Boston University
Cellular identity and cortical computation: from cell types to neural representations during abstract perceptual decision-making
Abstract
dc:description.abstractNeural circuits are composed of distinct cell types. However, how these cell types facilitate circuit computation is poorly understood. This dissertation addresses two problems: 1) how cell types can be identified with electrophysiology and 2) how neuronal populations in prefrontal cortex compute abstract perceptual decisions. In in vivo extracellular electrophysiology, cell type information is almost always inaccessible except at the coarsest level. In the first study (Chapter 2), I developed WaveMAP, a tool to identify putative cell types from action potential shapes. With this tool, I uncovered putative cell types in primate premotor cortex. Here I describe the functional properties of these putative types and conclude with predictions as to what molecularly defined, ground-truth cell types each may correspond to. In this study, only action potential shape was considered. However, there are other properties of cell types that are informative for their identification. In the second study (Chapter 3), I developed another tool—called PhysMAP—to perform multimodal identification of cell types from electrophysiology. I show that this method identifies batch effects and can be used to identify cell types in experiments without ground truth. With this, I uncovered the activity and real-time interactions of multiple cell types in a behavioral experiment without optotagging. Finally, most perceptual decision-making tasks conflate abstract decision formation with motor preparation because subjects know the required actions during stimulus viewing. In a third study (Chapter 4), I used Neuropixels and laminar probes to record from dorsolateral prefrontal cortex (DLPFC) neurons of rhesus macaques performing a discrimination task in which random target locations were revealed after a delay. This temporally dissociated stimulus evidence from action choice. Dimensionality reduction and cross-temporal decoding analyses revealed evidence-selective activity during stimulus viewing that persisted through the delay period. Applying WaveMAP, I found putative parvalbumin-positive interneurons encoded stimulus evidence through the delay, whereas excitatory units encoded evidence only during stimulus viewing (Chapter 5). Taken together, this body of work advances understanding of how prefrontal cortex and its cell types compute decisions abstracted from upcoming motor actions.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy
- Level thesis:degree_level
- doctoral
- Discipline thesis:degree_discipline
- Psychological & Brain Sciences
- Grantor
- Boston University
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Lee, Eric Kenji
- Advisor dc:contributor.advisor
-
- Chandrasekaran, Chandramouli
- ORCID dc:identifier.orcid
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- Attribution 4.0 International
- Licence dc:rights.uri
- Language dc:language.iso
- en_US
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/2144/53588