Amyotrophic lateral sclerosis (ALS), also known as Lou Gehrig’s disease, causes rapidly progressive upper and lower motor neuron degeneration, thereby disrupting the ability of the brain to control voluntary motor function leading to dysphagia (disordered swallowing), dysarthria (disordered speech), impaired limb function, poor respiratory function, and ultimately fatality (Kiernan et al., 2011). The disease is categorized by significant across-patient heterogeneity in onset region, pattern, and rate of progression (Ravits et al., 2007). There is currently no universal standard for early detection or for monitoring the progression of ALS (Nzwalo et al., 2014; Malekzadeh, 2021). Due to the lack of a biomarker, patients with ALS are often initially misdiagnosed (up to 45% of the time) and their diagnosis can be delayed up to 12 months (Iwasaki et al., 2001).

Regardless of the focality of motor neuron degeneration at clinical onset, progressive bulbar motor deterioration is common in most patients with ALS, which leads to dysarthria (Green et al., 2013). Thus, the identification of a speech-motor biomarker for early detection of ALS has been an active area of research recently (An et al., 2018; Vieira et al., 2019; Stegmann et al., 2020). The degree to which clinicians can identify speech impairments in ALS using perceptual characteristics of speech (e.g., listening for deviations in articulation, voice quality, resonance, and prosody) is only moderately reliable (Allison et al., 2017). Early detection and monitoring of the progression of bulbar symptoms based on behavioral observations remain limited because oral-motor functional changes may not occur until muscle weakness progresses to a critical level (DePaul and Brooks, 1993; Green et al., 2013). However, physiologically, these subtle symptoms could be identified earlier by quantifying the neural activity pattern changes during speech tasks.

There have been intense investigations for diagnostic and prognostic biomarkers in the brain that can provide evidence for ALS mechanisms and thus novel targets for therapeutic intervention. Studies using functional magnetic resonance imaging (fMRI) have shown evidence of increased functional connectivity in ALS patients (Konrad et al., 2002; Lulé et al., 2007; Verstraete et al., 2010; Agosta et al., 2013). Similar findings have been reported using resting-state electroencephalography (EEG) (Iyer et al., 2015; Fraschini et al., 2016; Dukic et al., 2021) and magnetoencephalography (MEG) (Proudfoot et al., 2018; Sorrentino et al., 2018). Using MEG during a spinal motor task, another study demonstrated intensified cortical beta desynchronization followed by a delayed rebound for participants with ALS (Proudfoot et al., 2017), which hinted that beta-band oscillation may be used as an early distinguishing cortical feature for ALS. Such prior neuroimaging studies have provided tremendously impactful insights toward a better understanding of the major mechanisms of neurodegeneration due to ALS in the attempt to identify a neural biomarker. However, most neuroimaging studies have focused on group-level connectivity analyses during resting-state or spinal motor tasks. How cortical activation is impacted by ALS during speech-motor tasks has not been investigated. In addition, it is unknown if single-trial detection of ALS from neural signals is possible. In theory, single-trial ALS detection could instantly diagnose ALS in real-time thereby strengthening medical treatments for ALS. Classifying ALS on a single-trial basis involves training a machine learning model with multiple samples/trials of a quantifiable objective marker that can efficiently predict a sample/trial as ALS or healthy after proper training. Single-trial detection using machine learning has shown great potential in several neural disorders including major depressive disorder (MDD) (Liu et al., 2022), autism spectrum disorder (ASD) (Ezabadi and Moradi, 2021), post-traumatic stress disorder (PTSD) (Georgopoulos et al., 2010), schizophrenia (Xu et al., 2013), amongst other neurologic disorders (Aoe et al., 2019).

In this study, we investigated cortical differences between healthy and ALS brain signals during overt (involving bulbar motor coordination) and imagined speech (without motor involvement). The assumption is that there is a cortical disturbance during motor functions in the early stage of ALS which has been shown in previous studies (Kew et al., 1993; Mills and Nithi, 1997; Geevasinga et al., 2016; Shibuya et al., 2016; Eisen et al., 2017). Here, we used speech-motor tasks to trigger the disturbance and then detect the presence of ALS using machine learning. To our knowledge, this is the first study to use functional neuroimaging data during speech tasks for ALS detection. We examined cortical differences between healthy controls and patients with ALS using the following approaches: (1) signal correlation across sensors, (2) band power distance estimation for individual neural oscillations, (3) functional connectivity analysis, and (4) single-trial classification of ALS and healthy samples using machine learning. These approaches have been widely used in the literature to examine cortical differences between neurotypical controls and patients with neural disorders (Bob et al., 2010; Proudfoot et al., 2017, 2018; Aoe et al., 2019). Using these approaches, we found significant cortical differences between patients with ALS and healthy controls, particularly in the beta band MEG activity, which we detected at the single-trial level.

