Year: 2026
Type: dissertation
Abstract: A complete map of human synapses is impossible to construct with existing technology due to limitations on accuracy, speed, and scale. This research proposes HEADTURNER (HElical Arrangement Device for Tissue Under Rotation for Neuroscientists, Engineers, and Researchers) as a new method of handling tissue that addresses these limitations and could reduce the time to collect a given brain volume by several hundred times. Prior methods discretely process individual slices of tissue by manual manipulation or the use of specialized conveyor belts, and they are prone to human errors and section loss or damage. The proposed method intakes a continuous ribbon of tissue and autonomously wraps it around a drum in a helical pattern, allowing all tissue to be visualized during downstream imaging with minimal human intervention. This work develops a first-principles framework to assess the feasibility of helical tissue collection for connectomics. Models from web handling, thin plate mechanics, and precision machine design are unified to characterize the material behavior, geometric constraints, and error sources governing the winding of compliant biological tissues at nanometer-scale thicknesses. Necessary bounds on tension, mandrel geometry, winding speed, and positional error are derived. The analysis defines the realm of acceptable parameter combinations for stable, repeatable winding without wrinkling or tracking errors, which are implemented in the design, fabrication, and characterization of a precision helical winding machine. Maintaining lateral errors within 500 μm and angular errors within 0.001 rad are suggested to prevent wrinkling errors in 1 mm wide tissue samples. Experimental measurements support the proposed models and demonstrate pitch variation within 250 μm and minimal defect formation. By establishing both the physical limits and practical implementation of continuous helical tissue collection, this thesis introduces a scalable pathway toward high throughput connectomics. Beyond neuroscience, the findings provide generalizable insights into precision winding of ultra-compliant materials, with implications for manufacturing processes involving thin films, biomedical devices, and nanostructured materials.
Author: Roeder, Gillian J.
Institution: Massachusetts Institute of Technology
Language: English
Subfield: Mechanical Engineering
Field: Engineering
Domain: Physical Sciences
Keywords: connectomics, web handling, compliant materials
Open Access status: green
OpenAlex ID: W7202009105
