In-Home Biomechanics Using Kinect Sensors
Often caregivers (including nurses, nursing aids, healthcare workers, etc.) suffer lower-back injuries that can be disabling and cause them to take time off work. Most of these injuries happen because caregivers often have to move or lift patients or clients. These injuries are expected to happen more often as our population gets older and more people require care at home. In patients’ homes, caregivers have to work in small, confined spaces. Most caregivers are without help to do very strenuous tasks, since much of the equipment designed to help caregivers is too large for small in-home spaces, and since they are usually unable to get help from another caregiver. This situation puts caregivers at a much higher risk of injury, and increases the need to develop effective assistive lifting equipment that can fit in people’s homes. Understanding the techniques caregivers use to provide care in such challenging spaces is key to developing a solution. However, most existing biomechanics evaluation tools use large, expensive equipment that often involves attaching markers to the caregiver, which can get in the way as the caregiver assists the client or patient.
Our first goal is to develop a portable, low-cost motion capture system for evaluating posture, which can fit in most confined spaces such as the bathroom. Our system uses multiple Kinect sensors (sensors that give information about color, depth and joint location) and a computer to collect and analyze data about the caregiver’s posture and joint movement. It will estimate the force put on the caregiver’s spine as they assist the client. Our second goal is to study and understand the techniques caregivers use to assist clients in the bathroom setting where most injuries occur. Our system will help us to identify postures, tasks and times that cause the greatest risk of injury, and estimate various factors such as how fast the caregiver lifts or lowers a patient, how fast the caregiver moves when not lifting or moving heavy weight, and the area where most of the heavy lifting and maneuvering is done. With the information we gather, we plan to develop a smart lift system that will fit in the home environment and reduce injuries.
The first step was to do a quick test to see if the Kinect was capable of giving us reasonable motion capture data. Here’s the resulting paper and AudioSlide presentation:
Next we developed a multi-sensor motion capture system and are now conducting a validation study to compare the accuracy of our system with a gold standard system (Vicon). We are also testing which camera heights and angles are the best to use for collecting our data. We then plan to set up our system in the Home Lab with 5 PSWs assisting an actor client to gather information about the caregivers’ postures and techniques as they assist the client, and the stresses these techniques put on each caregiver’s spine. We will also look at the various parameters needed to develop design specifications for a smart lift system.