Figure 2 shows the comparative histogram distribution of sensor-level signal correlations for ALS and healthy controls for each phrase (top for imagined speech and bottom for overt speech). A significantly larger number of sensors showed greater correlations for ALS compared to healthy controls across all phrases during both overt (one-sided, 2-sample t-test: t = 3.76, df = 28, p < 0.001) and imagined speech (one-sided, 2-sample t-test: t = 6.01, df = 28, p < 0.001). This is also evident by the higher variance in the distribution of correlations for ALS compared to healthy controls for both imagined and overt speech across all phrases. In other words, the majority of the correlations were near mean (i.e., zero correlation) for the controls compared to ALS. For imagined speech, 95% (Bayesian analysis based on Monte Carlo simulations) of the correlation values were in a range of −0.5 to 0.5 for healthy participants. The range was between −0.8 to 0.8 for ALS participants. For overt speech, the correlation range for healthy controls was approximately within the range − 0.8 to 0.8, which was greater for the ALS ranging from −1 to 1. A heatmap plot of the correlation distribution for each subject is shown in Figure 3 (Top for imagined speech and middle for overt speech), which depicts stronger correlations across the whole brain for participants with ALS compared to the healthy controls, especially for the first participant with ALS (A1) who also had the lowest speech intelligibility and speaking rate scores (Table 1). To interpret these correlation heatmaps, correlation density was calculated as the sum of all absolute correlation values over total number of sensor-pairs and shown for each participant in Figure 3—Bottom panel. Mean correlation density was higher for participants with ALS (Overt: 0.428; Imagination: 0.326) compared to healthy subjects (Overt: 0.292; Imagination: 0.192) averaged across trials, phrases, and participants as well as statistically across all phrases and participants (one-sided, 2 sample t-tests: overt: t = 6.13, df = 28, p < 0.001; imagined: t = 6.68, df = 28, p < 0.001). As expected, a stronger correlation for overt speech was observed compared to speech imagination, irrespective of healthy or ALS data.

Figure 4 shows the mean band-power distances between healthy and ALS for each during overt speech task (articulation), where panels (A) to (F) are for delta, theta, alpha, beta, gamma, and high gamma bands, respectively. For better visualization, the distances are shown as heatmaps where the color range from blue to red indicates minimum to maximum range of the normalized distance values. Each cell in the heatmap represents a pairwise distance between the band powers of a healthy sensor (y-axis) and an ALS sensor (x-axis) across all the phrases. The distances were significantly greater for the beta band powers than the other canonical bands for both imagination (1-way ANOVA: F = 208.55, p < 0.001; post-hoc Tukey tests: beta vs. rest: p < 0.0019) and articulation (1-way ANOVA: F = 206.59, p < 0.001; post-hoc Tukey tests: beta vs. rest: p < 0.0021, see Figure 4—Panel G and H). Also, a larger number of pairwise (ALS—healthy) dissimilarities were observed in the beta band. These oscillatory patterns were similar for each phrase and across all phrases for individual and group subject analysis irrespective of speech task, i.e., imagination or production. A couple of sensors showed the highest distance (solid red lines in the heatmap) which could be because those sensors were noisy.

Figure 5 shows the AEC-based beta-band functional connectivity for both groups during the production of the phrase ‘Do you understand me?’ in the form of heatmaps; showing the correlation range of −1 to 1 (from blue [minimum] to red [maximum]). Greater beta band connectivity was significant for ALS patients compared to healthy subjects (one sided, 2-sample t-test: p < 0.05; see Supplementary Figure S3 for distributions of connectivity strengths for both patients with ALS and healthy controls). This is notably apparent in the first patient (A1), who had more severe bulbar impairment than the other two (A2 and A3). Interestingly, similar patterns of increased connectivity were also prominent during speech imagination (Supplementary Figure S2). A more diverse connectivity pattern among the 3 patients with ALS compared to the healthy participants can be observed by visualizing the connectivity strengths.

