Exploring the interactions between parietal cortex and the midbrain during perceptual decision-making
Complex behavior is supported by cooperation among several brain areas. My thesis work focuses on the interactions between two brain areas — the lateral intraparietal area (LIP) and the midbrain superior colliculus (SC) — both of which are nodes in a network of visuomotor association areas. I record from these areas simultaneously as animals make decisions about a random-dot-motion stimulus and report their decision with an eye movement. I use high-density linear probes, including neuropixels probes and v-probes, to record from retinotopically-aligned neural populations in each area. These recordings allow me to characterize the differences in decision-related activity within each area and the ways in which they interact as the animal is forming and terminating its decision. I also use interventions like chemical inactivation to test hypotheses about each areas’ computational role in these decisions.
Paper: https://www.cell.com/neuron/fulltext/S0896-6273(23)00400-2
Identifying the strategies that psychophysical subjects use in decision-making tasks
Many tasks used to study decision-making encourage subjects to integrate evidence over time. Such tasks are useful to understand how the brain operates on multiple samples of information over prolonged timescales, but only if subjects actually integrate evidence to form their decisions. This project explored the behavioral observations that corroborate evidence-integration in a number of task-designs. I found that several commonly accepted signs of integration were also predicted by non-integration strategies. Furthermore, an integration model could fit data generated by non-integration models. I then identified the features of non-integration models that allowed them to mimic integration and used these insights to design a motion discrimination task that disentangled the models. This work was published in eLife in April 2020: https://elifesciences.org/articles/55365.
Revealing single-trial decision variables in area LIP with large-scale neural recordings
Decisions are well-characterized by models of noisy evidence accumulation. In the brain, neurons in area LIP display signatures of this evidence accumulation process in trial-averaged data. Evidence accumulation is a stochastic process. Thus, averaging across trials obfuscates the latent dynamics that hypothetically give rise to the decision, though this has been necessary historically because the number of simultaneously recorded neurons was limited by recording technologies. In close collaboration with postdoc Natalie Steinemann, I used newly-developed macaque neuropixels probes to record simultaneously from hundreds of neurons in LIP as animals made perceptual decisions. These large-scale recordings allowed us to observe the dynamics of LIP population activity on single-trials. These dynamics closely matched those predicted by models of noisy evidence accumulation (e.g. drift-diffusion models) and could be used to accurately predict the animals’ choices and reaction times on single trials.
Reviewed preprint: https://elifesciences.org/reviewed-preprints/90859#tab-content
Development of large-scale recording technology for use in non-human primates.
High-density, integrated silicon probes have transformed systems neuroscience in small animal models. These probes enable large-scale neural population recordings with single cell resolution. However, existing technologies have provided limited functionality in nonhuman primate species such as macaques, which offer close models of human cognition and behavior. Through a large collaboration between IMEC and the Howard Hughes Medical Institute, I was involved in designing and testing Neuropixels 1.0-NHP, a new generation of probes for use in non-human primates. The Neuropixels 1.0-NHP is a high channel-count linear electrode array designed to enable large-scale, simultaneous recording in superficial and deep structures within the macaque brain. Specifically, I developed a protocol for reliably and safely inserting these delicate probes into the macaque brain, assessed their performance and longevity, and designed custom hardware that interfaces the probes with commonly used, commercially available microdrives. In our paper, my colleagues and I demonstrated recordings from thousands of neurons within a single session and large-scale recordings from dozens of cortical and subcortical regions. We also demonstrated examples of the new classes of experiments that can be achieved with these probes. All of the protocols and hardware designs have been released open-source to the primate neuroscience community in order to facilitate fast, wide-spread adoption of this transformative recording technology.
Preprint: https://www.biorxiv.org/content/10.1101/2023.02.01.526664v3.abstract
gabestine.wordpress.com 