Quantitative Assessment of Resting-State for Mild Cognitive Impairment Detection: A Functional Near-Infrared Spectroscopy and Deep Learning Approach

Screening tools for subjective cognitive decline and mild cognitive impairment based on task-state prefrontal functional connectivity: a functional near-infrared spectroscopy study

BACKGROUND

Subjective cognitive decline (SCD) and mild cognitive impairment (MCI) carry the risk of progression to dementia, and accurate screening methods for these conditions are urgently needed. Studies have suggested the potential ability of functional near-infrared spectroscopy (fNIRS) to identify MCI and SCD. The present fNIRS study aimed to develop an early screening method for SCD and MCI based on activated prefrontal functional connectivity (FC) during the performance of cognitive scales and subject-wise cross-validation via machine learning.

METHODS

Activated prefrontal FC data measured by fNIRS were collected from 55 normal controls, 80 SCD patients, and 111 MCI patients. Differences in FC were analyzed among the groups, and FC strength and cognitive scale performance were extracted as features to build classification and predictive models through machine learning. Model performance was assessed based on accuracy, specificity, sensitivity, and area under the curve (AUC) with 95 % confidence interval (CI) values.

RESULTS

Statistical analysis revealed a trend toward more impaired prefrontal FC with declining cognitive function. Prediction models were built by combining features of prefrontal FC and cognitive scale performance and applying machine learning models, The models showed generally satisfactory abilities to differentiate among the three groups, especially those employing linear discriminant analysis, logistic regression, and support vector machine. Accuracies of 92.0 % for MCI vs. NC, 80.0 % for MCI vs. SCD, and 76.1 % for SCD vs. NC were achieved, and the highest AUC values were 97.0 % (95 % CI: 94.6 %-99.3 %) for MCI vs. NC, 87.0 % (95 % CI: 81.5 %-92.5 %) for MCI vs. SCD, and 79.2 % (95 % CI: 71.0 %-87.3 %) for SCD vs. NC.

CONCLUSION

The developed screening method based on fNIRS and machine learning has the potential to predict early-stage cognitive impairment based on prefrontal FC data collected during cognitive scale-induced activation.

Identification of functional near-infrared spectroscopy for older adults with mild cognitive impairment: a systematic review

Objective

Mild cognitive impairment (MCI), a common state of cognitive impairment without significant impairment in daily functioning among older adults, is mainly identified using various neuropsychological tests, clinical interviews, and collateral history with some subjective interferences. This systematic review aimed to investigate the functional near-infrared spectroscopy (fNIRS) features of older adults with MCI compared with those with normal cognitive function to assist in the diagnosis of MCI.

Methods

A literature search was conducted in electronic databases, including PubMed, Web of Science, Embase, and Cochrane Library, up to June 15, 2024. The data on article information (first author and year of publication), participant characteristics, task paradigms, regions of interest (ROIs), fNIRS device attributes, and results related to cerebral oxygenation and hemodynamics were extracted.

Results

Finally, 34 relevant studies were identified, involving 1033 patients with MCI and 1107 age-, sex-, and education-matched controls with normal cognitive function. We found that the studies frequently used working memory-related task paradigms and resting-state measurements. Also, the prefrontal cortex was a primary ROI, and the changes in oxygenated hemoglobin concentration were the most basic research attributes used to derive measures such as functional connectivity (FC), FC variability, slope, and other parameters. However, ROI activation levels differed inconsistently between patients with MCI and individuals with normal cognition across studies. In general, the activation levels in the ROI of MCI patients may be higher than, lower than, or comparable to those in the normal control group.

Conclusion

Research on fNIRS in elderly patients with MCI aims to provide an objective marker for MCI diagnosis. The current findings are mixed. However, these differences can be partly explained with the theoretical support from the interaction of cognitive load theory and scaffolding theory of aging and cognition, taking into account factors such as unspecified MCI subtypes, task difficulty, task design, monitoring duration, and population characteristics. Therefore, future studies should consider definite MCI subtypes, strict and well-designed paradigms, long-term monitoring, and large sample sizes to obtain the most consistent results, thereby providing objective references for the clinical diagnosis of MCI in elderly patients.

