BACKGROUND

About 2.6 million individuals in the United States use manual wheelchairs as their primary form of mobility in order to increase their independence and perform activities of daily living (1). However, due to the repetitive nature of the wheelchair stroke-cycle, injuries to the shoulder, elbow, wrist, and hand are extremely common. Over 73% of manual wheelchair users experience some type of shoulder pain (2). While age and activity level do correlate to injury rate, it is the repetitive trauma that occurs to the joint during wheelchair propulsion that is the main cause of these injuries (3). To date, the primary way that healthcare professionals are able to prevent shoulder injuries in wheelchair users is to ensure proper wheelchair fitting and to help minimize the force and frequency of the patient’s stroke cycle (4).

Unfortunately, there are no personal fitness tracking devices (e.g. Fitbit) designed for persons who use manual wheelchairs. In addition, there is only one clinical tool that is able to quantify these metrics, the SmartWheel; but, with a price-tag of $20,000 it is only able to be used in a clinical setting (5). This makes it impossible for healthcare professionals and wheelchair users to understand real world habits of wheelchair users. In addition, the SmartWheel weighs 9lbs, almost 25% of the total weight of a standard wheelchair, and it replaces one of the wheelchairs wheels while data is being collected. Therefore, it makes the individual’s wheelchair significantly more difficult to propel and can make it unbalanced. The SmartWheel is shown in Figure 1.

While there are various devices that can track similar metrics for bicycles or for lower extremity ambulators, these are exceedingly difficult to adapt for use on a wheelchair (6-7). The inspiration of the SmartHub project was the lack of devices currently available for tracking daily personal fitness and related metrics for manual wheelchair users.

PROBLEM STATEMENT

There are currently no personal fitness tracking devices available for manual wheelchair users; however, information on personal fitness, such as stroke force and frequency, can be used by healthcare professionals and manual wheelchair users to prevent upper-extremity injuries. There are also no devices on the market that can properly collect fitness data for research use within this population. Therefore, the goal of this project was to create a personal fitness tracking device designed for manual wheelchairs that is able to measure average velocity, distance traveled per day, periods of daily activity, average stroke frequency, and average pushing force.

DEVICE DESIGN

 

The main design objective of this project was to create a personal fitness tracker for use on manual wheelchairs. However, the device also needed to be safe to use, easy to use, lightweight, inexpensive, and durable.

Safety is an obvious design constraint; however, it came up many times during the design process. For example, whether or not this device should have a real-time display. We decided that a real-time display could potentially cause accidents, if individuals were focused on the device and not where they were going. Ease-of-use is also an important design constraint given that the population the device is intended for often has reduced manual dexterity and muscle weakness. Therefore, device installation and data retrieval need to involve minimal user effort and be intuitive. In addition, the user-interface for viewing personal fitness metrics needs to be easy to navigate and understand.

The device weight, cost, and durability were also important design considerations. In order to ensure that the device did not interfere with an individual’s normal propulsion habits, it needs to be lightweight. Our goal was to make a device that weighs less than 10% of standard wheelchair weight, or 3.5 lbs. In addition, individuals would not be able to get support from Medicare or Medicaid when purchasing the device. We wanted it to be priced competitively with similar devices sold for lower extremity ambulators such as the Fitbit or Nike+ FuelBand. This means it needs to be built for less than $150 (8-9). Finally, we wanted to make a device that could be taken anywhere someone may need their wheelchair. So, the device needs to be water-resistant and firmly attachable to the wheelchair frame.

DEVICE DEVELOPMENT

Device Overview SmartHub on Wheelchair (Low)

The SmartHub consists of two parts that must be attached to the wheelchair: the main control board and a magnet. The board attaches between the frame of the wheelchair and the wheel, using either Velcro or double sided tape. The magnet is attached to one of the spokes on the wheel. The only restriction on device placement is that the magnet must be able to pass over the control board when the wheel is spun. An image of this placement can be seen in Figure 2 and a video of the device being used on a wheelchair can be seen below.

