An interpretable tinnitus prediction framework using gap-prepulse inhibition in auditory late response and electroencephalogram.

BACKGROUND AND OBJECTIVE: Tinnitus is a neuropathological condition that results in mild buzzing or ringing of the ears without an external sound source. Current tinnitus diagnostic methods often rely on subjective assessment and require intricate medical examinations. This study aimed to propose an interpretable tinnitus diagnostic framework using auditory late response (ALR) and electroencephalogram (EEG), inspired by the gap-prepulse inhibition (GPI) paradigm. METHODS: We collected spontaneous EEG and ALR data from 44 patients with tinnitus and 47 hearing loss-matched controls using specialized hardware to capture responses to sound stimuli with embedded gaps. In this cohort study of tinnitus and control groups, we examined EEG spectral and ALR features of N-P complexes, comparing the responses to gap durations of 50 and 20 ms alongside no-gap conditions. To this end, we developed an interpretable tinnitus diagnostic model using ALR and EEG metrics, boosting machine learning architecture, and explainable feature attribution approaches. RESULTS: Our proposed model achieved 90 % accuracy in identifying tinnitus, with an area under the performance curve of 0.89. The explainable artificial intelligence approaches have revealed gap-embedded ALR features such as the GPI ratio of N1-P2 and EEG spectral ratio, which can serve as diagnostic metrics for tinnitus. Our method successfully provides personalized prediction explanations for tinnitus diagnosis using gap-embedded auditory and neurological features. CONCLUSIONS: Deficits in GPI alongside activity in the EEG alpha-beta ratio offer a promising screening tool for assessing tinnitus risk, aligning with current clinical insights from hearing research.

Estimation of patient-reported outcome measures based on features of knee joint muscle co-activation in advanced knee osteoarthritis.

Electromyography (EMG) is considered a potential predictive tool for the severity of knee osteoarthritis (OA) symptoms and functional outcomes. Patient-reported outcome measures (PROMs), such as the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and visual analog scale (VAS), are used to determine the severity of knee OA. We aim to investigate muscle activation and co-contraction patterns through EMG from the lower extremity muscles of patients with advanced knee OA patients and evaluate the effectiveness of an interpretable machine-learning model to estimate the severity of knee OA according to the WOMAC (pain, stiffness, and physical function) and VAS using EMG gait features. To explore neuromuscular gait patterns with knee OA severity, EMG from rectus femoris, medial hamstring, tibialis anterior, and gastrocnemius muscles were recorded from 84 patients diagnosed with advanced knee OA during ground walking. Muscle activation patterns and co-activation indices were calculated over the gait cycle for pairs of medial and lateral muscles. We utilized machine-learning regression models to estimate the severity of knee OA symptoms according to the PROMs using muscle activity and co-contraction features. Additionally, we utilized the Shapley Additive Explanations (SHAP) to interpret the contribution of the EMG features to the regression model for estimation of knee OA severity according to WOMAC and VAS. Muscle activity and co-contraction patterns varied according to the functional limitations associated with knee OA severity according to VAS and WOMAC. The coefficient of determination of the cross-validated regression model is 0.85 for estimating WOMAC, 0.82 for pain, 0.85 for stiffness, and 0.85 for physical function, as well as VAS scores, utilizing the gait features. SHAP explanation revealed that greater co-contraction of lower extremity muscles during the weight acceptance and swing phases indicated more severe knee OA. The identified muscle co-activation patterns may be utilized as objective candidate outcomes to better understand the severity of knee OA.

Interpreting Stroke-Impaired Electromyography Patterns through Explainable Artificial Intelligence.

