Mamontovs M., Sisojevs A. and Tatarinov A.. Bone porosity and thickness evaluation by ultrasonic signals using pattern recognition.. n Proceedings of the International Conferences on Interfaces and Human Computer Interaction 2024, Game and Entertainment Technologies 2024 and Computer Graphics, Visualization, Computer Vision and Image Processing 2024 – Part of the 18th Multi Conference on Computer Science and Information Systems 2024, 200-204 pp. IADIS Press, 2024.
Bibtex citation:
Bibtex citation:
@inproceedings{16914_2024,
author = {Mamontovs M. and Sisojevs A. and Tatarinov A.},
title = {Bone porosity and thickness evaluation by ultrasonic signals using pattern recognition.},
journal = {n Proceedings of the International Conferences on Interfaces and Human Computer Interaction 2024, Game and Entertainment Technologies 2024 and Computer Graphics, Visualization, Computer Vision and Image Processing 2024 – Part of the 18th Multi Conference on Computer Science and Information Systems 2024},
pages = {200-204},
publisher = {IADIS Press},
year = {2024}
}
author = {Mamontovs M. and Sisojevs A. and Tatarinov A.},
title = {Bone porosity and thickness evaluation by ultrasonic signals using pattern recognition.},
journal = {n Proceedings of the International Conferences on Interfaces and Human Computer Interaction 2024, Game and Entertainment Technologies 2024 and Computer Graphics, Visualization, Computer Vision and Image Processing 2024 – Part of the 18th Multi Conference on Computer Science and Information Systems 2024},
pages = {200-204},
publisher = {IADIS Press},
year = {2024}
}
Abstract: The fragility of long bones in osteoporosis is linked to a decrease in cortical layer thickness (CTh) and the development of internal porosity (P) within the cortex. This study is dedicated to investigate whether it is possible and how accurately P and CTh can be determined using a pattern recognition methods, utilizing discrete Fourier transform (DFT) analysis of ultrasonic signals transmitted through a soft tissue layer. As input data, ultrasound scans of phantom of cortical bone with different grades of CTh and P were used. Each of the phantom scans was chosen in turn as a test object, while other specimens composed the training set. For analysis, three different existing approaches were applied firstly and the later one was developed by improving the existing ones. The analysis was carried out in three categories: P and CTh ratio, P only and CTh only. Based on the results of each analysis, a method was obtained that predicts P and CTh better than other applied ones. A comparison of the methods for P and CTh determining and choosing the most effective ones are shown.