Abstract
Electrical impedance tomography (EIT) is an imaging method for characterizing the inner conductivity distribution of an object based on the measured boundary voltages resulting from the injection of an AC signal, followed by an image reconstruction procedure. An algorithm tries to solve an ill-posed inverse problem making it challenging to reconstruct an accurate image. To overcome this, we propose a genetic algorithm (GA) for the image reconstruction with a non-blind search method considering prior knowledge about the possible conductivity distribution in the initial search space. To validate the algorithm, experiments have been conducted in a water tank. The algorithm’s performance was evaluated regarding image quality and processing time, being able to minimize the corresponding quality function to 0.0505 with 100 generations using the non-blind search and the uniform crossover/random mutation. Compared to traditional methods, the GA achieves significantly better image quality. It has been implemented as an image reconstruction algorithm for gesture recognition. EIT measurements have been conducted with six persons performing American sign numbers (0–9) resulting in 1800 reconstructed images. They were classified by a previously developed convolutional neural network (CNN), reaching a 92 % accuracy, which is a very good achievement in the case of multiple subjects.
Zusammenfassung
Die elektrische Impedanz Tomographie (EIT) ist ein bildgebendes Verfahren zur Charakterisierung der inneren Leitfähigkeitsverteilung eines Objekts basierend auf gemessenen Spannungen, die aus der Injektion eines AC-Wellenform Signales resultieren, gefolgt von einem Bildrekonstruktionsverfahren. Ein Algorithmus versucht, ein schlecht gestelltes inverses Problem zu lösen, das die Rekonstruktion eines genauen Bildes erschwert. Um dieses Problem zu lösen, schlagen wir einen genetischen Algorithmus (GA) für die Bildrekonstruktion mit einer nicht blinden Suchmethode vor, die Vorwissen über die mögliche Leitfähigkeitsverteilung im initialen Suchraum miteinbringt. Zur Validierung des Algorithmus wurden Experimente in einem Wassertank durchgeführt. Die Leistung des Algorithmus wurde im Hinblick auf die Bildqualität und die Verarbeitungszeit bewertet, wobei die entsprechende Qualitätsfunktion mit 100 Generationen unter Verwendung der nicht blinden Suche und der einheitlichen Kreuzung/Zufallsmutation auf 0,0505 minimiert werden konnte. Im Vergleich zu traditionellen Methoden erreicht der GA eine deutlich bessere Bildqualität. Er wurde als Bildrekonstruktionsalgorithmus für die Gestenerkennung implementiert. Es wurden EIT-Messungen mit sechs Personen durchgeführt, die amerikanische Zeichen (0–9) ausführten, die zu 1800 rekonstruierten Bildern führten. Sie wurden von einem zuvor entwickelten Convolutional Neural Network (CNN) klassifiziert und erreichten eine Genauigkeit von 92 %. Dies stellt im Falle mehrerer Personen eine sehr gute Leistung dar.
Funding source: Deutsche Forschungsgemeinschaft
Award Identifier / Grant number: SFB 1410/1
Funding source: Deutscher Akademischer Austauschdienst
Award Identifier / Grant number: #57424451
Funding statement: The German Research Foundation ‘DFG’ (Deutsche Forschungsgemeinschaft) for funding within the project SFB 1410/1 Hybrid societies, 2019 and German Academic Exchange Service (DAAD) by funds from the Federal Ministry for Economic Cooperation and Development (BMZ) Germany, within the project number #57424451.
About the authors
Mariem Hafsa received her diploma engineering degree in applied computer science from the National Engineering School of Sousse, Sousse, Tunisia in 2020. She did her graduation project at the Professorship Measurement and Sensor Technology (MST) at Chemnitz University of Technology, Chemnitz, Germany. Her main research topics are electrical impedance tomography, artificial intelligence, image reconstruction algorithms and medical imaging.