Electromyography (EMG) proves invaluable myoelectric manifestation in identifying neuromuscular alterations resulting from ischemic strokes, serving as a potential marker for diagnostics of gait impairments caused by ischemia. This study aims to develop an interpretable machine learning (ML) framework capable of distinguishing between the myoelectric patterns of stroke patients and those of healthy individuals through Explainable Artificial Intelligence (XAI) techniques. The research included 48 stroke patients (average age 70.6 years, 65% male) undergoing treatment at a rehabilitation center, alongside 75 healthy adults (average age 76.3 years, 32% male) as the control group. EMG signals were recorded from wearable devices positioned on the bicep femoris and lateral gastrocnemius muscles of both lower limbs during indoor ground walking in a gait laboratory. Boosting ML techniques were deployed to identify stroke-related gait impairments using EMG gait features. Furthermore, we employed XAI techniques, such as Shapley Additive Explanations (SHAP), Local Interpretable Model-Agnostic Explanations (LIME), and Anchors to interpret the role of EMG variables in the stroke-prediction models. Among the ML models assessed, the GBoost model demonstrated the highest classification performance (AUROC: 0.94) during cross-validation with the training dataset, and it also overperformed (AUROC: 0.92, accuracy: 85.26%) when evaluated using the testing EMG dataset. Through SHAP and LIME analyses, the study identified that EMG spectral features contributing to distinguishing the stroke group from the control group were associated with the right bicep femoris and lateral gastrocnemius muscles. This interpretable EMG-based stroke prediction model holds promise as an objective tool for predicting post-stroke gait impairments. Its potential application could greatly assist in managing post-stroke rehabilitation by providing reliable EMG biomarkers and address potential gait impairment in individuals recovering from ischemic stroke.

An Explainable EEG-Based Human Activity Recognition Model Using Machine-Learning Approach and LIME.

Electroencephalography (EEG) is a non-invasive method employed to discern human behaviors by monitoring the neurological responses during cognitive and motor tasks. Machine learning (ML) represents a promising tool for the recognition of human activities (HAR), and eXplainable artificial intelligence (XAI) can elucidate the role of EEG features in ML-based HAR models. The primary objective of this investigation is to investigate the feasibility of an EEG-based ML model for categorizing everyday activities, such as resting, motor, and cognitive tasks, and interpreting models clinically through XAI techniques to explicate the EEG features that contribute the most to different HAR states. The study involved an examination of 75 healthy individuals with no prior diagnosis of neurological disorders. EEG recordings were obtained during the resting state, as well as two motor control states (walking and working tasks), and a cognition state (reading task). Electrodes were placed in specific regions of the brain, including the frontal, central, temporal, and occipital lobes (Fz, C1, C2, T7, T8, Oz). Several ML models were trained using EEG data for activity recognition and LIME (Local Interpretable Model-Agnostic Explanations) was employed for interpreting clinically the most influential EEG spectral features in HAR models. The classification results of the HAR models, particularly the Random Forest and Gradient Boosting models, demonstrated outstanding performances in distinguishing the analyzed human activities. The ML models exhibited alignment with EEG spectral bands in the recognition of human activity, a finding supported by the XAI explanations. To sum up, incorporating eXplainable Artificial Intelligence (XAI) into Human Activity Recognition (HAR) studies may improve activity monitoring for patient recovery, motor imagery, the healthcare metaverse, and clinical virtual reality settings.

Explainable Artificial Intelligence Model for Stroke Prediction Using EEG Signal.

