Goal

The project’s initial goal was to collect data from Chronic Obstructive Pulmonary Disease (COPD) patients, analyze it and find a way to predict the exacerbation periods typical to the disease. At first the data was collected using Emfit under-mattress bed sensor which provided information about the patients’ heart rates and breathing frequencies during their sleep. However midway through the project’s timespan the goal was changed due to the lack of data from exacerbation periods. The new goal was to create different ways of gathering data from the patients in addition to the existing bed sensor. Basically the project changed from data analysis project to a research and embedded systems project.

Result

The end result of the project is a bit lacking due to obvious time constraints but still sufficient. The team managed to figure out two ways of gathering additional data from the COPD patients: collecting daily blood oxygen level (SpO2) values and counting coughing sounds. Both of these techniques use Raspberry Pi with specific additions as their core. SpO2 values are collected using pulse oximeter and coughing noises are captured by recording patient’s sleep with 4-microphone array ReSpeaker and then analyzing if the noise is coughing or not. Although neither of these mediums are ready for hospital implementation, they can still be used as a basis for upcoming research projects on this subject.

Additionally the team delivered a research study on the subject and some mockup data analysis on what could be done after enough viable data has been collected.