Connectome-based prediction of future episodic memory performance for individual amnestic mild cognitive impairment patients

The amnestic mild cognitive impairment progression to probable Alzheimer's disease is a continuous phenomenon. Here we conduct a cohort study and apply machine learning to generate a model of predicting episodic memory development for individual amnestic mild cognitive impairment patient that incorporates whole-brain functional connectivity. Fifty amnestic mild cognitive impairment patients completed baseline and 3-year follow-up visits including episodic memory assessments (e.g. Rey Auditory Verbal Learning Test Delayed Recall) and resting-state functional MRI scanning. Using a multivariate analytical method known as relevance vector regression, we found that the baseline whole-brain functional connectivity features failed to predict the baseline Rey Auditory Verbal Learning Test Delayed Recall scores (r = 0.17, P = 0.082). Nonetheless, the baseline whole-brain functional connectivity pattern could predict the longitudinal Rey Auditory Verbal Learning Test Delayed Recall score with statistically significant accuracy (r = 0.50, P < 0.001). The connectivity that contributed most to the prediction (i.e. the top 1% connectivity) included within-default mode connections, within-limbic connections and the connections between default mode and limbic systems. More importantly, these connections with the highest absolute contribution weight mainly displayed long anatomical distances (i.e. Euclidean distance >75 mm). These 'neural fingerprints' may be appropriate biomarkers for amnestic mild cognitive impairment patients to optimize individual patient management and longitudinal evaluation in a timely fashion.

Resting-State Brain Network Characteristics Related to Mild Cognitive Impairment: A Preliminary fNIRS Proof-of-Concept Study

BACKGROUND

This study investigates the reliability of functional near-infrared spectroscopy (fNIRS) in detecting resting-state brain network characteristics in patients with mild cognitive impairment (MCI), focusing on static resting-state functional connectivity (sRSFC) and dynamic resting-state functional connectivity (dRSFC) patterns in MCI patients and healthy controls (HCs) without cognitive impairment.

METHODS

A total of 89 MCI patients and 83 HCs were characterized using neuropsychological scales. Subject sRSFC strength and dRSFC variability coefficients were evaluated via fNIRS. The study evaluated the feasibility of using fNIRS to measure these connectivity metrics and compared resting-state brain network characteristics between the two groups. Correlations with Montreal Cognitive Assessment (MoCA) scores were also explored.

RESULTS

sRSFC strength in homologous brain networks was significantly lower than in heterologous networks (p < 0.05). A significant negative correlation was also observed between sRSFC strength and dRSFC variability at both the group and individual levels (p < 0.001). While sRSFC strength did not differentiate between MCI patients and HCs, the dRSFC variability between the dorsal attention network (DAN) and default mode network (DMN), and between the ventral attention network (VAN) and visual network (VIS), emerged as sensitive biomarkers after false discovery rate correction (p < 0.05). No significant correlation was found between MoCA scores and connectivity measures.

CONCLUSIONS

fNIRS can be used to study resting-state brain networks, with dRSFC variability being more sensitive than sRSFC strength for discriminating between MCI patients and HCs. The DAN-DMN and VAN-VIS regions were found to be particularly useful for the identification of dRSFC differences between the two groups.

CLINICAL TRIAL REGISTRATION

ChiCTR2200057281, registered on 6 March, 2022; https://www.chictr.org.cn/showproj.html?proj=133808.

Resting-state functional near-infrared spectroscopy in neurodegenerative diseases - A systematic review

OBJECTIVE

To systematically review and summarize alterations found in resting-state activity as measured via functional near-infrared spectroscopy (fNIRS) in neurodegenerative diseases.

BACKGROUND

fNIRS is a novel and emerging neuroimaging method suitable for a variety of study designs. Resting-state is the measure of brain activity in the absence of a task, which has been investigated for yielding information about neurodegenerative diseases, mainly using magnetic resonance imaging. We aimed to systematically review the usage of resting-state fNIRS (rsfNIRS) in neurodegenerative diseases.

INCLUSION CRITERIA

Studies investigating people diagnosed with a neurodegenerative disease and resting-state activity obtained with fNIRS using at least two channels.

METHODS

We searched three databases for publications. After the screening, 16 studies were included in the systematic review. The quality of the studies was assessed, and data were extracted. Data were qualitatively synthesized and in the case of at least 10 similar studies, a meta-analysis was planned.

RESULTS

Most studies investigated Mild cognitive impairment (50%), followed by Alzheimer's disease (25%). Other neurodegenerative diseases encompassed Parkinson's disease, Multiple sclerosis, and Amyotrophic lateral sclerosis. All studies reported oxygenated hemoglobin. Still, studies were heterogeneous in terms of study design, measurement duration, fNIRS device, montage, pre-processing, and analyses. A meta-analysis was not considered possible due to this heterogeneity.

CONCLUSION

rsfNIRS shows promise in neurodegenerative disease, as most studies have observed resting-state alterations when compared to healthy controls. However, inconsistencies across studies limit data comparison and meta-analysis. Hence, we strongly advocate the application of fNIRS reporting guidelines and the establishment of rsfNIRS-specific guidelines. This will ensure reliable and comparable results in future research.