State-of-the-art healthcare technologies are incorporating advanced Artificial Intelligence (AI) models, allowing for rapid and easy disease diagnosis. However, most AI models are considered "black boxes," because there is no explanation for the decisions made by these models. Users may find it challenging to comprehend and interpret the results. Explainable AI (XAI) can explain the machine learning (ML) outputs and contribution of features in disease prediction models. Electroencephalography (EEG) is a potential predictive tool for understanding cortical impairment caused by an ischemic stroke and can be utilized for acute stroke prediction, neurologic prognosis, and post-stroke treatment. This study aims to utilize ML models to classify the ischemic stroke group and the healthy control group for acute stroke prediction in active states. Moreover, XAI tools (Eli5 and LIME) were utilized to explain the behavior of the model and determine the significant features that contribute to stroke prediction models. In this work, we studied 48 patients admitted to a hospital with acute ischemic stroke and 75 healthy adults who had no history of identified other neurological illnesses. EEG was obtained within three months following the onset of ischemic stroke symptoms using frontal, central, temporal, and occipital cortical electrodes (Fz, C1, T7, Oz). EEG data were collected in an active state (walking, working, and reading tasks). In the results of the ML approach, the Adaptive Gradient Boosting models showed around 80% accuracy for the classification of the control group and the stroke group. Eli5 and LIME were utilized to explain the behavior of the stroke prediction model and interpret the model locally around the prediction. The Eli5 and LIME interpretable models emphasized the spectral delta and theta features as local contributors to stroke prediction. From the findings of this explainable AI research, it is expected that the stroke-prediction XAI model will help with post-stroke treatment and recovery, as well as help healthcare professionals, make their diagnostic decisions more explainable.

Quantitative Evaluation of EEG-Biomarkers for Prediction of Sleep Stages.

Electroencephalography (EEG) is immediate and sensitive to neurological changes resulting from sleep stages and is considered a computing tool for understanding the association between neurological outcomes and sleep stages. EEG is expected to be an efficient approach for sleep stage prediction outside a highly equipped clinical setting compared with multimodal physiological signal-based polysomnography. This study aims to quantify the neurological EEG-biomarkers and predict five-class sleep stages using sleep EEG data. We investigated the three-channel EEG sleep recordings of 154 individuals (mean age of 53.8 ± 15.4 years) from the Haaglanden Medisch Centrum (HMC, The Hague, The Netherlands) open-access public dataset of PhysioNet. The power of fast-wave alpha, beta, and gamma rhythms decreases; and the power of slow-wave delta and theta oscillations gradually increases as sleep becomes deeper. Delta wave power ratios (DAR, DTR, and DTABR) may be considered biomarkers for their characteristics of attenuation in NREM sleep and subsequent increase in REM sleep. The overall accuracy of the C5.0, Neural Network, and CHAID machine-learning models are 91%, 89%, and 84%, respectively, for multi-class classification of the sleep stages. The EEG-based sleep stage prediction approach is expected to be utilized in a wearable sleep monitoring system.

Driving-Induced Neurological Biomarkers in an Advanced Driver-Assistance System.

Physiological signals are immediate and sensitive to neurological changes resulting from the mental workload induced by various driving environments and are considered a quantifying tool for understanding the association between neurological outcomes and driving cognitive workloads. Neurological assessment, outside of a highly-equipped clinical setting, requires an ambulatory electroencephalography (EEG) headset. This study aimed to quantify neurological biomarkers during a resting state and two different scenarios of driving states in a virtual driving environment. We investigated the neurological responses of seventeen healthy male drivers. EEG data were measured in an initial resting state, city-roadways driving state, and expressway driving state using a portable EEG headset in a driving simulator. During the experiment, the participants drove while experiencing cognitive workloads due to various driving environments, such as road traffic conditions, lane changes of surrounding vehicles, the speed limit, etc. The power of the beta and gamma bands decreased, and the power of the delta waves, theta, and frontal theta asymmetry increased in the driving state relative to the resting state. Delta-alpha ratio (DAR) and delta-theta ratio (DTR) showed a strong correlation with a resting state, city-roadways driving state, and expressway driving state. Binary machine-learning (ML) classification models showed a near-perfect accuracy between the resting state and driving state. Moderate classification performances were observed between the resting state, city-roadways state, and expressway state in multi-class classification. An EEG-based neurological state prediction approach may be utilized in an advanced driver-assistance system (ADAS).

Prediction of Myoelectric Biomarkers in Post-Stroke Gait.

