A novel device to detect the infection of wound

The commonly found biomarkers of wound infection that can be tracked non intrusively are

1. Odors: Bacteria and microbes in wounds generate volatile compounds and are often the cause of bad smell emitted by dead organic matter or decomposed tissue. The main volatile gases emitted are ammonia and amines; acetone and its variants; alcohols such as ethanol.

2. Temperature: A clinically diagnosed wound infection shows an elevation of temperature difference between 4 to 5 degrees between the wound and healthy skin. Wounds with just signs of infection show a temperature difference of 1.5 to 2.2 degrees.

3. Moisture: Most research suggests that dressings that create and maintain a moist environment, are now considered to provide the optimal conditions for wound healing. However an infected wound that is often accompanied by increased discharge which would increase the moisture of the dressing and surrounding skin area.

Based on the above research, I decided to build a hardware device was built using the Arduino Mega2560, connected to MQ gas sensors which detect Ammonia, CO₂ and Acetone, as well as a sensor to measure temperature and humidity (DHT-11). The circuit was then connected wirelessly to an Android phone via a Bluetooth chip, to communicate with the App. Standard programs and libraries were modified with the input values gathered from sensor specification manual for generating gas, temperature and humidity values in their respective units (ppm, centigrade, % water vapor).

The Android app developed included different features to display and analyze data. The app captured the concentration of the biomarkers in fresh air which was then used to normalize the same biomarkers captured from the wound. Data is stored onto the device, then saved to the cloud (Firebase) when network service is available.

The circuit box and the sensor pads were designed using Autodesk Fusion 360

Conclusion

In conclusion, I considered my device a success since it met all the design criteria - it was non-intrusive, small and portable, was capable of taking accurate readings of all biometrics, and could store data on the device and cloud for analysis. Moreover, it met the budget of $200, only costing $176.43.

For the testing of a simulated wound using rotting fish, the readings were stable and showed a rise only when the fish began to decompose after 12-24 hours. In addition, the higher values of ammonia recorded were consistent with the slightly putrid smell produced by the fish. This was also accompanied by much higher levels of carbon dioxide and acetone, which are both odorless gases that are difficult to detect. Thus, this device would work on wounds if all or any of the gases are being released as the result of infection.

Moisture detection : The device was able to detect varying levels of humidity as shown by the data during independent testing with moist paper towels as well as testing with decomposed fish. Although it is inconclusive whether or not the bacterial decomposition was increasing the level of moisture, an increased wetness level was detected by the DHT11 moisture sensor.

Temperature Detection: The temperature readings for wound infection could not be simulated and any temperature changes due to the decomposition of fish is inconclusive. However the sensitivity of the DHT11 temperature sensor with independent testing on skin showed that the device could be relied for measuring different levels of skin temperature.

Mentors and Advisors:

Mr. Taylor, Computer Science, Homestead High School

Mr. Sam Fung, Chemistry, Homestead High School