Application Dependencies and Incident Management
Real-Time Contactless Medical Sensing
The goal of this project was to create a prototype mobile application supporting the following functions:
- Use of a video camera on a mobile device to do static face recognition on the strongest capture point for micro-blushes of a patient, the forehead. Eliminate the need for physical sensors, cables, and proprietary machines for these readings.
- Capture separated video frames into color channels, analyze the strongest color channel, and utilize this for waveform data collection.
- Create a “blood flow algorithm” to analyze the waveform and respond with an accurate “ground truth” reading of pulse rate and SPo2 calculations.
- Xiaoyue Hong
- Nanxi Li
- Longfei Wang
- Zhehao Wu
- Yue Xing
- Jun Zhao
InteloMed is a Pittsburgh-area health care company that offers CVInsight® Patient Monitoring & Informatics System, a non-invasive monitoring system that provides real-time information about cardiovascular conditions. Using IoT (Internet of Things), InteloMed can capture and analyze data retrieved from medical patients through sensors. This data consists of a pulse rate and SPo2 (Blood Oxygen Level) reading.
The CMU MSIT eBusiness Technology team was challenged to remove the physical sensor worn by patients from this type of data capture. They accomplished this by utilizing a simple video camera attached to commonly used devices (mobile devices, cell phones, tablets). These devices needed to perform facial recognition, accurately capture and read the physical nature of the human body’s micro-blushes, and create an algorithm to determine the pulse rate and SPo2 readings from a waveform created from these micro-blushes. In addition, the devices needed to capture and store the acquired data locally and to a cloud service which allows for real-time data access.
On August 25, the six-member student team demonstrated their prototype at the MSIT eBusiness Technology program’s Practicum Competition. They did this in real time by measuring a team member’s heart rate using only a camera pointed at their head with zero human contact. After this same individual elevated their heart rate through exercise, the screen then correctly showed the increased heart rate through their device. Additionally, the team proved the ability to measure SPo2 levels. To demonstrate this, an individual from the team held their breath, and the device measured the gradually decreasing SPo2 numbers. While the visual demonstration was impressive, the potential benefits that this can provide the health care industry is extraordinary.
Congratulations to Team InteloMed for winning the 2017 Practicum Competition along with $14,000 for their project “Real-Time Contactless Medical Sensing”.