The University of Auckland

Project #90: Artificial Intelligence algorithms to automatically prescribe Robot-Botox therapies

Back

Description:

Background: Cerebral palsy (CP) is the most common cause of childhood disability and is associated with spasticity (over-tight muscles that can lead to permanent joint contractures), muscle weakness and disordered motor control. Botox has emerged as clinical best practise for treating spasticity, as injections reduce stiffness for a period of time, however then intense physiotherapy is needed to gain more control of the muscles. In our research group we have developed novel wearable robot exoskeletons that can be worn to deliver various types of physiotherapy to patients in their homes with them playing computer games.

 

Aim: The aim of the project is to use artificial intelligence (AI) algorithms to develop an automatic therapy prescription controller that takes into consideration the ability of the patient and other measures to decide what physiotherapy should be delivered to the patient. This algorithm should learn and adapt as the patient progresses.

 

Objectives

1.     Implement various physiotherapy algorithms with adjustable parameters to allow the robot to deliver adaptable exercise programmes

2.     Implement diagnosis algorithms that can automatically detect a patients performance

3.     Develop AI framework that can intelligently adapt the exercise programme based on diagnosis and other subject-specific metrics in order to bring the patient closer to a healthy persons ability

4.     Develop a model of a “virtual patient” to simulate the AI algorithm with a number of different use cases (including CP and BTX injections) to validate that it will work and adapt to bring a patient to normal ability

5.     Pilot testing on a number of healthy individuals to analyse the algorithms ability to adapt and train a person.

Type:

Undergraduate

Outcome:

Prerequisites

None

Specialisations

Categories

Supervisor

Team

Lab

Mechatronics PG + Dynamics and Control Lab (201.562, Lab)