During the last two decades, a large variety of upper limb exoskeletons have been developed. Out of these, majority are platform based systems which might be the reason for not being widely adopted for post-stroke rehabilitation. Despite the potential benefits of platform-based exoskeletons as being rugged and reliable, stroke patients prefer to have a portable and user-friendly device that they can take home. However, the types of actuator as well as the actuation mechanism used in the exoskeleton are the inhibiting factors why portable exoskeletons are mostly non-existent for patient use. This paper presents a quantitative analysis of the actuation systems available for developing portable upper arm exoskeletons with their specifications. Finally, it has been concluded from this research that there are not many stand-alone arm exoskeletons which can provide all forms of rehabilitation, therefore, a generic solution has been proposed as the rehabilitation strategy to get best out of the portable arm exoskeletons.

A risk assessment tool has been developed for automated estimation of level of neuropathy based on the clinical characteristics of patients. The smart tool is based on risk factors for diabetic neuropathy, which utilizes vibration perception threshold (VPT) and a set of clinical variables as potential predictors. Significant risk factors included age, height, weight, urine albumin-to-creatinine ratio, glycated hemoglobin, total cholesterol, and duration of diabetes. The continuous-scale VPT was recorded using a neurothesiometer and classified into three categories based on the clinical thresholds in volts (V): low risk (0-20.99 V), medium risk (21-30.99 V), and high risk (≥31 V). The initial study had shown that by just using patient data (n = 5088) an accuracy of 54% was achievable. Having established the effectiveness of the “classical” method, a special Neural Network based on a Proportional Odds Model was developed, which provided the highest level of prediction accuracy (>70%) using the simulated patient data (n = 4158). In the absence of any assessment devices or trained personnel, it is possible to establish with reasonable accuracy a diagnosis of diabetic neuropathy by means of the clinical parameters of the patient alone.

Patients suffering from stroke need to undergo a standard and intensive rehabilitation therapy. The rehabilitation training consists of three sequential stages: the first stage is controlled joint movement under external actuator, the second stage deals with supporting the movements by providing assistive force, and the last stage provides variety and difficulty to exercises. Most of the exoskeletons developed so far for rehabilitation are restricted to a particular type of activity. Although a few exoskeletons incorporate different modes of rehabilitation, those are software controlled requiring sensory data acquisition and complex control architecture. To bridge this gap, a portable elbow exoskeleton has been developed for delivering three stages of rehabilitation in a single structure without affecting the range of motion and safety features. Use of electric motor and springs have been arranged in the actuation mechanism to minimize the energy consumption. The developed exoskeleton enhances torque to weight ratio compared to existing models, and all the three modes of rehabilitation have been controlled using a single motor.

This work is to build upon the concept of matching a person's weight, height and age to their overall body shape to create an adjustable three-dimensional model. A versatile and accurate predictor of body size and shape and ligament thickness is required to improve simulation for medical procedures. A model which is adjustable for any size, shape, body mass, age or height would provide ability to simulate procedures on patients of various body compositions. Three methods are provided for estimating body circumferences and ligament thicknesses for each patient. The first method is using empirical relations from body shape and size. The second method is to load a dataset from a magnetic resonance imaging (MRI) scan or ultrasound scan containing accurate ligament measurements. The third method is a developed artificial neural network (ANN) which uses MRI dataset as a training set and improves accuracy using error back-propagation, which learns to increase accuracy as more patient data is added. The ANN is trained and tested with clinical data from 23,088 patients. The ANN can predict subscapular skinfold thickness within 3.54 mm, waist circumference 3.92 cm, thigh circumference 2.00 cm, arm circumference 1.21 cm, calf circumference 1.40 cm, triceps skinfold thickness 3.43 mm. Alternative regression analysis method gave overall slightly less accurate predictions for subscapular skinfold thickness within 3.75 mm, waist circumference 3.84 cm, thigh circumference 2.16 cm, arm circumference 1.34 cm, calf circumference 1.46 cm, triceps skinfold thickness 3.89 mm. These calculations are used to display a 3D graphics model of the patient's body shape using OpenGL and adjusted by 3D mesh deformations. A patient-specific epidural simulator is presented using the developed body shape model, able to simulate needle insertion procedures on a 3D model of any patient size and shape. The developed ANN gave the most accurate results for body shape, size and ligament thickness. The resulting simulator offers the experience of simulating needle insertions accurately whilst allowing for variation in patient body mass, height or age.