The device currently has a micro-SD card that can be removed and inserted into a card reader to download data onto a computer. However, the group plans to incorporate a card reader into the device so that a cable can be plugged into it directly to download data and charge its battery. This will make it much easier for individuals who have limited manual dexterity to use. A video of the micro-SD card being removed from the device can be seen in the video below.

The battery life is around 20 hours, so it can be used for an entire day before it needs to be recharged. Currently, the SmartHub weighs less than 1lb and can be produced for $140. The group is also planning on 3D-printing a case for the SmartHub so that it can be protected from the elements when used outside.

Electronics and Programming SD Card being put into SmartHub (Low)

The control board contains three main electronic components: an Arduino Pro-Mini, a reed switch, and an accelerometer. The Arduino is the main controller for the device. The reed switch is used to calculate distance traveled, average velocity, and time active and the accelerometer is used to calculate stroke frequency and average pushing force. The control board also contains a port to hold a micro-SD card for data storage and a USB connection to charge the battery. An image of the main control board can be seen in Figure 3.

The calculations for distance traveled, time active, average velocity, stroke frequency, and pushing force are performed in a Matlab executable file when the user downloads their data from the device. Distance traveled is calculated using the known circumference of the wheel and counting the number of times the magnet passes over the reed switch. Time active is determined by looking for periods where the magnet passes the reed switch more than once in a five second interval. If the magnet has not passed within five seconds, the wheel has not made a full revolution in that time, and the user has entered into a period of inactivity. The average velocity is determined by knowing the time between periods when the magnet passes the reed switch. A video of the magnet passing the reed switch is shown below, the flashing green light indicates when the magnet passes.

Stroke frequency and pushing force are calculated using accelerometer data. When the user pushes on the handrim, there is a peak in the linear acceleration recorded by the wheelchair. These peaks can be used to determine push frequency. The magnitudes of these peaks can be used, along with the known weight of the user, to determine an approximate pushing force. Figure 4 shows a sample of the filtered accelerometer data that is used to calculate stroke frequency.

User-Interface & Data Downloading Reed Switch (Low)

As previously mentioned, the user interacts with the device via a MATLAB executable file. The initial set-up screen can be seen in the Figure 5. The user only needs to enter their personal data the first time they use the device. This set-up screen is used to gather information needed to make the distance and force calculations for the wheelchair.

Once the user has downloaded the data onto their computer, they can run the executable file. This file will give a visual printout of their daily activity periods, as well as, list the values for their personal health metrics. The group is working on being able to automatically save this information to a log, so that the user can have a file showing their daily usage habits.

VALIDATION

The SmartHub’s accuracy is currently being tested by comparing its reported metrics to those reported by a SmartWheel, which is gold standard for device able to measure them. Preliminary results from controlled trials show that the SmartHub’s metrics are within 10% of those metrics reported by the SmartWheel. Plans have been made to send the device home with several different individuals for a day to see how well it is able to report metrics on a full day’s worth of data.

In addition, the SmartHub was presented at the Hite Symposium, an occupational therapy conference that features selected student projects and brings in occupational therapists from around central Ohio. The SmartHub was well received at the Hite Symposium and many individuals wanted us to build more for use in local clinics. The group intends to continue working on this device in the future and will hopefully be able to provide them with several devices.

DISCUSSION AND DEVICE IMPLICATIONS

Because there are currently no devices on the market that are able to be taken out of the clinic to track a manual wheelchair user’s fitness, the SmartHub fills a void in the market. The ability to track an individual’s wheelchair habits in the community opens up a new opportunity to collect data on an individual’s day-to-day wheelchair use. For researchers, this type of data could be used to better understand how individuals develop upper-extremity injuries and how to prevent them. The number of research studies that can be performed with data collected can be endless. They can track effects on quality of life, effects on the cardiovascular system, recommend proper number of strokes per day, along with endless other possibilities. For clinicians, it can be used to help rehabilitate patients who have sustained upper extremity injuries. For wheelchair manufacturers, it would provide data that can be used to design more efficient wheelchairs. For individuals who use manual wheelchairs for mobility, this information will allow them to easily track personal health goals.