LEFT: Temperature Sensor, RIGHT: Rotary Encoder
Powerpoint presentation for demo
The three main sensors
The workstation prior to moving the circuitry on to the band
The Android app and Android Studio in the background
Firebase real time database
The 3-axis accelerometer
Deep vein thrombosis is a preventable and treatable disease, and yet it affects nearly one million Americans annually, costs the US economy $10 billion per year, and is one of the most common preventable causes of death in hospital, most often due to its risk of developing into a pulmonary embolism. With such drastic and unimproving consequences for a harm that could be mitigated, we felt we needed to do something about this issue and find a way to improve outcomes and treatment for DVT development in hospital.
What it does
Bandwidth is a sensory band designed to be worn by hospitalized patients at risk of developing deep vein thrombosis (DVT). The device detects subtle changes in volume of a limb and localized temperature over time, and relays this information to Bandwidth's app, which will be running on a computer in the hospital unit's nursing station. Upon detecting sensory changes for patient(s) that surpass research-validated thresholds, the Bandwidth app can send a notification to the nurse that the patient may be developing DVT and may require an assessment. This ultimately promises to improve hospitals' ability to catch and address DVT earlier, thereby reducing risks for long-term DVT consequences and pulmonary emboli. Finally, the device also measures movement of the limb throughout the day, and can nofity the nurse of a patient who might be at risk for DVT due to greater inactivity during the day, allowing staff to intervene in the meantime.
How we built it
The hardware end of the sensor band was built using a Raspberry Pi 3, a rotary encoder, a temperature sensor and an accelerometer. These pieces of the puzzle were all wired together in one large circuit. The Raspberry Pi relayed all of the data it received from the sensors, using the code we wrote in python, to a Firebase real time database, where sensor values were stored and updated. We then created an android app, fully created in Android studio in Java. The app retrieved the data from Firebase in real time and displayed for the end user (i.e. a nurse) the data in an organized and simple manner, with all different patients one tap away.
Challenges we ran into
Among a few roadblocks, the largest one was definitely transferring data from the sensors and the Raspberry Pi to the Android app. After many attempts with all sorts of Bluetooth code, hard-wiring the phone to the Raspberry Pi, and a brief stint with Wifi Direct, Google's cloud services came in handy. We utilized Google's Firebase real time database for the storage and transfer of sensor data of the band to then be retrieved by the request of the android app.
Accomplishments that we're proud of
We are proud to have prototyped a technology that incorporates both hardware and software components in such a short amount of time, and to have developed something that has the potential of addressing such a prevalent medical issue. Creating a sensory band with a Raspberry Pi 3, a rotary encoder, a temperature sensor, an accelerometer, a Firebase database, and a fully functioning android app with real-time user updates was quite the feat. Overcoming the technological obstacles for DVT is what we are proud of.
What we learned
We learned through this project of the value and promise that lies in identifying and strategically tackling the gaps in medical care. As well, we learned many new aspects of hardware development, databases and app development.
What's next for Bandwidth
We would like to develop this technology for use in-hospital, to have a real, and positive impact on the quality of life and quality of care that patients will experience. We also see other applications for this technology in research on subtle sensory changes in asymptomatic DVT, as well as in acting as an inexpensive option for monitoring progression of edema or hydrocephalus.