EEG were used as input features for machine learning approaches. Results in this mouse model were promising, suggesting similar approaches may be applicable to detect TBI in humans in practical scenarios.

Sign in / Sign up

Export Citation Format

Share Document