Biology subjects

Zhong, X. Z.

Publications and source records attributed to Zhong, X. Z..

7 recordsLinked to original sources

bioRxiv2026

Intranasal photobiomodulation as an energy-efficient, largely parameter-insensitive alternative to transcranial photobiomodulation

BackgroundPhotobiomodulation (PBM) is an emerging non-invasive light-based brain stimulation technique that can alter cortical oscillations and is currently being pursued for improving cognition and treating neurological and psychiatric conditions. Nearly all human EEG evidence comes from transcranial PBM (tPBM) applied to the forehead, where light must traverse the scalp and thick skull, requiring protocols to compensate with high surface irradiance. Intranasal PBM (iPBM) can reach the anterior skull base at a fraction of that irradiance and has also been shown to modulate cerebrospinal fluid dynamics, yet it has been studied almost exclusively as an adjunct to tPBM, leaving its cortical effects in isolation, and its energy-efficiency relative to the transcranial route, unknown. ObjectiveTo define the spatiotemporal EEG response to pulsed iPBM delivered alone, determine whether stimulation parameters or individual biology moderate it, and compare the energy-efficiency of iPBM and tPBM in the same participants. MethodsHigh-density EEG was collected from forty-six healthy young adults during pulsed iPBM and tPBM spanning a parameter space of varying wavelengths, pulsation frequencies, and irradiances. Percent change in band power from a within-session pre-stimulation baseline was tested with spatiotemporal cluster-based permutation tests. Linear mixed-effects models with backward elimination assessed stimulation and biological moderators (sex, nostril-to-cortex distance). Energy-efficiency, defined as the percent change in band power per J/cm2 of delivered surface energy, was compared between routes within each subject in delivery route-specific cluster regions of interest (ROI) (Wilcoxon signed-rank tests, Benjamini-Hochberg false discovery rate). ResultsiPBM alone produced significant spatiotemporal clusters in theta, beta, and gamma power, with anterior increases and posterior decreases; no delta or alpha clusters survived correction. Beta and gamma effects appeared at stimulation onset and persisted even after stimulation ended, whereas theta effects strengthened after stimulation ended. No predictor survived elimination in any band, time window, or cluster ROI: response magnitude was independent of wavelength, pulsation frequency, irradiance, sex, and nostril-to-cortex distance. Notably, although iPBM delivered roughly twenty times less surface energy than tPBM ([~]0.6-1.1 vs [~]12-24 J/cm2), it produced EEG changes of similar magnitude, and its energy-efficiency exceeded that of tPBM in seven of eight eligible comparisons, with median iPBM-to-tPBM efficiency ratios of 14-32 (all FDR q<0.05) ConclusionsDelivered in isolation, pulsed iPBM elicits a robust cortical EEG signature closely resembling that of tPBM, is insensitive to the stimulation parameters and individual factors tested, and achieves this at a small fraction of the delivered surface energy. As a result, delivery route, not surface irradiance alone, should be treated as a primary variable in PBM dose reporting and protocol design.

physiology↗

bioRxiv2026

Characterization and Validation of Compressed Sensing for Time-of-Flight MRI Angiography of the Human Brain at 3T and 7T

BackgroundCerebral vasculature is a key biomarker of brain health, and time-of-flight (TOF) magnetic resonance angiography (MRA) provides noninvasive assessment of vascular anatomy. However, conventional TOF-MRA requires long scan times, increasing patient burden and susceptibility to motion artifacts. Compressed sensing (CS) offers a feasible acceleration strategy. PurposeTo quantitatively evaluate CS acceleration in TOF-MRA at 3T and 7T using automated whole-FOV vascular segmentation and semi-automatic segmentation of representative vessels. Study typeProspective Population23 healthy human participants (3T) and 8 healthy human participants (7T). Field Strength/SequenceCS TOF-MRA (CS factors 4 and 8 at 3T; 8 at 7T) was compared against non-accelerated (CS0) TOF-MRA. AssessmentVisual comparison and vascular segmentation were performed using automated whole-FOV methods and semi-automatic segmentation of the posterior cerebral artery and anterior choroidal artery. Statistical TestsContrast-to-noise ratio (CNR), voxel count, and vessel diameter were assessed using two-tailed paired t-tests. ResultsWhole-FOV CNR differed significantly across CS factors at 3T (CS0 > CS4: p < 0.001, d = 0.77; CS0 < CS8: p = 0.008, d = 0.36; CS4 < CS8: p < 0.001, d = 1.11) and 7T (CS0 < CS8: p = 0.002, d = 0.54), with semi-automatic segmentation yielding consistent findings (p < 0.01 for all comparisons). The diameter measurements for segmented vessels are also higher with high CS-factors (PCA 7T: left: p = 0.006, d = 0.93, right: p = 0.045, d = 0.43; AChA 7T: left: p < 0.001, d = 0.66, right: p = 0.009, d = 1.06; PCA 3T: p < 0.001 for all comparison dLeft = 0.52 (CS0 vs. CS4), 0.56 (CS4 vs. CS8), 1.11 (CS0 vs. CS8) and dRight = 0.78 (CS0 vs. CS4), 0.57 (CS4 vs. CS8), 1.17 (CS0 vs. CS8)). Data ConclusionCS shows promise for enhancing clinical applicability of TOF-MRA, with advantages most pronounced at 7T.