Bilel Ben Atitallah received his diploma engineering degree in electrical engineering from the National Engineering School of Sfax, Sfax, Tunisia in 2018. He did his graduation project at the Professorship Measurement and Sensor Technology (MST) at Chemnitz University of Technology, Chemnitz, Germany. Since 2020, he is a member of the CRC hybrid societies. His main research focus is on impedance spectroscopy, electrical impedance tomography, artificial intelligence, and methods for gesture recognition.
Taha Ben Salah received his PhD degree in telecommunications, microwave antenna and propagation in 2009 from the National Engineering School of Tunis, Tunis, Tunisia and the engineering degree in computer science in 2000 from the National School of Computer Science, Tunis, Tunisia. He was the director of the computing engineering department at the National School of Engineers of Sousse until 2019. He is a specialist of radio-frequency modeling (microwaves, antennas & wave propagation) and a certified software architect promoting Libre software initiatives. He worked in many international and national software development projects.
Najoua Essoukri Ben Amara was the director of ENISo, University of Sousse, Tunisia, from 2008 to 2011. Since 2011, she has been the president of the Tunisian Association of Innovative Techniques of Sousse. She is currently a full Professor of electrical engineering at the National School of Engineers of Sousse, ENISo. She was a coordinator of several European projects (Euromed 3C3 and Tempus). Her research focus is on pattern recognition, document analysis, multimodal biometric, medical image processing, computer vision, with application to the segmentation of documents, biometric recognition, individuals, and detection/monitoring multi-object.
Olfa Kanoun graduated in electrical engineering and information technology from the Technical University of Munich, Munich, Germany, in 1996. She carried out her Ph. D. until 2001 at the Institute of Measurement and Automation Technology, University of the Bundeswehr Munich, Munich. Since 2007, she has been a full professor in measurement and sensor technology at Chemnitz University of Technology, Chemnitz, Germany. Her research interests include impedance spectroscopy, sensors based on carbonaceous materials, and energy aware wireless sensors. She received six best paper awards on the international conferences. She published more than 400 papers in peer-reviewed scientific journals, book chapters, and international conferences.
Acknowledgment
The authors would like to thank the German Research Foundation ‘DFG’ (Deutsche Forschungsgemeinschaft) for funding within the project SFB 1410/1 Hybrid societies, 2019; applicant Prof. Dr.-Ing Olfa Kanoun. This research was also supported by the German Academic Exchange Service (DAAD) by funds from the Federal Ministry for Economic Cooperation and Development (BMZ) Germany, within the project number #57424451.
References
1. R. Barioul, A. Al Qudah, B. Ben Atitallah, O. Bouchaala, & O. Kanoun, (2020, June). Surface Electrical Impedance Myography Measurements for Recognition of Numbers in Sign Language. In 2020 IEEE International Conference on Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA) (pp. 1–6).10.1109/CIVEMSA48639.2020.9132745Search in Google Scholar
2. B. Ben Atitallah, R. Barioul, A. Ghribi, N. Derbel, & O. Kanoun, (2021). Electrodes Placement Investigation for Hand Gesture Recognition Based on Impedance Measurement. In 2020 17th International Multi-Conference on Systems, Signals & Devices (SSD).Search in Google Scholar
3. B. Ben Atitallah, R. Ramalingame, N. Derbel, & O. Kanoun, (2019, March). Muscle Movement Tracking Using Nanocomposite Based Pressure Sensor. In 2019 16th International Multi-Conference on Systems, Signals & Devices (SSD) (pp. 483–488).10.1109/SSD.2019.8893179Search in Google Scholar
