PROJECT 40

City: Kyiv

Project manager: Bohdan Salenko

Technology readiness level: 7

An innovative tool for checking wound channels has been developed. The functional dependence of controlled signals on the shape and type of a foreign object in the wound channel has been revealed.

The feasibility of using the frequency-amplitude characteristics of a controlled signal as a parameter has been proven. The study proved the functional dependence of the noise emission signal spectrum at the moment of mechanical contact with a foreign object in the wound on its type; it was determined that the use of an acoustic analyzer by decomposing the signal based on fast Fourier transform allows to accurately identify a foreign object in the wound and determine its location and type. 

In order to improve the accuracy of identification, machine learning algorithms have been proposed, which are implemented on an original converter that allows the type of fragment larger than 3-5 mm to be determined with a probability of 75-80%.

Project presentation in PDF format  >>>

---

FOR THE JURY >>>                                                                      FOR INVESTORS >>>

---

You can ask your questions about the project through the comments form  _⇩