Home

Electrophysiology

Since April 2020, I have been working with Dr Susanna Mierau in the Synapse and Network Development Group affiliated with the Department of Neurology at Brigham and Women's Hospital and Harvard Medical School (Boston, Massachusetts, USA) and the Department of Physiology, Development and Neuroscience at the University of Cambridge (Cambridge, England).

The group's research focuses on synaptic and network maturation in cortical circuits during early postnatal development and how this process is disrupted in neurodevelopmental disorders including Rett syndrome and autism spectrum disorder (ASD). We have expanded our work with cellular-scale networks from Mecp2-deficient mice to recording network activity from human cerebral organoids. Our ultimate goal is to identify new targets for circuit-based pharmacologic therapies for Rett syndrome and genetic causes of ASD.

Since joining the group as an undergraduate student, I have been primarily involved in developing computational methods for the analysis of microelectrode array (MEA) recordings. These included:

• Image analysis: Estimating MEA chip coverge and cellular network density from 2D culture microphotographs. Collaboration with Leo Nagy (Cambridge Clinical School, UK) and Richard Turner (PDN, Cambridge, UK).
• Spike detection: WATERS - A novel neuronal spike detection method based on continuous wavelet transform with automatic data-driven template selection. Validated in 2D in cortical neuronal cultures and 3D human cerebral and spinal cord organoids. Benchmarked on synthetic data, achieving superior performance to currently used spike detection methods, especially under low signal-to-noise ratio (SNR) conditions. Collaboration with Tim Sit (Sainsbury Wellcome Centre, UCL, UK), Andrey Vinogradov (Tampere University, Finland) and Alex Dunn (PDN, Cambridge, UK).
• Functional connectivity: High-performance MEX implementation of Spike Time Tiling Coefficient and probabilistic thresholding of adjacency matrices. Includes options for parallel computing. Collaboration with Hugo Smith (Cambridge Clincal School, UK).
• Dimensionality analysis of network dynamics using effective rank. Extending the existing methodology from correlation matrices to spatiotemporal neuronal activity. Benchmarking the two approaches and examining the effects of applying variance stabilising transformation. Tracking changes in complexity and synchronicity of network patterns of activity during neurodevelopment in healthy and Mecp2-deficient networks. Collaboration with Tim Sit (Sainsbury Wellcome Centre, UCL, UK).
• Network control theory applied to cellular-scale network dynamics. Based on seminal work on controllability of complex networks and its applications to network neuroscience. Extended the average and modal controllability measures to functional connectivity in 2D cortical cultures, and introduced a novel metric based on the volume of controllability Gramian (feasibility of driving the network dynamics to arbitrary states with unit input energy). Collaboration with Tim Sit (Sainsbury Wellcome Centre, UCL, UK).
• Neuronal activity in cortical-spinal neuraxis organoid constructs. I performed the analysis of MEA recordings from hybrid functional constructs of human air-liquid interface cerebral organoids (ALICOs) and spinal cord organoids. Examined the responsivity and effects of electrical stimulation, and the directionality of functional connections between the components of the constructs. Courtesy of Dr George Gibbons and Lakatos Group (Departament of Clinical Neurosciences, Cambridge).
• I also help with managing the group's website and codebase.

Most of my contributions made their way to our MEA network analysis pipeline (MEA-NAP). MEA-NAP is a streamlined diagnostic and analytic tool for cellular-scale network activity data obtained using microelectrode arrays. It combines methods for studying network function using graph theoretical and other network metrics that are commonly applied at the whole brain level (e.g., fMRI data) and in other fields of network science.