Electromyography (EMG) is sensitive to neuromuscular changes resulting from ischemic stroke and is considered a potential predictive tool of post-stroke gait and rehabilitation management. This study aimed to evaluate the potential myoelectric biomarkers for the classification of stroke-impaired muscular activity of the stroke patient group and the muscular activity of the control healthy adult group. We also proposed an EMG-based gait monitoring system consisting of a portable EMG device, cloud-based data processing, data analytics, and a health advisor service. This system was investigated with 48 stroke patients (mean age 70.6 years, 65% male) admitted into the emergency unit of a hospital and 75 healthy elderly volunteers (mean age 76.3 years, 32% male). EMG was recorded during walking using the portable device at two muscle positions: the bicep femoris muscle and the lateral gastrocnemius muscle of both lower limbs. The statistical result showed that the mean power frequency (MNF), median power frequency (MDF), peak power frequency (PKF), and mean power (MNP) of the stroke group differed significantly from those of the healthy control group. In the machine learning analysis, the neural network model showed the highest classification performance (precision: 88%, specificity: 89%, accuracy: 80%) using the training dataset and highest classification performance (precision: 72%, specificity: 74%, accuracy: 65%) using the testing dataset. This study will be helpful to understand stroke-impaired gait changes and decide post-stroke rehabilitation.

Quantitative Evaluation of Task-Induced Neurological Outcome after Stroke.

Electroencephalography (EEG) can access ischemic stroke-derived cortical impairment and is believed to be a prospective predictive method for acute stroke prognostics, neurological outcome, and post-stroke rehabilitation management. This study aims to quantify EEG features to understand task-induced neurological declines due to stroke and evaluate the biomarkers to distinguish the ischemic stroke group and the healthy adult group. We investigated forty-eight stroke patients (average age 72.2 years, 62% male) admitted to the rehabilitation center and seventy-five healthy adults (average age 77 years, 31% male) with no history of known neurological diseases. EEG was recorded through frontal, central, temporal, and occipital cortical electrodes (Fz, C1, C2, T7, T8, Oz) using wireless EEG devices and a newly developed data acquisition platform within three months after the appearance of symptoms of ischemic stroke (clinically confirmed). Continuous EEG data were recorded during the consecutive resting, motor (walking and working activities), and cognitive reading tasks. The statistical results showed that alpha, theta, and delta activities are biomarkers classifying the stroke patients and the healthy adults in the motor and cognitive states. DAR and DTR of the stroke group differed significantly from those of the healthy control group during the resting, motor, and cognitive tasks. Using the machine-learning approach, the C5.0 model showed 78% accuracy for the resting state, 89% accuracy in the functional motor walking condition, 84% accuracy in the working condition, and 85% accuracy in the cognitive reading state for classification the stroke group and the control group. This study is expected to be helpful for post-stroke treatment and post-stroke recovery.

Quantifying Physiological Biomarkers of a Microwave Brain Stimulation Device.

Physiological signals are immediate and sensitive to neural and cardiovascular change resulting from brain stimulation, and are considered as a quantifying tool with which to evaluate the association between brain stimulation and cognitive performance. Brain stimulation outside a highly equipped, clinical setting requires the use of a low-cost, ambulatory miniature system. The purpose of this double-blind, randomized, sham-controlled study is to quantify the physiological biomarkers of the neural and cardiovascular systems induced by a microwave brain stimulation (MBS) device. We investigated the effect of an active MBS and a sham device on the cardiovascular and neurological responses of ten volunteers (mean age 26.33 years, 70% male). Electroencephalography (EEG) and electrocardiography (ECG) were recorded in the initial resting-state, intermediate state, and the final state at half-hour intervals using a portable sensing device. During the experiment, the participants were engaged in a cognitive workload. In the active MBS group, the power of high-alpha, high-beta, and low-beta bands in the EEG increased, and the power of low-alpha and theta waves decreased, relative to the sham group. RR Interval and QRS interval showed a significant association with MBS stimulation. Heart rate variability features showed no significant difference between the two groups. A wearable MBS modality may be feasible for use in biomedical research; the MBS can modulate the neurological and cardiovascular responses to cognitive workload.