Diagnostic machine learning applications on clinical populations using functional near infrared spectroscopy: a review

Functional near-infrared spectroscopy (fNIRS) and its interaction with machine learning (ML) is a popular research topic for the diagnostic classification of clinical disorders due to the lack of robust and objective biomarkers. This review provides an overview of research on psychiatric diseases by using fNIRS and ML. Article search was carried out and 45 studies were evaluated by considering their sample sizes, used features, ML methodology, and reported accuracy. To our best knowledge, this is the first review that reports diagnostic ML applications using fNIRS. We found that there has been an increasing trend to perform ML applications on fNIRS-based biomarker research since 2010. The most studied populations are schizophrenia (n = 12), attention deficit and hyperactivity disorder (n = 7), and autism spectrum disorder (n = 6) are the most studied populations. There is a significant negative correlation between sample size (>21) and accuracy values. Support vector machine (SVM) and deep learning (DL) approaches were the most popular classifier approaches (SVM = 20) (DL = 10). Eight of these studies recruited a number of participants more than 100 for classification. Concentration changes in oxy-hemoglobin (ΔHbO) based features were used more than concentration changes in deoxy-hemoglobin (ΔHb) based ones and the most popular ΔHbO-based features were mean ΔHbO (n = 11) and ΔHbO-based functional connections (n = 11). Using ML on fNIRS data might be a promising approach to reveal specific biomarkers for diagnostic classification.

Abnormal prefrontal functional network in adult obstructive sleep apnea: A resting-state fNIRS study

To assess prefrontal brain network abnormality in adults with obstructive sleep apnea (OSA), resting-state functional near infrared spectroscopy (rs-fNIRS) was used to evaluate 52 subjects, including 27 with OSA and 25 healthy controls (HC). The study found that patients with OSA had a decreased connection edge number, particularly in the connection between the right medial frontal cortex (MFG-R) and other right-hemisphere regions. Graph-based analysis also revealed that patients with OSA had a lower global efficiency, local efficiency, and clustering coefficient than the HC group. Additionally, the study found a significant positive correlation between the Montreal Cognitive Assessment (MoCA) score and both the connection edge number and the graph-based indicators in patients with OSA. These preliminary results suggest that prefrontal rs-fNIRS could be a useful tool for objectively and quantitatively assessing cognitive function impairment in patients with OSA.

Comparison of the Influence of Dual-Task Activities on Prefrontal Activation and Gait Variables in Older Adults with Mild Cognitive Impairment during Straight and Curved Walking

Background and Objectives: Mild cognitive impairment (MCI) is an early stage of dementia in which everyday tasks can be maintained; however, notable challenges may occur in memory, focus, and problem-solving skills. Therefore, motor-cognitive dual-task training is warranted to prevent cognitive decline and improve cognition in aging populations. This study aimed to determine the influence of such dual-task activities during straight and curved walking on the activities of the prefrontal cortex and associated gait variables in older adults with MCI. Materials and Methods: Twenty-seven older adults aged ≥65 years and identified as having MCI based on their scores (18-23) on the Korean Mini-Mental State Examination were enrolled. The participants performed four task scenarios in random order: walking straight, walking straight with a cognitive task, walking curved, and walking curved with a cognitive task. The activation of the prefrontal cortex, which is manifested by a change in the level of oxyhemoglobin, was measured using functional near-infrared spectroscopy. The gait speed and step count were recorded during the task performance. Results: Significant differences were observed in prefrontal cortex activation and gait variables (p < 0.05). Specifically, a substantial increase was observed in prefrontal cortex activation during a dual task compared with that during a resting-state (p < 0.013). Additionally, significant variations were noted in the gait speed and step count (p < 0.05). Conclusions: This study directly demonstrates the impact of motor-cognitive dual-task training on prefrontal cortex activation in older adults with MCI, suggesting the importance of including such interventions in enhancing cognitive function.

Correlation Between Prefrontal Functional Connectivity and the Degree of Cognitive Impairment in Alzheimer's Disease: A Functional Near-Infrared Spectroscopy Study

Background

The development of Alzheimer's disease (AD) can be divided into subjective cognitive decline (SCD), mild cognitive impairment (MCI), and dementia. Early recognition of pre-AD stages may slow the progression of dementia.

Objective

This study aimed to explore functional connectivity (FC) changes of the brain prefrontal cortex (PFC) in AD continuum using functional near-infrared spectroscopy (fNIRS), and to analyze its correlation with cognitive function.

Methods

All participants underwent 48-channel fNIRS at resting-state. Based on Brodmann partitioning, the PFC was divided into eight subregions. The NIRSIT Analysis Tool (v3.7.5) was used to analyze mean ΔHbO2 and FC. Spearman correlation analysis was used to examine associations between FC and cognitive function.