4. B. Ben Atitallah, M. B. Abbasi, R. Barioul, D. Bouchaala, N. Derbel, & O. Kanoun, (2020, October). Simultaneous Pressure Sensors Monitoring System for Hand Gestures Recognition. In 2020 IEEE Sensors (pp. 1–4).10.1109/SENSORS47125.2020.9278884Search in Google Scholar
5. A. Widodo, (2018, November). Experimental study of one step linear Gauss-Newton algorithm for improving the quality of image reconstruction in high-speed Electrical Impedance Tomography (EIT). In IoP Journal of Physics: Conference Series (Vol. 1120, No. 1, p. 012067).10.1088/1742-6596/1120/1/012067Search in Google Scholar
6. O. Kanoun, U. Trö ltzsch, & H.-R. Tränkler, (2006, June). Benefits of evolutionary strategy in modeling of impedance spectra. In Electrochimica Acta (Vol. 51, pp. 1453–1461). 2006. Elsevier.10.1016/j.electacta.2005.02.123Search in Google Scholar
7. T. A. Khan, & S. H. Ling, (2019). Review on electrical impedance tomography: Artificial intelligence methods and its applications. In Algorithms (Vol. 12, No. 5, p. 88).10.3390/a12050088Search in Google Scholar
8. S. W. Huang, H. M. Cheng, & S. F. Lin, (2019, July). Improved imaging resolution of electrical impedance tomography using artificial neural networks for image reconstruction. In 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (pp. 1551–1554).10.1109/EMBC.2019.8856781Search in Google Scholar PubMed
9. A. Roy, D. Dutta, & K. Choudhury, (2013). Training artificial neural network using particle swarm optimization algorithm. In International Journal of Advanced Research in Computer Science and Software Engineering (Vol. 3, No. 3).Search in Google Scholar
10. M. Zhou, (2019). Electrical Impedance Tomography based on Genetic Algorithm. arXiv preprint arXiv:1901.04872.Search in Google Scholar
11. M. Hafsa, B. Ben Atitallah, B. T. Salah, Essoukri Ben Amara, N., & O. Kanoun, (2020) Hand Gesture Recognition based on Electrical Impedance Tomography Measurements using Genetic Algorithms. In 2020 13th International Workshop on Impedance Spectroscopy (IWIS) ISBN: 978-3-9817630-9-6 (pp. 58–59).Search in Google Scholar
12. PyEIT: python-based, open-source EIT framework. DoA: 19/09/2020 https://github.com/liubenyuan/pyEIT.Search in Google Scholar
13. N. M. Razali, & J. Geraghty, (2011, July). Genetic algorithm performance with different selection strategies in solving TSP. In Proceedings of the world congress on engineering (Vol. 2, No. 1, pp. 1–6). Hong Kong: International Association of Engineers.Search in Google Scholar
14. S. Saha, & S. Bandyopadhyay, (2008). Application of a New Symmetry-Based Cluster Validity Index for Satellite Image Segmentation. In IEEE Geoscience and Remote Sensing Letters (Vol. 5, No. 2, pp. 166–170).10.1109/ICIT.2006.15Search in Google Scholar
15. V. A. Barbosa, R. R. Ribeiro, A. R. Feitosa, V. L. Silva, A. D. Rocha, R. C. Freitas, & W. P. Santos, (2019). Reconstruction of electrical impedance tomography using fish school search, non-blind search, and genetic algorithm. In Biotechnology: Concepts, Methodologies, Tools, and Applications (pp. 2021–2038). IGI Global.10.4018/978-1-5225-8903-7.ch082Search in Google Scholar
16. M. Hafsa, B. B. Atitallah, T. B. Salah, N. Essoukri Ben Amara, & O. Kanoun, (2021, March). A parallel Architecture of a Genetic Algorithm for EIT Image Reconstruction. In 2021 18th International Multi-Conference on Systems, Signals & Devices (SSD) (pp. 965–970). IEEE.10.1109/SSD52085.2021.9429349Search in Google Scholar
17. B. Ben Atitallah, Z. Hu, D. Bouchaala, M. A. Hussain, A. Ismail, N. Derbel, & O. Kanoun, Hand Sign Recognition System based on EIT Imaging and Robust CNN Classification. In 2021 IEEE Sensors Journal, DOI: 0.1109/JSEN.2021.3130982.Search in Google Scholar
© 2022 Walter de Gruyter GmbH, Berlin/Boston
