Biology subjects

Miller, G. A.

Publications and source records attributed to Miller, G. A..

5 recordsLinked to original sources

bioRxiv2026

Coupled neuro-fluid dynamics in human bioelectromagnetism

Electroencephalography (EEG) and magnetoencephalography (MEG) are widely interpreted as direct, non-invasive measures of neural population activity, assuming surrounding tissue and cerebrospinal fluid (CSF) act as passive volume conductors that distorts but do not contribute to signal generation. Using conductive-fluid phantom experiments, structural magnetic resonance imaging, and in vivo EEG and MEG recordings, the present report challenges this assumption. Mechanically driven conductive fluid generated structured, oscillatory, source-localizable electromagnetic signals resembling M/EEG in the absence of neural tissue, and interindividual differences in CSF volume systematically moderated M/EEG signal amplitude. These findings suggest that, at the scales accessed by invasive and non-invasive electrophysiology, electromagnetic and mechanical processes form a coupled system in which contributions of neural currents and fluid motion cannot readily be uniquely separated.

neuroscience↗

bioRxiv2025

Simulation and empirical evaluation of biologically-informed neural network performance

Biologically-informed neural networks (BiNNs) offer interpretable deep learning models for biological data, but the dataset characteristics required for strong performance remain poorly understood. For instance, we previously developed P-NET, a BiNN with an architecture based on the Reactome pathway database, and applied this model to predict metastatic status of patients with prostate cancer using somatic mutation and copy number information. It seems likely that including additional relevant signal - e.g., germline variation in this context - should improve model performance, but we currently lack a principled approach to assess whether BiNNs will successfully detect this signal. Here, we developed two simulation frameworks to evaluate the factors that influence BiNN performance - including signal type, signal strength, feature sparsity, and sample size - and empirically tested how integrating germline and somatic data affects the models ability to predict prostate cancer metastatic status. Simulations revealed that small sample size, weak signal strength, and especially extreme feature sparsity limit BiNN performance, and that the model preferentially uses linear over nonlinear signal. Empirically, P-NET performed poorly on sparse germline data, and while adding germline to somatic data did not improve prediction, it improved gene prioritization and model interpretation. Broadly, our simulation frameworks enable systematic evaluation of how dataset-level characteristics affect BiNN performance and provide a principled framework for benchmarking novel methods.

bioinformatics↗

bioRxiv2025

Resting state periodic and aperiodic brain oscillations from birth to preschool years: Aperiodic maturity predicts developmental course

The non-invasive assessment of resting-state (RS) neural activity via electrophysiology provides information on brain function and brain health. Our understanding of RS neural activity in older children and adults is limited by our poor understanding of the maturation of RS activity (oscillatory and non- oscillatory) from infancy to preschool ages. The present study used MEG with source imaging and an adapted Dark Room eyes-open task to assess oscillatory and non-oscillatory RS activity. 107 typically developing children 2 to 68 months were enrolled. For the Dark-Room eyes-open task, each child alternated between viewing an Inscapes video without audio for 20s and then resting with their eyes open for 30s in total darkness. This was repeated for 6 cycles. Whole-brain RS activity maps were computed using Minimum Norm Estimates, with RS power spectra divided into estimates of periodic measures (dominant frequency and power) and aperiodic measures (the exponent - slope of the 1/f function; the offset - vertical displacement of the 1/f function). An infant dominant peak was observed more often in the Dark-Room (94% of children) than in the Video-On condition (83% of children). The Dark-Room condition elicited a 36% increase in dominant oscillation activity than the Video-On condition. The maturation of the parietal-occipital periodic dominant frequency increased non-linearly as a function of age. The maturation of aperiodic measures decreased nonlinearly as a function of age, with aperiodic measures as well as their maturation rate differing across the brain. Finally, more mature aperiodic values predicted better adaptive behaviors and daily living skills. Present findings demonstrate that (1) the use of an appropriate Dark-Room eyes-open task provides measures of young child RS periodic activity with excellent SNR, (2) an understanding of the development of infant RS activity is best achieved via obtaining measures in brain source space in order to detect regional differences in aperiodic activity, and (3) a more mature aperiodic value predicts higher developmental behavior scores.