Results

Compared with HC group, the mean ΔHbO2 and FC were different between multiple subregions in the AD continuum. Both mean ΔHbO2 in the left dorsolateral PFC and average FC decreased sequentially from SCD to MCI to AD groups. Additionally, seven pairs of subregions differed in FC among the three groups: the differences between the MCI and SCD groups were in heterotopic connectivity; the differences between the AD and SCD groups were in left intrahemispheric and homotopic connectivity; whereas the MCI and AD groups differed only in homotopic connectivity. Spearman correlation results showed that FCs were positively correlated with cognitive function.

Conclusions

These results suggest that the left dorsolateral PFC may be the key cortical impairment in AD. Furthermore, there are different resting-state prefrontal network patterns in AD continuum, and the degree of cognitive impairment is positively correlated with reduced FC strength.

Effectiveness assessment of repetitive transcranial alternating current stimulation with concurrent EEG and fNIRS measurement

Transcranial alternating current stimulation (tACS) exhibits the capability to interact with endogenous brain oscillations using an external low-intensity sinusoidal current and influences cerebral function. Despite its potential benefits, the physiological mechanisms and effectiveness of tACS are currently a subject of debate and disagreement. The aims of our study are to (i) evaluate the neurological and behavioral impact of tACS by conducting repetitive sham-controlled experiments and (ii) propose criteria to evaluate effectiveness, which can serve as a benchmark to determine optimal individual-based tACS protocols. In this study, 15 healthy adults participated in the experiment over two visiting: sham and tACS (i.e., 5 Hz, 1 mA). During each visit, we used multimodal recordings of the participants' brain, including simultaneous electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS), along with a working memory (WM) score to quantify neurological effects and cognitive changes immediately after each repetitive sham/tACS session. Our results indicate increased WM scores, hemodynamic response strength, and EEG power in theta and delta bands both during and after the tACS period. Additionally, the observed effects do not increase with prolonged stimulation time, as the effects plateau towards the end of the experiment. In conclusion, our proposed closed-loop scheme offers a promising advance for evaluating the effectiveness of tACS during the stimulation session. Specifically, the assessment criteria use participant-specific brain-based signals along with a behavioral output. Moreover, we propose a feedback efficacy score that can aid in determining the optimal stimulation duration based on a participant-specific brain state, thereby preventing the risk of overstimulation.

Lateralization difference in functional activity during Stroop tasks: a functional near-infrared spectroscopy and EEG simultaneous study

Introduction

Conflict monitoring and processing is an important part of the human cognitive system, it plays a key role in many studies of cognitive disorders.

Methods

Based on a Chinese word-color match Stroop task, which included incongruent and neutral stimuli, the Electroencephalogram (EEG) and functional Near-infrared Spectroscopy (fNIRS) signals were recorded simultaneously. The Pearson correlation coefficient matrix was calculated to analyze brain connectivity based on EEG signals. Granger Causality (GC) method was employed to analyze the effective connectivity of bilateral frontal lobes. Wavelet Transform Coherence (WTC) was used to analyze the functional connectivity of the bilateral hemisphere and ipsilateral hemisphere.

Results

Results indicated that brain connectivity analysis on EEG signals did not show any significant lateralization, while fNIRS analysis results showed the frontal lobes especially the left frontal lobe play the leading role in dealing with conflict tasks. The human brain shows leftward lateralization while processing the more complicated incongruent stimuli. This is demonstrated by the higher functional connectivity in the left frontal lobe and the information flow from the left frontal lobe to the right frontal lobe.

Discussion

Our findings in brain connectivity during cognitive conflict processing demonstrated that the dual modality method combining EEG and fNIRS is a valuable tool to excavate more information through cognitive and physiological studies.

An Isolated CNN Architecture for Classification of Finger-Tapping Tasks Using Initial Dip Images: A Functional Near-Infrared Spectroscopy Study

This work investigates the classification of finger-tapping task images constructed for the initial dip duration of hemodynamics (HR) associated with the small brain area of the left motor cortex using functional near-infrared spectroscopy (fNIRS). Different layers (i.e., 16-layers, 19-layers, 22-layers, and 25-layers) of isolated convolutional neural network (CNN) designed from scratch are tested to classify the right-hand thumb and little finger-tapping tasks. Functional t-maps of finger-tapping tasks (thumb, little) were constructed for various durations (0.5 to 4 s with a uniform interval of 0.5 s) for the initial dip duration using a three gamma functions-based designed HR function. The results show that the 22-layered isolated CNN model yielded the highest classification accuracy of 89.2% with less complexity in classifying the functional t-maps of thumb and little fingers associated with the same small brain area using the initial dip. The results further demonstrated that the active brain area of the two tapping tasks from the same small brain area are highly different and well classified using functional t-maps of the initial dip (0.5 to 4 s) compared to functional t-maps generated for delayed HR (14 s). This study shows that the images constructed for initial dip duration can be helpful in the future for fNIRS-based diagnosis or cortical analysis of abnormal cerebral oxygen exchange in patients.

