Abstract
Introduction
There is a recognized need for sensitive, quantitative and cost-effective clinical exams that assess neuromuscular health. Correctly identifying movement impairments, and the extent and source of the impairment, is critical for diagnosing movement disorders and prescribing an appropriate course of action. Movement deficits caused by traumatic brain injury (TBI), for example, are difficult to diagnose because they are both common and complex and involve diffuse neural networks. Movement deficits that involve diffuse neural networks are not yet understood as well as those caused by focal lesions that are seen in stroke victims. Consequently, the clinician generally recommends a standard rehabilitation protocol that is not tailored to the problem or the individual.
A variety of batteries to evaluate motor deficits already exist but are lacking in one or more of these criteria. For example the extent of tremor is most often determined using a rating scale (Goetz, 2008) (Trouillas, 1997). Though rating scales have been validated and are performed by highly experienced and trained clinicians, they are nonetheless subjective. The scales are also course, from 0-4, with each rating defined differently depending on the test. Robotic systems (Debert, 2012) are quite sensitive and provide quantitative information, however, they are cost prohibitive for most clinics and rehabilitation centers. Optoelectronic systems, electromagnetic systems, inertial measurement units, and electrogoniometers identity and quantify movements well, but require painstaking placement of markers, electrodes, or sensors in precise locations on the patient’s body by trained personnel. The setup time for such systems are impractical in the clinical setting, especially when it compounds over repeated assessments.
Recent advances in gaming and animation provide an opportunity for sensitive, quantitative, and low-cost movement assessments. Systems such as the X-box Kinect, Organic Motion, and Leap Motion employ markerless motion capture that capture body movements, thus providing a practical solution for recording movement during a motor assessment. Such recordings result in a large quantity of data for in-depth movement analyses. Also impressive is that this technology is inexpensive and commercially available.
We leveraged markerless motion capture technology to fulfill two research aims: 1) Develop a quantitative motor assessment (QMA) that is clinically relevant, user-friendly, low-cost, and highly sensitive, and 2) establish a normative database for each of the test measures to allow comparisons relative to a healthy norm. The QMA involves an integrated system, including a motion capture device and custom software. The particular motion capture device we used (Leap Motion, San Francisco, CA) is capable of tracking the position of a fingertip, palm of the hand and tools such as pens or dowels. As the person being tested performs each movement in the assessment, the system records their hand and finger position with a resolution of 0.01mm and a sampling frequency of 100Hz. Other motion capture devices with such capabilities can equally serve as a solution in our system.
The QMA automates conventional tests, which have served the medical community relatively well in the absence of current technology. Each test is described in detail in Kincaid et al. We seeded a normative database for the QMA by administering these tests to 101 healthy control subjects, and compared QMA tests results to those of conventional tests.