Wearable rehabilitative technology for the movement measurement of patients with arthritis

Abstract

Rheumatoid Arthritis (RA) is an auto-immune disorder that is characterised by pain, stiffness, swelling and deformity of affected joints. Drug treatments and therapies aim to reduce painful RA symptoms, stop disease progression and place the disease into remission. Early identification of RA is important to initiate treatment. RA symptoms can be similar to other related musculo-skeletal and muscular disorders. Identifying it from tell-tale symptoms has been the long-term goal of international teams of clinicians and researchers using varying approaches. Joint stiffness is a common complaint of RA sufferers and is regarded as one of the first symptoms of RA. However its unpredictability between patients and its difficulty of measurement has reduced the importance that joint stiffness plays as an RA identifier. Consequently there is a need for an ambulatory system capable of objectively measuring daily changes in patient movement to identify the symptom of joint stiffness.

This thesis describes the causes of RA and the clinical techniques used to identify and diagnose it. Understanding the condition gives a better understanding of the requirements demanded from a system capable of objective joint stiffness measurement. Other objective measurement devices have been designed by research teams to measure the physical symptoms of joint stiffness through grip strength, hand volume and passive resistance to torque. Unknown physiological effects and limited repeatability of results restricted device adoption into clinical practise. This thesis examines each system.

Joint movement and limitation is assessed by range of motion measurement in the clinical setting. Angular kinematics describes movement in terms of motion or displacement without regard to required force. Kinematics therefore provide alternative views on joint range of motion measurement to allow inter-joint comparisons. Goniometric measurement devices measure joint range of motion. Contact and non-contact systems capable of measuring the full cycle of joint movement are evaluated in this thesis for suitability as a finger joint measurement system. Data gloves are identified as a suitable device to measure finger joint movement and have been used in virtual reality, gesture recognition and rehabilitation.

This thesis focuses on the development of a new objective measurement system capable of determining finger joint movement to evaluate and detect changes in joint range of motion and flexibility and help identify the characteristics of joint stiffness. This new system is comprised of exclusively designed data gloves collaboratively designed by the author that are controlled and analysed by an intelligent suite of software also written by the author to control these bespoke data gloves. The software system is named ‘DigitEase’. DigitEase is capable of measuring finger joint movement in high resolution to evaluate and detect changes in joint range of motion and flexibility and help identify the characteristics of joint stiffness. This thesis details the new data glove hardware and details the algorithms and intelligent techniques employed by DigitEase for detailed objective finger joint measurement. The new data gloves address the common problems associated with data glove wearability, usability and accuracy. Data gloves designed by the author and described in this thesis are self-calibrating, easy to donn and doff, and can wirelessly transmit sensor data to the DigitEase system. Using this data glove system to measure joint mobility removes common problems with intertester reliability associated with clinical assessment techniques such as goniometric measurement and visual examination. The DigitEase software manages all aspects of data glove functionality including user access, exercise management, movement recording and segmentation, and data analysis. Tailored exercise routines are completed by the patient as prescribed by the clinician. Detailed analysis of recorded movement identifies minute changes throughout each recorded session to measure changes in joint range of motion and flexibility.

Information produced by this system must be accurate and reliable. Results produced by DigitEase software algorithms are heavily influenced by the raw data from each data glove sensor. Testing strategies examine each newly designed data glove for accuracy and reliability through specially devised testing strategies. Results are compared to current clinical measurement techniques to assess each data glove as a viable alternative. A neural network is used to create full angular range for each data glove sensor and improve sensor linearity and accuracy. Repeatability and accuracy testing examine the benefits provided by the NN angular method over traditional sensor output.

Date of Award	Aug 2015
Original language	English
Supervisor	Joan Condell (Supervisor) & Kevin Curran (Supervisor)