Diagnosis of Chronic Musculoskeletal Pain by Using Functional Near-Infrared Spectroscopy and Machine Learning

Chronic pain (CP) has been found to cause significant alternations of the brain's structure and function due to changes in pain processing and disrupted cognitive functions, including with respect to the prefrontal cortex (PFC). However, until now, no studies have used a wearable, low-cost neuroimaging tool capable of performing functional near-infrared spectroscopy (fNIRS) to explore the functional alternations of the PFC and thus automatically achieve a clinical diagnosis of CP. In this case-control study, the pain characteristics of 19 chronic pain patients and 32 healthy controls were measured using fNIRS. Functional connectivity (FC), FC in the PFC, and spontaneous brain activity of the PFC were examined in the CP patients and compared to those of healthy controls (HCs). Then, leave-one-out cross-validation and machine learning algorithms were used to automatically achieve a diagnosis corresponding to a CP patient or an HC. The current study found significantly weaker FC, notably higher small-worldness properties of FC, and increased spontaneous brain activity during resting state within the PFC. Additionally, the resting-state fNIRS measurements exhibited excellent performance in identifying the chronic pain patients via supervised machine learning, achieving F1 score of 0.8229 using only seven features. It is expected that potential FC features can be identified, which can thus serve as a neural marker for the detection of CP using machine learning algorithms. Therefore, the present study will open a new avenue for the diagnosis of chronic musculoskeletal pain by using fNIRS and machine learning techniques.

Illuminating neurodegeneration: a future perspective on near-infrared spectroscopy in dementia research

SIGNIFICANCE

Dementia presents a global healthcare crisis, and neuroimaging is the main method for developing effective diagnoses and treatments. Yet currently, there is a lack of sensitive, portable, and low-cost neuroimaging tools. As dementia is associated with vascular and metabolic dysfunction, near-infrared spectroscopy (NIRS) has the potential to fill this gap.

AIM

This future perspective aims to briefly review the use of NIRS in dementia to date and identify the challenges involved in realizing the full impact of NIRS for dementia research, including device development, study design, and data analysis approaches.

APPROACH

We briefly appraised the current literature to assess the challenges, giving a critical analysis of the methods used. To assess the sensitivity of different NIRS device configurations to the brain with atrophy (as is common in most forms of dementia), we performed an optical modeling analysis to compare their cortical sensitivity.

RESULTS

The first NIRS dementia study was published in 1996, and the number of studies has increased over time. In general, these studies identified diminished hemodynamic responses in the frontal lobe and altered functional connectivity in dementia. Our analysis showed that traditional (low-density) NIRS arrays are sensitive to the brain with atrophy (although we see a mean decrease of 22% in the relative brain sensitivity with respect to the healthy brain), but there is a significant improvement (a factor of 50 sensitivity increase) with high-density arrays.

CONCLUSIONS

NIRS has a bright future in dementia research. Advances in technology - high-density devices and intelligent data analysis-will allow new, naturalistic task designs that may have more clinical relevance and increased reproducibility for longitudinal studies. The portable and low-cost nature of NIRS provides the potential for use in clinical and screening tests.

Screening diagnosis of executive dysfunction after ischemic stroke and the effects of transcranial magnetic stimulation: A prospective functional near-infrared spectroscopy study

BACKGROUND

Post-ischemic stroke executive impairment (PISEI) is a serious obstacle for patients to returning to their society and is currently difficult to screen early and clinically ineffective.

AIM

The aim of the study was to clarify whether functional near-infrared spectroscopy (fNIRS) can be used as a rapid screening tool for PISEI and to explore the efficacy of transcranial magnetic stimulation (TMS) in PISEI patients and the changes in brain function.

METHODS

A single-blind, randomized controlled study design was used to detect hemodynamic differences by fNIIRS in 16 PISEI patients and 16 healthy subjects during the resting state and Stroop task, respectively. After 3 days, all subjects received a single TMS intervention and underwent simultaneous fNIRS testing for the Stroop task before and 3 days after the TMS intervention.

RESULTS

PISEI patients had significantly higher HbO₂ content in the left dorsolateral prefrontal cortex (DLPFC), the right pre-motor cortex (PMC) and the right primary sensorimotor cortex (SM1) during the Stroop task compared to the resting state (F = 141.966, p < 0.001), but significantly lower than healthy subjects (T = -3.413, p = 0.002). After TMS intervention, PISEI patients' time and error number scores on the Stroop test were significantly enhanced, and the functional activity of the above-mentioned brain regions was significantly more active than at baseline, while the strength of their functional connections with each other was markedly increased.