Figure 6 shows the median single-trial classification accuracy for the healthy versus ALS group for both speech imagination and overt speech tasks. The best performance (median accuracy ~98%) was obtained using beta bands and was similar for both speech tasks. The performance using each individual frequency range of interest (excluding delta) was significantly higher than chance level (50%) and was also higher when compared to performance using all frequency information (all: 0.3–250 Hz). The distribution of the test performance for different folds (i.e., for each pair of ALS-healthy single-trial test accuracy) is shown in Supplementary Figure S4. The performance accuracy was lowest for the first ALS patient (A1) (mean across folds = 65% for overt speech; 83% for imagined speech), likely because this participant’s speech symptoms were severe compared to the other two participants with ALS. Although median performance was highest for beta band, statistically, 1-way ANOVA based comparison did not show a significant difference between the performances of different bands (F = 1.33, p = 0.05), possibly due to the low sample size. For the case of imagined speech, performances obtained with theta and gamma band were comparable to beta band performance.

The evidence of greater inter-sensor correlation for ALS compared to healthy participants is a clear distinguishable marker between the two groups. This has been previously observed with M/EEG resting state (Proudfoot et al., 2019) and motor imagery studies (Yang et al., 2018). This difference in sensor correlations was apparent across all phrases and participants which further illustrates that this feature is independent of stimuli and an across-subject observation. A stronger correlation during the overt speech task compared to the imagined speech task indicated greater cortical activity for producing overt speech compared to speech imagination, which was true for both healthy and ALS groups and was expected. The signal artifacts introduced by movement during the production of speech gestures could have also contributed to the higher sensor correlation during overt speech production (Dash et al., 2018). Participant (A1) with the most severe symptoms (lowest speech intelligibility and speaking rate scores: Table 1) showed the strongest correlation (Figure 3) suggesting that the proposed approach may be useful as a marker of disease progression. We must note that a larger sample size is needed to statistically validate this observation. Further, sensor correlation differences could also arise from the differences in the head positions inside the scanner. Mapping the sensor data into source space and performing the correlations across parcels/voxels would be a better way to remove these confounds, as planned for future studies.

Beta band has been traditionally associated with motor function, and the observed differences in the beta band power during overt speech are consistent with the hypotheses that ALS is associated with cortical hyperexcitability, possibly due to the loss of inhibitory interneuron (Proudfoot et al., 2017). Our results reproduced the importance of beta band for identifying ALS during a bulbar motor task (speech). The prominent beta band differences during the speech imagination task (which does not involve motor execution) (Supplementary Figure S1) suggest that the beta band during speech tasks that involve motor planning could be a potential neural biomarker of ALS. Crucially, the pattern of band-power differences was similar for both imagination and overt speech in the beta band, possibly indicating a functional similarity between the two speech tasks. Clear differences in band power were also observed in the theta and the gamma band; however, they were less prominent compared to the beta band differences.

This finding of significantly greater beta band connectivity in the ALS group compared to healthy controls was expected since beta band functional connectivity changes have been previously shown (Verstraete et al., 2010; Agosta et al., 2013; Proudfoot et al., 2017) both during resting state as well as for spinal motor tasks. An increase in beta band functional connectivity has been hypothesized as the result from loss of intracortical inhibitory influence supported in vivo by neurophysiology findings of accentuated cortical beta-desynchronization during movement preparation and diminished post-movement beta-rebound (Proudfoot et al., 2017). This inhibitory influence may lead to compensatory mechanisms in early-stage ALS resulting in higher functional connectivity. Behaviorally, it may be explained as a compensatory mechanism for speech tasks in early-stage ALS due to the recruitment of larger neural networks, supported by tongue kinematic studies (Green et al., 2013; Kuruvilla-Dugdale and Mefferd, 2017; Teplansky et al., 2019). Additionally, disruption of efficient motor control networks in ALS may lead to higher cognitive control demands and attention increases, both of which are known to modulate beta-band oscillation power and connectivity (Cheyne and Ferrari, 2013; Riddle et al., 2021). This study provides the first evidence of increased beta band connectivity during a speech-motor task. Interestingly, increased connectivity was also prominent during speech imagination (Supplementary Figure S2) indicating higher beta band connectivity during speech planning, thereby strengthening the role of beta band as a neural biomarker for ALS. Similar to the observations with band-power differences, beta-band functional connectivity was greatest for our most severe patient (participant A1), providing additional confidence in the specificity of this marker. Accumulating the connectivity strength (i.e., correlation values) of all three subjects and for all 5 phrases showed greater connectivity strength for the ALS group than the control group (Supplementary Figure S2) indicating beta-band connectivity to be an across-subject marker.