neuroscience↗

bioRxiv2021

Alpha oscillations link action to cognition: An oculomotor account of the brain's dominant rhythm

Power modulations in alpha oscillations (8-14Hz) have been associated with most human cognitive functions and psychopathological conditions studied. These reports are often inconsistent with the prevailing view of a specific relationship of alpha oscillations to attention and working memory (WM). We propose that conceptualizing the role of alpha oscillations in oculomotor control resolves this inconsistency. This proposition is based on a review of results across species (human Npooled=295, one non-human primate, honey bee N=5), experimental conditions (rest, attention, and working memory), and recording techniques (EEG, ECOG, eye-tracking, and MEG) that encourage the following relationships between alpha oscillations and eye-movement control: (i) saccade initiation prompts power decrease in brain circuits associated with saccadic control; (ii) the direction of a saccade is consistent with alpha lateralization, both during task and resting conditions; (iii) the phase of alpha activity informs saccade occurrence and biases miniature eye movements during fixation (e.g. fixational tremor); and (iv) oculomotor action differentiates WM load. A new theory on how alpha oscillations link oculomotor action to cognition is proposed. Generalizing across tasks and species: low oculomotor activity is associated with high alpha power and vice versa. Alpha oscillations regulate how long to look at a given target and how fast to saccade to a next. By ensuring steady gaze position, any potential input outside foveal vision is "suppressed".

neuroscience↗

bioRxiv2017

Meta-analysis and Commentary: Preemptive Correction of Arteriovenous Access Stenosis

BackgroundA recent meta-analysis by Ravani and colleagues (Ravani, P., et al., Am J Kidney Dis, 2016. 67(3): p. 446-60.) studied the effect of pre-emptive correction of arterio-venous dialysis vascular access versus deferred care, based on data from 11 trials. The authors reported a non-significant protective treatment effect of pre-emptive correction on access loss, while showing a significant protective effect on thrombosis rates conferred by pre-emptive correction. We revisit this analysis, including data extraction and effects of a heterogenous study population.\n\nMethodsWe repeated data extraction from all referenced publications in the meta-analysis by Ravani et al. and corrected event counts where applicable. We repeated the meta-analyses with access loss as the outcome for studies that recruited patients with arterio-venous fistulae (AVF) and grafts (AVG), respectively, using a random effects model with relative risk (RR) and risk difference (RD) of access loss as the outcomes of interest. We repeated data extraction from all referenced publications, and corrected event counts where applicable.\n\nResultsOur conclusions differ from the original findings in two ways. First, after some amendment of the event counts extracted from Mayer et al. (Vascular and Endovascular Surgery 1993), we find a significant overall positive effect of pre-emptive correction on arterio-venous access loss in the overall study population [RR 0.80 (95% CI 0.64 to 0.99), RD -0.07 (95% CI -0.12 to -0.02); Figure 1]. Secondly, we highlight the impact of heterogeneous study populations on the meta-analysis. Whereas the data do not conclusively show a benefit of pre-emptive correction for arteriovenous grafts (AVG; RR = 0.87, 95% CI: 0.69 - 1.11), they show a strong protective effect for arteriovenous fistulae (AVF; RR = 0.5, 95% CI: 0.29 to 0.86).\n\nO_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=128 SRC=\"FIGDIR/small/179580_fig1.gif\" ALT=\"Figure 1\">\nView larger version (41K):\norg.highwire.dtl.DTLVardef@af3c13org.highwire.dtl.DTLVardef@911a50org.highwire.dtl.DTLVardef@568cf1org.highwire.dtl.DTLVardef@3f3c08_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOFigure 1:C_FLOATNO Meta-analysis of access loss, overall and by access type using risk ratio (RR) as the measure of association.\n\nC_FIG\n\nDiscussion and ConclusionThese findings corroborate clinical arguments such as superior long-term patency of AVF and the nature of AVG failure that often involve infectious causes. The available data indicate mild or no benefit of pre-emptive correction for AVG, but strongly support tight monitoring of dialysis accesses and preemptive intervention and correction upon the slightest suspicion of access stenosis for AVF.

epidemiology↗