CONCLUSIONS

fNIRS helped screen and diagnose PISEI. A single TMS session benefited PISEI patients with effects lasting 3 days, which may be attributed to activation of the left DLPFC, right PMC and right SM1 brain regions.

Disturbed functional connectivity and topological properties of the frontal lobe in minimally conscious state based on resting-state fNIRS

Background

Patients in minimally conscious state (MCS) exist measurable evidence of consciousness. The frontal lobe is a crucial part of the brain that encodes abstract information and is closely related to the conscious state. We hypothesized that the disturbance of the frontal functional network exists in MCS patients.

Methods

We collected the resting-state functional near-infrared spectroscopy (fNIRS) data of fifteen MCS patients and sixteen age- and gender-matched healthy controls (HC). The Coma Recovery Scale-Revised (CRS-R) scale of MCS patients was also composed. The topology of the frontal functional network was analyzed in two groups.

Results

Compared with HC, the MCS patients showed widely disrupted functional connectivity in the frontal lobe, especially in the frontopolar area and right dorsolateral prefrontal cortex. Moreover, the MCS patients displayed lower clustering coefficient, global efficiency, local efficiency, and higher characteristic path length. In addition, the nodal clustering coefficient and nodal local efficiency in the left frontopolar area and right dorsolateral prefrontal cortex were significantly reduced in MCS patients. Furthermore, the nodal clustering coefficient and nodal local efficiency in the right dorsolateral prefrontal cortex were positively correlated to auditory subscale scores.

Conclusion

This study reveals that MCS patients' frontal functional network is synergistically dysfunctional. And the balance between information separation and integration in the frontal lobe is broken, especially the local information transmission in the prefrontal cortex. These findings help us to understand the pathological mechanism of MCS patients better.

Differences in Brain Activity and Body Movements Between Virtual Reality and Offline Exercise: Randomized Crossover Trial

BACKGROUND

Virtual reality (VR) has been suggested to be effective at enhancing physical exercises because of its immersive characteristics. However, few studies have quantitatively assessed the range of motion and brain activity during VR exercises.

OBJECTIVE

We hypothesized that 3D immersive VR could stimulate body movement and brain activity more effectively than standard exercises and that the increased range of motions during 3D immersive VR exercises would be associated with orbitofrontal activation.

METHODS

A randomized crossover trial was conducted to compare exercises with and without VR. A total of 24 healthy males performed the same motions when exercising with and without 3D immersive VR, and the recorded videos were used for motion analysis. Hemodynamic changes in the prefrontal cortex were assessed using functional near-infrared spectroscopy.

RESULTS

There were significant differences in the total angle (z=-2.31; P=.02), length (z=-2.78; P=.005), calorie consumption (z=-3.04; P=.002), and change in accumulated oxygenated hemoglobin within the right orbitofrontal cortex (F^1,94=9.36; P=.003) between the VR and offline trials. Hemodynamic changes in the right orbitofrontal cortex were positively correlated with the total angle (r=0.45; P=.001) and length (r=0.38; P=.007) in the VR exercise; however, there was no significant correlation in the offline trial.

CONCLUSIONS

The results of this study suggest that 3D immersive VR exercise effectively increases the range of motion in healthy individuals in relation to orbitofrontal activation.

TRIAL REGISTRATION

Clinical Research Information Service KCT0008021; https://cris.nih.go.kr/cris/search/detailSearch.do/23671.

Reliable Fast (20 Hz) Acquisition Rate by a TD fNIRS Device: Brain Resting-State Oscillation Studies

A high power setup for multichannel time-domain (TD) functional near infrared spectroscopy (fNIRS) measurements with high efficiency detection system was developed. It was fully characterized based on international performance assessment protocols for diffuse optics instruments, showing an improvement of the signal-to-noise ratio (SNR) with respect to previous analogue devices, and allowing acquisition of signals with sampling rate up to 20 Hz and source-detector distance up to 5 cm. A resting-state measurement on the motor cortex of a healthy volunteer was performed with an acquisition rate of 20 Hz at a 4 cm source-detector distance. The power spectrum for the cortical oxy- and deoxyhemoglobin is also provided.

Screening for Alzheimer's disease using prefrontal resting-state functional near-infrared spectroscopy

INTRODUCTION

Alzheimer's disease (AD) is neurodegenerative dementia that causes neurovascular dysfunction and cognitive impairment. Currently, 50 million people live with dementia worldwide, and there are nearly 10 million new cases every year. There is a need for relatively less costly and more objective methods of screening and early diagnosis.