Evidence of a high single-trial ALS detection accuracy with beta-band suggests that the neural mechanisms for ALS could be specific to spectral content, particularly to the beta band during speech tasks. From the previous qualitative analyses (sensor correlation, band power difference, and functional connectivity) beta-band was expected to perform the best for single-trial classification. The median accuracy with beta band was superior when both overt and imagined phrases were considered, although theta and gamma band also showed comparable performance with beta band for the case of imagined phrases. A recent study suggested covert speech emphasizes both beta and gamma band (Moon et al., 2022), which may explain why gamma band also obtained high accuracies. In short, greater performance accuracies during the speech imagination task suggest that neural signals derived while imagining speech may be optimal for diagnosing early-onset ALS, whereas overt speech may be more appropriate for evaluating the rate of disease progression. Crucially, this is the first demonstration of ALS detection from single-trial neural signals.

In terms of behavior, individuals with ALS exhibited larger onset latency and duration in overt speech tasks when compared to their healthy counterparts (2-sample t-tests, t = 3.09, p = 0.002, N = 900 [3 participants × 5 phrases × 60 trials]; Supplementary Figure S5), as one would expect the patients to take longer time to complete the task. It is plausible that these behavioral effects manifest in elevated sensor correlation and functional connectivity strength for ALS patients as opposed to healthy controls. However, there is notable convergence in these behaviors at the single-trial level between the two population groups, with more than 44% overlap in onset time and over 22% overlap in duration. The behavioral difference was mostly driven by the first ALS participant (A1) with the lowest speaking rate and speech intelligibility. Consequently, relying solely on behavioral indicators for single-trial detection proves to be inefficient. In addition, similar cortical differences were also observed during the covert speech task, a scenario where these behavioral markers are absent. Further, covert speech segments are immune to movement artifacts that can be present during overt speech and bias the results. Hence, the optimal approach for single-trial ALS detection involves analyzing neural activity during covert speech tasks.

Although these results are encouraging, this study suffers from a very small sample size and the omission of a non-ALS clinical control group. Future studies should include larger cohorts and include another patient population with a movement disorder, e.g., Parkinson’s disease and other motor neural diseases, in order to reveal the specificity of these detection methods. If validated, neuromagnetic signals during speech tasks with machine learning would open a new direction for assisting the diagnosis of ALS. As bulbar onset of ALS represents about 30% of the total and spinal onset accounts for about 70% (Van Es et al., 2017), we plan to combine neural signals during speech and spinal motor tasks (e.g., finger tapping) in future studies. A further step is to combine neuromagnetic signals with (speech) audio (An et al., 2018). Finally, individuals with ALS and other neurological diseases that show some similar symptoms such as Parkinson’s disease will be included for differential analysis.

In this study, we investigated the neuromagnetic pattern differences between individuals with ALS and healthy subjects during imagined and overt speech tasks, towards identifying a potential neural biomarker. Our preliminary results showed a greater number of sensors with larger correlations, a higher dissimilarity in the beta band power, and a larger beta band connectivity for ALS patients compared to healthy controls. Single-trial ALS detection analysis resulted in the highest median classification accuracy using beta band features, which were significant across trials, phrases, and participants for both speech imagination and articulation. The preliminary results of this study provide a proof of concept for the use of beta band as a potential neural biomarker during speech tasks and machine learning for early detection of ALS.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by IRBs of University of Texas at Austin and University of Texas at Dallas. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

DD implemented the algorithms and drafted the manuscript. PF and JW designed the experimental paradigm for data collection. DH performed ALS diagnosis. DD, KT, PF, AB-F, CC, DH, SA, and JW interpreted the results and performed subsequent editing. All authors contributed to the article and approved the submitted version.