Neural State Analytics
Nikolai Kapralov
In M/EEG analyses, it is often convenient to extract a time series of activity from the brain regions of interest (ROI). However, multiple pipelines can be used for extraction of ROI activity, and there is no consensus on the gold standard among existing pipelines. My Ph.D. project addresses this problem via several complementary approaches. First, we performed a multiverse analysis to compare different pipelines using the same dataset and illustrate how the choice of the pipeline affects the estimated values of SNR and connectivity (Kapralov et al., 2024, J Neural Eng). Second, we used cross-talk function (CTF) to quantify the amount of remaining field spread and observed that its effect on the estimation of activity and connectivity is non-uniform across the cortex (Kapralov et al., in prep.). Currently, we investigate how CTF can be used to optimize the extraction of ROI activity for template and individual head models. Overall, our results illustrate the challenges associated with extraction of ROI activity and provide solutions that can be fine-tuned to the needs of performed analysis.
Beta oscillations (~13–30 Hz) are crucial for sensorimotor and cognitive processes, yet their exact functional role remains debated. In this project I will investigate beta bursts features in sensorimotor system. Using magnetoencephalography (MEG) and ultra-high-field 7T MRI, I will explore how distinct beta burst types relate to functional roles and how burst features correlate with cortical myelination content. This will improve our understanding of beta bursts and their potential as biomarkers for neurological conditions such as Parkinson’s disease.
The interaction between brain and body is increasingly being recognized as an important feature for our understanding of cognitive functions, emotions, and overall well-being. In my research, I investigate electrophysiological markers for cardiac activity, primarily heartbeat evoked potentials (HEP) and the methods that are used to study them. Since HEPs are prone to being confounded by heartbeat-unrelated activity, part of my work is focused on the development of robust methods for their analysis (Steinfath et al., 2025, IMAG). In addition, since the HEP field is challenged by variable findings, we (Maria Azanova, Nikolai Kapralov, and I) asked ourselves the question of whether this heterogeneity could be related to the diversity of methods used. To answer this question, we systematically reviewed 132 HEP studies with a focus on the methods used from data acquisition to statistical analysis (Steinfath, Azanova, Kapralov et al., 2025, BioRxiv). Lastly, we empirically investigate the influence of data analysis choices on HEP and the HEP to anxiety association in a multiverse analysis of a large population-based dataset (together with Maria Azanova, in prep.).