METHODS

Functional near-infrared spectroscopy (fNIRS) systems are a promising solution for the early Detection of AD. For a practical clinically relevant system, a smaller number of optimally placed channels are clearly preferable. In this study, we investigated the number and locations of the best-performing fNIRS channels measuring prefrontal cortex activations. Twenty-one subjects diagnosed with AD and eighteen healthy controls were recruited for the study.

RESULTS

We have shown that resting-state fNIRS recordings from a small number of prefrontal locations provide a promising methodology for detecting AD and monitoring its progression. A high-density continuous-wave fNIRS system was first used to verify the relatively lower hemodynamic activity in the prefrontal cortical areas observed in patients with AD. By using the episode averaged standard deviation of the oxyhemoglobin concentration changes as features that were fed into a Support Vector Machine; we then showed that the accuracy of subsets of optical channels in predicting the presence and severity of AD was significantly above chance. The results suggest that AD can be detected with a 0.76 sensitivity score and a 0.68 specificity score while the severity of AD could be detected with a 0.75 sensitivity score and a 0.72 specificity score with ≤5 channels.

DISCUSSION

These scores suggest that fNIRS is a viable technology for conveniently detecting and monitoring AD as well as investigating underlying mechanisms of disease progression.

Use of Hypoxic Respiratory Challenge for Differentiating Alzheimer's Disease and Wild-Type Mice Non-Invasively: A Diffuse Optical Spectroscopy Study

Alzheimer's disease is one of the most critical brain diseases. The prevalence of the disease keeps rising due to increasing life spans. This study aims to examine the use of hemodynamic signals during hypoxic respiratory challenge for the differentiation of Alzheimer's disease (AD) and wild-type (WT) mice. Diffuse optical spectroscopy, an optical system that can non-invasively monitor transient changes in deoxygenated (ΔRHb) and oxygenated (ΔOHb) hemoglobin concentrations, was used to monitor hemodynamic reactivity during hypoxic respiratory challenges in an animal model. From the acquired signals, 13 hemodynamic features were extracted from each of ΔRHb and -ΔOHb (26 features total) for more in-depth analyses of the differences between AD and WT. The hemodynamic features were statistically analyzed and tested to explore the possibility of using machine learning (ML) to differentiate AD and WT. Among the twenty-six features, two features of ΔRHb and one feature of -ΔOHb showed statistically significant differences between AD and WT. Among ML techniques, a naive Bayes algorithm achieved the best accuracy of 84.3% when whole hemodynamic features were used for differentiation. While further works are required to improve the approach, the suggested approach has the potential to be an alternative method for the differentiation of AD and WT.

Deep learning in neuroimaging data analysis: Applications, challenges, and solutions

Methods for the analysis of neuroimaging data have advanced significantly since the beginning of neuroscience as a scientific discipline. Today, sophisticated statistical procedures allow us to examine complex multivariate patterns, however most of them are still constrained by assuming inherent linearity of neural processes. Here, we discuss a group of machine learning methods, called deep learning, which have drawn much attention in and outside the field of neuroscience in recent years and hold the potential to surpass the mentioned limitations. Firstly, we describe and explain the essential concepts in deep learning: the structure and the computational operations that allow deep models to learn. After that, we move to the most common applications of deep learning in neuroimaging data analysis: prediction of outcome, interpretation of internal representations, generation of synthetic data and segmentation. In the next section we present issues that deep learning poses, which concerns multidimensionality and multimodality of data, overfitting and computational cost, and propose possible solutions. Lastly, we discuss the current reach of DL usage in all the common applications in neuroimaging data analysis, where we consider the promise of multimodality, capability of processing raw data, and advanced visualization strategies. We identify research gaps, such as focusing on a limited number of criterion variables and the lack of a well-defined strategy for choosing architecture and hyperparameters. Furthermore, we talk about the possibility of conducting research with constructs that have been ignored so far or/and moving toward frameworks, such as RDoC, the potential of transfer learning and generation of synthetic data.

Multi-View Feature Enhancement Based on Self-Attention Mechanism Graph Convolutional Network for Autism Spectrum Disorder Diagnosis

Functional connectivity (FC) network based on resting-state functional magnetic resonance imaging (rs-fMRI) has become an important tool to explore and understand the brain, which can provide objective basis for the diagnosis of neurodegenerative diseases, such as autism spectrum disorder (ASD). However, most functional connectivity (FC) networks only consider the unilateral features of nodes or edges, and the interaction between them is ignored. In fact, their integration can provide more comprehensive and crucial information in the diagnosis. To address this issue, a new multi-view brain network feature enhancement method based on self-attention mechanism graph convolutional network (SA-GCN) is proposed in this article, which can enhance node features through the connection relationship among different nodes, and then extract deep-seated and more discriminative features. Specifically, we first plug the pooling operation of self-attention mechanism into graph convolutional network (GCN), which can consider the node features and topology of graph network at the same time and then capture more discriminative features. In addition, the sample size is augmented by a "sliding window" strategy, which is beneficial to avoid overfitting and enhance the generalization ability. Furthermore, to fully explore the complex connection relationship among brain regions, we constructed the low-order functional graph network (Lo-FGN) and the high-order functional graph network (Ho-FGN) and enhance the features of the two functional graph networks (FGNs) based on SA-GCN. The experimental results on benchmark datasets show that: (1) SA-GCN can play a role in feature enhancement and can effectively extract more discriminative features, and (2) the integration of Lo-FGN and Ho-FGN can achieve the best ASD classification accuracy (79.9%), which reveals the information complementarity between them.