bioengineering↗

bioRxiv2025

The fMRI response to transcranial photobiomodulation: the effect of wavelength, irradiance, frequency and skin tone on the BOLD and CBF response in healthy young adults

BackgroundTranscranial photobiomodulation (tPBM) utilizes near-infrared light to penetrate the skull to stimulate neural tissue. However, the in vivo physiological response and the factors influencing this response in the human brain have yet to be understood. MethodsIn this study, we utilize functional magnetic resonance imaging (fMRI) to evaluate the effect of tPBM on the blood-oxygenation (BOLD) and cerebral blood flow (CBF), while varying stimulation parameters such as wavelength, irradiance, and frequency. We further examine the influence of skin tone and sex. We further model the neurovascular interactions underlying the response. ResultsOur results show that the fMRI responses to tPBM is not restrained to the site of irradiation, but quickly spreads to distal sites. Certain regions display an fMRI response sustained after tPBM cessation. Importantly, the responses are dependent on biological and stimulation parameters. Lastly, biophysical modeling revealed a consistent neurovascular coupling-like behaviour underlying these responses. ConclusionEmpirical characterizations of dose dependence are critically important to brain stimulation methods in general but have yet to be demonstrated in most cases. This is the first tPBM study to do just that, establishing the foundation for precision medicine using tPBM, and sets a valuable precedent for the field of brain stimulation.

bioengineering↗

bioRxiv2025

Real-Time EEG Response to Pulsed Transcranial Photobiomodulation in Healthy Young Adults: Effects of Stimulation Parameters and Skin Tone

ObjectivesTranscranial photobiomodulation (tPBM) modulates cortical activity, but how specific stimulation parameters shape the electrophysiological response remains unclear. We tested whether pulsed forehead tPBM produces frequency band-specific EEG changes, whether pulsation frequency entrains endogenous oscillations, and whether wavelength, pulsation frequency, irradiance, sex, and skin tone significantly modulate the response. Materials and MethodsForty-six healthy young adults (24M/22F, 20-32 years) each completed four 12-minute EEG recordings (256-channel; PRE-DURING-POST, 4 min each) with pulsed NIR light delivered to the right forehead. The parameter space included two wavelengths (808 nm, 1064 nm), two pulsation frequencies (10 Hz, 40 Hz), and three irradiances (100, 150, 200 mW/cm2). Forehead skin tone (ITA) and sex were included as biological moderators. EEG band power (delta through gamma) was expressed as percent change from a pre-stimulus baseline. Spatiotemporal cluster-based permutation tests (10,000 permutations, FWER alpha=0.05) identified significant electrode clusters. Linear mixed-effects models with backward elimination and FDR correction (q=0.05) quantified parameter and biological contributions. ResultstPBM produced bilateral, bidirectional EEG changes: frontal gamma enhancement and posterior theta, alpha, and beta suppression, emerging gradually during stimulation and persisting post-stimulation. Wavelength qualitatively shaped the response: 808 nm drove broadband posterior suppression while 1064 nm selectively enhanced frontal gamma. Contrary to entrainment predictions, 10 Hz pulsation produced significantly larger gamma increases and theta reductions than 40 Hz. Sex was the most consistent biological modulator: females showed greater alpha suppression during stimulation and males showed greater posterior beta and gamma suppression. Irradiance and skin tone were not significant. ConclusionsPulsed forehead tPBM produces robust, parameter-dependent EEG changes that are not consistent with neural entrainment. Wavelength determines the character of the cortical response, pulsation frequency modulates its magnitude, potentially through photobiological rather than oscillatory mechanisms, and sex is a significant biological moderator. These findings contribute toward a human-specific foundation for informing optical tPBM parameter selection.

physiology↗

bioRxiv2025

The Link to Oxidative Metabolism Varies across rs-fMRI Metrics: A Whole-Brain Assessment Using Macrovascular Correction

