Rodríguez Rodríguez, ArmandoVento Álvarez, José RaúlGalarza Galarza, MarkoVanegas Tenezaca, Evelyn DayanaraSchuster, KayBravo Acha, MikelLópez-Amo Sáinz, Manuel2025-10-212025-10-212025-07-03Rodriguez, A., Vento Alvarez, J. R., Galarza, M., Vanegas-Tenezaca, E., Schüster, K., Bravo Acha, M. l, López-Amo, M. (2025). Torsion sensor using a high-birefringence nine-hole optical fiber. IEEE Sensors Journal, 25(16), 30692-30700. https://doi.org/10.1109/JSEN.2025.3577300.1530-437X10.1109/JSEN.2025.3577300https://academica-e.unavarra.es/handle/2454/55386This article presents a novel high-birefringence fiber-based torsion sensor based on a microstructured optical fiber with nine holes and seven cores microstructured holes and cores optical fiber (MHCF) embedded into a Sagnac interferometer (SI). A segment of this fiber is inserted into a symmetric SMF-MMF-MHCF-MMF-SMF arrangement, which provides efficient coupling to the multiple cores of the birefringent fiber and, consequently, multimode interference (MMI). Fast Fourier transform (FFT) spectral data analysis is employed to enhance measurement stability and reduce dependence on optical source variations. The sensor demonstrates a linear response to torsion angles between −50◦ and +50◦ , with a 16-mrad/◦ sensitivity. The high sensitivity and good linearity of the sensor are enhanced through the application of machine learning (ML) techniques.application/pdfeng© 2025 The Authors. This work is licensed under a Creative Commons Attribution 4.0 License.Machine learning (ML)Microstructured optical fiberSagnac interferometer (SI)Spatial frequencyTorsion sensorinfo:eu-repo/semantics/article2025-10-21info:eu-repo/semantics/openAccess