Transfer Learning Approaches for Neuroimaging Analysis: A Scoping Review

Deep learning algorithms have been moderately successful in diagnoses of diseases by analyzing medical images especially through neuroimaging that is rich in annotated data. Transfer learning methods have demonstrated strong performance in tackling annotated data. It utilizes and transfers knowledge learned from a source domain to target domain even when the dataset is small. There are multiple approaches to transfer learning that result in a range of performance estimates in diagnosis, detection, and classification of clinical problems. Therefore, in this paper, we reviewed transfer learning approaches, their design attributes, and their applications to neuroimaging problems. We reviewed two main literature databases and included the most relevant studies using predefined inclusion criteria. Among 50 reviewed studies, more than half of them are on transfer learning for Alzheimer's disease. Brain mapping and brain tumor detection were second and third most discussed research problems, respectively. The most common source dataset for transfer learning was ImageNet, which is not a neuroimaging dataset. This suggests that the majority of studies preferred pre-trained models instead of training their own model on a neuroimaging dataset. Although, about one third of studies designed their own architecture, most studies used existing Convolutional Neural Network architectures. Magnetic Resonance Imaging was the most common imaging modality. In almost all studies, transfer learning contributed to better performance in diagnosis, classification, segmentation of different neuroimaging diseases and problems, than methods without transfer learning. Among different transfer learning approaches, fine-tuning all convolutional and fully-connected layers approach and freezing convolutional layers and fine-tuning fully-connected layers approach demonstrated superior performance in terms of accuracy. These recent transfer learning approaches not only show great performance but also require less computational resources and time.

Transfer learning for non-image data in clinical research: A scoping review

BACKGROUND

Transfer learning is a form of machine learning where a pre-trained model trained on a specific task is reused as a starting point and tailored to another task in a different dataset. While transfer learning has garnered considerable attention in medical image analysis, its use for clinical non-image data is not well studied. Therefore, the objective of this scoping review was to explore the use of transfer learning for non-image data in the clinical literature.

METHODS AND FINDINGS

We systematically searched medical databases (PubMed, EMBASE, CINAHL) for peer-reviewed clinical studies that used transfer learning on human non-image data. We included 83 studies in the review. More than half of the studies (63%) were published within 12 months of the search. Transfer learning was most often applied to time series data (61%), followed by tabular data (18%), audio (12%) and text (8%). Thirty-three (40%) studies applied an image-based model to non-image data after transforming data into images (e.g. spectrograms). Twenty-nine (35%) studies did not have any authors with a health-related affiliation. Many studies used publicly available datasets (66%) and models (49%), but fewer shared their code (27%).

CONCLUSIONS

In this scoping review, we have described current trends in the use of transfer learning for non-image data in the clinical literature. We found that the use of transfer learning has grown rapidly within the last few years. We have identified studies and demonstrated the potential of transfer learning in clinical research in a wide range of medical specialties. More interdisciplinary collaborations and the wider adaption of reproducible research principles are needed to increase the impact of transfer learning in clinical research.

Refined prefrontal working memory network as a neuromarker for Alzheimer's disease

Detecting Alzheimer's disease (AD) is an important step in preventing pathological brain damage. Working memory (WM)-related network modulation can be a pathological feature of AD, but is usually modulated by untargeted cognitive processes and individual variance, resulting in the concealment of this key information. Therefore, in this study, we comprehensively investigated a new neuromarker, named "refined network," in a prefrontal cortex (PFC) that revealed the pathological features of AD. A refined network was acquired by removing unnecessary variance from the WM-related network. By using a functional near-infrared spectroscopy (fNIRS) device, we evaluated the reliability of the refined network, which was identified from the three groups classified by AD progression: healthy people (N=31), mild cognitive impairment (N=11), and patients with AD (N=18). As a result, we identified edges with significant correlations between cognitive functions and groups in the dorsolateral PFC. Moreover, the refined network achieved a significantly correlating metric with neuropsychological test scores, and a remarkable three-class classification accuracy (95.0%). These results implicate the refined PFC WM-related network as a powerful neuromarker for AD screening.