Abstract— Spasticity is a common impairment following an upper motor neuron lesion in conditions such as stroke and brain injury. A clinical issue is how to best quantify and measure spasticity. Recently, research has been performed to develop new methods of spasticity quantification using various systems. This paper follows up on previous work taking a closer look at the role of transversal forces obtained via rehabilitation robot for motions in the para-sagittal plane. Results from 45 healthy individuals and 40 individuals with acquired brain injury demonstrate that although the passive upper motions are vertical, horizontal forces into and away from the individual’s body demonstrate a relationship with the Modified Ashworth Scale. This finding leads the way to new avenues of spasticity quantification and monitoring. I. INTRODUCTION Spasticity assessment plays an important role in determining the best course of action for regaining function in the acquired brain injury (ABI) population. As a component of the upper motor neural syndrome, it is the most common symptom of ABI, stroke, traumatic spinal cord injury, multiple-sclerosis, and other neuro-muscular disorders [1, 2]. Typically, it is characterized by a velocity-dependent resistance to passive motion [3] and is known be indicative of recovery [4]. Those who have severe ABI’s often have severe spasticity. This impacts their potential for rehabilitation and secondary complications such as contractures and skin breakdown. In addition, spasticity impacts caregivers who provide personal care and assistance mobilizing the person. Currently, the most common method for assessing spasticity is the Modified Ashworth Scale (MAS) [4] a 6 point scale (0, 1, 1+, 2, 3, 4) where 0 and 4 represents little and high levels of spasticity respectively. Although simple and easy to use, the MAS is an ordinal scale and such, has limited sensitivity [5]. New methods of spasticity assessment have been explored, often with new devices and techniques to quantify velocity dependent resistance by EMG readings [5, 6], or utilizing rehabilitation robotics [7, 8]. Several systems, however, focus on the delivery of physical therapy [9, 10, 11] and aspects regarding spasticity assessment as a *Research supported by the Natural Science and Engineering Council of Canada (NSERC) Discovery Grant. N. Seth and H. A. Abdullah are with the University of Guelph, Guelph, ON N1G,2W1 CA (519-824-4120 x53346; e-mail: habdulla@uoguelph.ca). D. Johnson is with Hamilton Health Sciences Regional Rehabilitation Centre, Hamilton, ON, L8L 0A4. secondary objective as indicated in their design. The well- known strength of robotics such as their repeatability, accuracy, and tireless motions, however, make them an ideal tool to be developed for the primary purpose of spasticity assessment. As such, more robotic systems capable of targeting this aspect should be developed given their abilities of collecting temporal force, distance, and velocity readings. Given the issues regarding the MAS, the most commonly used clinical scale [1, 6], providing more detailed representations of spasticity through such systems will provide better answers regarding treatment decisions and ultimately lead to faster recovery. The majority of systems seek to quantifying spasticity through isolating and approximating the velocity component of resistive force readings [12, 13]. Studies involving EMG readings while presenting promising results, pose relevant clinical issues such as the need for skilled personnel, invasive procedures for the case of needlepoint EMG, or issues regarding placement, sensitivity, surface conductivity, and artifacts in the case of surface EMG [14]. Recently, analysis has demonstrated that using healthy control data as a baseline of health, rehabilitation robotics focused on upper limb spasticity assessment can successfully differentiate between healthy controls and slow-to-recover ABI patients [15]. This includes patients with MAS Bicep and Tricep scores of 0, which is normally considered to be healthy with no abnormal tone or spasticity. One aspect not explored in literature is how resistance to functional passive motion changes in the patients’ transverse plane. Abnormal tone often presents findings with resistance into or away from the body, however, it is not intuitive to study this dimension for functional motions with the upper limb that are in the para-sagittal plane. A better of these forces and how they related to the MAS scale may provide greater insight into the condition of individual’s tone and spasticity exhibit small incremental changes that cannot be reflected in current clinical scales. In this paper, we describe further analysis from previous work [15] and present a new focus on the force perpendicular to the motion of the limb, an area not often studied. Findings further demonstrate the case for why rehabilitation robots, specifically utilizing multi-dimensional data collection, should be developed for spasticity assessment protocols. Transverse Forces versus Modified Ashworth Scale for Upper Limb Flexion/Extension in Para-Sagittal Plane Nitin Seth, Denise Johnson, and Hussein A. Abdullah* 2017 International Conference on Rehabilitation Robotics (ICORR) QEII Centre, London, UK, July 17-20, 2017. 978-1-5386-2295-7/17/$31.00 ©2017 IEEE 765