Biosignals

This study aims to present a simplified and resource-efficient computational model for predicting activity-dependent conduction velocity changes in unmyelinated axons, serving as a complementary tool to Hodgkin-Huxley models. Our approach is based on the concept of ‘memory’, where the speed of action potentials is modulated by prior activity. We utilized microneurography data from 95 mechano-insensitive C-fibres of healthy human participants, including both sexes, across various stimulation protocols to optimize model parameters. The model incorporates linear long-term and non-linear short-term memory components, effectively predicting propagation speed by convolving the history of recorded action potentials with the memory function. The proposed one-dimensional and two-dimensional memory functions yielded low mean squared errors in predicting the propagation speed of subsequent action potentials. This computational framework provides insights into dynamics of unmyelinated axons under varying conditions, enhancing our understanding of signal processing along the axon and its short-term memory capabilities. Additionally, our model demonstrates rapid computation times suitable for real-time applications in electrophysiological experiments. This study introduces a novel model that simulates activity-dependent conduction velocity changes in unmyelinated axons, which is crucial for effective signal processing during conduction. Unlike Hodgkin-Huxley models that are computationally intensive and complex, our approach leverages fibre ‘memory’ to capture how prior activity influences conduction. With fewer parameters required to fit diverse datasets, including patient data, our highly efficient model enables faster simulations than Hodgkin-Huxley models and facilitates the analysis of spike train propagation over long distances, and it is therefore suitable for modelling peripheral axons that extend up to 1 m.

Introduction:

Traditional pharmaco-electroencephalography (EEG) studies have mainly examined the effects of psychotropic medications at the level of individual drugs or broad drug classes, limiting biological specificity and clinical translation. This study aimed to determine whether modeling EEG spectral power changes according to the engagement of distinct neurotransmitter systems provides a more mechanistic understanding of psychotropic drug effects in a real-world clinical population.

Methods:

We analyzed 4,128 EEG sessions from 2,083 patients in the Temple University Hospital EEG Corpus, a large heterogeneous dataset. EEG data were preprocessed and segmented into canonical frequency bands (delta, theta, alpha, beta, and gamma). Psychotropic medication data were systematically extracted and coded at the receptor level for serotonin, dopamine, norepinephrine, histamine, and acetylcholine systems using the Neuroscience-based Nomenclature framework. Receptor profiles were summarized to represent each patient’s overall neurotransmitter engagement (agonistic, neutral, antagonistic, or mixed). Linear mixed-effects models were applied to assess relationships between neurotransmitter profiles and log-transformed spectral power while controlling for electrode location and patient-level variability.

Results:

Frequency- and region-specific EEG patterns were identified across neurotransmitter systems. Dopamine antagonists were associated with higher delta and theta power at central electrode locations and lower alpha power at occipital and temporal locations, whereas dopamine agonists were associated with higher delta activity at occipital locations and increased frontal gamma power. Serotonin antagonists showed associations with elevated slow-wave and alpha power, while serotonin agonists were linked to increased frontal alpha, decreased occipital alpha, and enhanced temporal gamma power. Both norepinephrine antagonists and agonists showed positive relationships with delta power, with a broader topographical pattern for antagonists. Theta power was positively associated with norepinephrine antagonists and negatively associated with norepinephrine agonists. Norepinephrine antagonists were related to lower temporal alpha and higher frontal and parietal gamma power. Histamine antagonists and mixed histaminergic agents were associated with lower delta, theta, and alpha power. Acetylcholine antagonists were linked to higher delta, theta, and alpha power across electrode locations.

Discussion:

Modeling psychotropic medication effects on EEG at the neurotransmitter receptor level offers a biologically grounded and clinically relevant improvement over traditional drug class-based approaches. This neurotransmitter-centric framework enhances mechanistic interpretability and may support the development of EEG biomarkers for personalized, mechanism-based psychiatric care.

Sorting spikes from noisy single-channel in-vivo extracellular recordings is challenging, particularly due to the lack of ground truth data. Microneurography, an electrophysiological technique for studying peripheral sensory systems, employs experimental protocols that time-lock a subset of spikes. Stable propagation speed of nerve signals enables reliable sorting of these spikes. Leveraging this property, we established ground truth labels for data collected in two European laboratories and designed a proof-of-concept open-source pipeline to process data across diverse hardware and software systems. Using the labels derived from the time-locked spikes, we employed a supervised approach instead of the unsupervised methods typically used in spike sorting. We evaluated multiple low-dimensional representations of spikes and found that raw signal features outperformed more complex approaches, which are effective in brain recordings. However, the choice of the optimal features remained dataset-specific, influenced by the similarity of average spike shapes and the number of fibers contributing to the signal. Based on our findings, we recommend tailoring lightweight algorithms to individual recordings and assessing the “sortability feasibility” based on achieved accuracy and the research question before proceeding with sorting of non-time-locked spikes in future projects.

This work aims to improve FAIR-ness of the microneurography research by integrating the local (meta)data to existing research data infrastructures. In the previous work, we developed an odML based solution for local metadata storage of microneurography data. However, this solution is limited to a narrow community. As a next step, we propose the integration into the Local Data Hubs, data-sharing services within NFDI4Health infrastructure. We outline a first concept, that streams chosen data from the established odMLtables GUI.