Neural Repair and Robotics
: The In-cage automation of a test for dexterity after stroke.

Student thesis: Doctoral ThesisDoctor of Philosophy


Stroke is a chronic and debilitating disease that affects millions of individuals worldwide. The collective medical and socioeconomic costs reach into billions of pounds annually. Pre-clinical research into neurorestorative treatments for this disease often involves the behavioural assessment of rats using the Single Pellet Reaching Task (SPRT). This task is executed by placing a pellet in front of a rat beyond a slit, offsetting the pellet to the left or right so that it can only be reached with the right or left paw respectively. This allows for independent evaluation of the more or less-affected forelimb after unilateral stroke. The animal is scored on the number of pellets it consumes by reaching through the slit. After a cortical stroke, rats will drop more pellets than naïve rats, and despite some spontaneous recovery, without treatment, recovery is incomplete. It is an effective assessment method which yields information about motor skills, coordination, dexterity, reach, and grasping ability. However, it is labour intensive and time-consuming, resulting in lower powered experiments than that desired by researchers.

This thesis describes the development of robotic equipment to provide the in-cage automated training, rehabilitation and assessment of the grasping of sucrose pellets by large cohorts of rats. The goal of this new assessment paradigm is to help accelerate the identification of therapies for patients with central nervous system (CNS) injuries that impact dexterity.

With the use of rapid prototyping methods and technologies, an in-cage robotic apparatus (RatBot™) has been developed and validated to automatically train rats in the SPRT and assess their grasping ability before and after a CNS injury. The apparatus comprises of a programmable 45mg sucrose pellet dispenser with features to prevent pellet jamming and reduce grinding, and a programmable pellet positioning system which can present the pellet to a variety of locations within the reach space of a rat. Pellet consumption statistics, rat identification and trial participation information are automatically gathered and persisted to a database for remote access and offline analysis.

With minimal infrastructure and maintenance requirements, an array of RatBots has been shown to be sensitive to the individual reaching capabilities of the rats tested, successfully identifying grasping deficits in two independent cohorts of rats with unilateral cortical stroke. The device can operate in parallel in both bench top as well as overnight unattended in-cage operation.
Date of Award2018
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorLawrence Moon (Supervisor) & Thrishantha Nanayakkara (Supervisor)

Cite this