One of the major obstacles to the clinical application of resting-state functional magnetic resonance imaging (rs-fMRI) is the complex nature of its measurements, which limits interpretability. An approach to enhance the interpretability of the rs-fMRI metrics is to link them to more fundamental brain physiology, especially cerebral metabolism. Previous studies have established associations between glucose metabolism (CMRglu) and rs-fMRI measurements. In spite of this, oxidative metabolism (CMRO2) is more closely related to cerebral blood flow (CBF) and thus the BOLD signal, and its relationship with CMRglu is complex. Additionally, most currently published rs-fMRI metrics are uncorrected for macrovascular contribution, which may obscure the neuronal contributions. In this study, we measured resting CMRO2 (along with the oxygen extraction fraction, OEF and cerebral blood flow, CBF) using gas-free calibrated fMRI. We used linear mixed-effects (LME) models to examine associations between CMRO2 and various rs-fMRI metrics before and after macrovascular correction. We found that: 1) significant associations exist between CMRO2 and multiple rs-fMRI metrics, with the strongest association found for the global functional density (gFCD) and the weakest for seed-based functional connectivity (FC); 2) associations with rs-fMRI metrics also varied for OEF and CBF; 3) significant sex differences were observed in the above associations; 4) the use of macrovascular correction substantially strengthened the goodness fit of all LME models examined. This latter improvement further validates the use of macrovascular correction in rs-fMRI. These results provide a framework for linking rs-fMRI metrics to fundamental brain physiology, thus improving interpretability of rs-fMRI measurements. This is the first study to formally link whole-brain MRI-based baseline CMRO2 and rs-fMRI metrics, and helps to push the envelope for rs-fMRI in future clinical applications.

neuroscience↗

bioRxiv2022

Generating dynamic carbon-dioxide from the respiratory-volume time series: A feasibility study using neural networks

In the context of fMRI, carbon dioxide (CO2) is a well-known vasodilator that has been widely used to monitor and interrogate vascular physiology. Moreover, spontaneous fluctuations in end-tidal carbon dioxide (PETCO2) reflects changes in arterial CO2 and has been demonstrated as the largest physiological noise source in the low-frequency range of the resting-state fMRI (rs-fMRI) signal. Increasing appreciation for the role of CO2 in fMRI has given rise to methods that use it for physiological denoising or estimating cerebrovascular reactivity. However, the majority of rs-fMRI studies do not involve CO2 recordings, and most often only heart rate and respiration are recorded. While the intrinsic link between these latter metrics and CO2 led to suggested possible analytical models, they have not been widely applied. In this proof-of-concept study, we propose a deep learning approach to reconstruct CO2 and PETCO2 data from respiration waveforms in the resting state. We demonstrate that the one-to-one mapping between respiration and CO2 recordings can be well predicted using fully convolutional networks (FCNs), achieving a Pearson correlation coefficient (r) of 0.946 {+/-} 0.056 with the ground truth CO2. Moreover, dynamic PETCO2 can be successfully derived from the predicted CO2, achieving r of 0.512 {+/-} 0.269 with the ground truth. Importantly, the FCN-based methods outperform previously proposed analytical methods. In addition, we provide guidelines for quality assurance of respiration recordings for the purposes of CO2 prediction. Our results demonstrate that dynamic CO2 can be obtained from respiration-volume using neural networks, complementing the still few reports in deep-learning of physiological fMRI signals, and paving the way for further research in deep-learning based bio-signal processing.

bioinformatics↗

bioRxiv2020

Variations in the frequency and amplitude of resting-state EEG and fMRI signals in normal adults: The effects of age and sex

Frequency and amplitude features of both resting-state electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) are crucial metrics that reveal patterns of brain health in aging. However, the association between these two modalities is still unclear. In this study, we examined the peak frequency and standard deviation of both modalities in a dataset comprising healthy young (35.5{+/-}3.4 years, N=134) and healthy old (66.9{+/-}4.8 years, N=51) adults. Both age and sex effects were examined using non-parametric analyses of variance (ANOVA) and Tukeys Honest Significant Difference (HSD) post-hoc comparisons in the cortical and subcortical regions. We found that, with age, EEG power decreases in the low frequency band (1-12 Hz) but increases in the high frequency band (12-30 Hz). Moreover, EEG frequency generally shifts up with aging. For fMRI, fluctuation amplitude is lower but fluctuation frequency is higher in older adults, but in a manner that depends on the fMRI frequency range. Furthermore, there are significant sex effects in EEG power (female > male), but the sex effect is negligible for EEG frequency as well as fMRI power and frequency. We also found that the fMRI-EEG power ratio is higher in young adults than old adults. However, the mediation analysis shows the association between EEG and fMRI parameters in aging is negligible. This is the first study that examines both power and frequency of both resting EEG and fMRI signals in the same cohort. In conclusion, both fMRI and EEG signals reflect age-related and sex-related brain differences, but they likely associate with different origins.

physiology↗