Mindfulness Monitor pt 5

This series of posts will be documenting the development of Group 17’s (Evan Oskierko-Jeznacki, Christina Kim, Jiaang Hu) final project for ESE 519 Real-time Embedded Systems. Advisor: Dr. Nalaka S. Gooneratne, M.D., M.Sc.

Our proposal for the mindfulness monitor is built around three primary input vectors. The first, and primary, is the ECG signal. The second most significant is the respiratory rate signal. The third, and least significant is hand movement. During consultations with Dr. Gooneratne, we decided to separate the ECG and respiratory rates as much as possible, in terms of how these signals are input into the mindfulness monitor. Initially, we had considered using pulse oximetry instead of the detecting electrical pulses. On the hardware side, this would have been much simpler, however we decided to avoid this approach because of cardiopulminary coupling.

Initially, we had proposed to use and/or reverse engineer an existing respiratory belt, provided by Dr. Gooneratne. These belts work by detecting the change in resistance caused by the lengthening or shortening of an embedded conductive element. We had spent some time trying to use the provided Embla XActrace belts (below), but without the necessary belt amplification pack could not make any headway. We also did not want to damage the belt we were provided so we began looking for alternatives.

Ultimately, we settled on building our own respiratory belts. To do this, we purchased conductive rubber bands and embedded them into elastic medical bands (e.g. ace bandages). The bands have a set resistance, when tension is applied to the band and they lengthen. As they lengthen, the conductive particles embedded within the rubber spread apart and the resistance changes. We measure this change across a voltage divider into the ADC circuit. Given that the band is applied tightly across the chest, the device can detect both deep breaths as well normal, resting breathing that cause relatively small chest expansions. Furthermore, to increase the accuracy of this portion of the mindfulness monitor input, we chose to implement two respiratory belts (as is typical in most clinical grade respiratory belt systems). The resultant value is the average of the readings taken from both of these belts.

Preliminary results of the respiratory portion of the circuit are below. The noticeable spike occurs when I sat down. At this point, two deep breaths are noticeable followed by normal breathing.


Leave a Comment