This project necessitates the design of a wearable, wireless ECG unit for real-time performance during dynamic and unpredictable body movement, including contemporary dance. A brief survey of existing technologies for heart rate detection, outlined below, has informed the development of our current prototype.
Requirements and Constraints
- Low profile, sturdy, sweat resistant, amenable to extreme movement
- Low cost (<$60)
- Easy to acquire
- Rechargeable/replaceable battery with long life (3+ hours)
- Wide transmission range (wide enough to cover a large stage)
- Transmission of the raw ECG waveform
Wearable ECG Unit Prototypes
Prototype 2: Developed with Emmanuel Flety at IRCAM, November 2014
This unit consists of:
- AD8232 chip on the Sparkfun Single-lead ECG
- Sensor Cable – Electrode Pads (3 connector)
- ECG Electrodes, adhesive
- X-OSC, wireless i/o board
- Lithium Polymer Battery, 3.7 V, 2000mAH
- 3-D printed box
- Athletic belt, tape, or specialized clothing to hold unit on body
Before settling on the above hardware, we evaluated a number of other potential solutions which were ultimately rejected as they do not fulfill the requirements and constraints of our project as outlined above.
While this unit is reliable, easy to acquire, relatively inexpensive, and transmits its data using a simple protocol (Bluetooth LE), it sends only the heart rate, not the raw ECG signal. Since we are interested in the phase of the signal as well as the variability, access to the raw ECG is important. We also found the short transmission range (10m) of Bluetooth LE and this unit’s susceptibility to occlusion problematic as they would require each dancer to carry a cell phone.
While this board is intriguing, we decided not to evaluate it as the RMCM01 is no longer manufactured and seems unlikely to be in the future.
This inexpensive and easy to use device worked relatively well and produced a clear signal which, unlike an ECG, was not susceptible to muscle noise. However, we found that the device had to be positioned carefully with just the right amount of pressure and that it was easily displaced during the execution of fairly modest choreography.
While our initial desire was to use medical ECGs, their cost was prohibitive and their reliance on proprietary (and expensive) software made them impractical.
OpenBCI is a compelling platform and we had a chance to try out an evaluation board, but as it is designed for EEG, it proved to be overkill and cost prohibitive for our purposes.
A quick search for “Wearable ECG” shows that there is no shortage of interest and many interesting projects under development. While our current prototype meets our needs, we will keep a close watch on the new devices that emerge over the next few months/years.
As we are interested in the relationship between breath and heart activity, we have experimented with two different respiration sensors to date. Both measure the displacement of the chest/ribs/belly during breathing.
BioPak Respiration Belt Transducer
Textile Respiration Belt on Tank-top
Developed by Adrian Freed at CNMAT, August 2014