Master’s thesis research by Uğur Emin Baynal. Part of our project led by Prof. Dr. Aydın Akan, submitted to TÜSEB and granted pre-acceptance.
State of the Art
fNIRS uses light to measure changes in oxygenated and deoxygenated haemoglobin in the cortex, non-invasively and portably. We explained how it works in our fNIRS post. Limitations running from hardware to software can, however, increase bias in fNIRS research [1]:
- Motion artefacts are the largest source of corrupted measurements, and the balance between hardware and algorithmic solutions is still unclear [2].
- Extracerebral contributions. Because light first passes through the scalp and skull, systemic factors such as blood pressure can produce signals that look like brain activity [3].
- Optode–scalp contact is harder to achieve through dark, thick or curly hair. This is not only a technical issue but an inclusivity issue that leads to these people being excluded from research [1].
Signal quality is mostly assessed after recording, during analysis, when a corrupted channel can no longer be recovered. Machine learning methods that detect bad channels automatically are emerging [4, 5], but they run offline on desktop computers.
Excellence
Objectives
- O1. Develop a wearable fNIRS system that covers the whole head.
- O2. Turn signal quality control from a post-recording step into part of how the device behaves during recording.
- O3. Enable reliable measurements from people with a wide range of hair and skin characteristics.
Ambition and Progress Beyond the State of the Art
- Moving quality assessment from offline analysis onto the device itself, through embedded AI, aims to catch problems while they can still be corrected.
- Addressing optode contact as an inclusivity problem aims to widen who can take part in fNIRS research.
Methodology
The work combines the design of a whole-head wearable fNIRS system with embedded AI methods that run on the device during recording.
References
- Doherty, E. J. et al. (2023). Interdisciplinary views of fNIRS: Current advancements, equity challenges, and an agenda for future needs of a diverse fNIRS research community. Frontiers in Integrative Neuroscience, 17, 1059679. doi:10.3389/fnint.2023.1059679
- Huang, R., Hong, K.-S., Yang, D., Huang, G. (2022). Motion artifacts removal and evaluation techniques for functional near-infrared spectroscopy signals: A review. Frontiers in Neuroscience, 16, 878750. doi:10.3389/fnins.2022.878750
- Caldwell, M., Scholkmann, F., Wolf, U., Wolf, M., Elwell, C., Tachtsidis, I. (2016). Modelling confounding effects from extracerebral contamination and systemic factors on functional near-infrared spectroscopy. NeuroImage, 143, 91–105. doi:10.1016/j.neuroimage.2016.08.058
- Gerloff, C., Yücel, M. A., Mehlem, L., Konrad, K., Reindl, V. (2024). NiReject: toward automated bad channel detection in functional near-infrared spectroscopy. Neurophotonics, 11(4), 045008. doi:10.1117/1.NPh.11.4.045008
- Guglielmini, S., Chen, Z., Wolf, M. (2025). DL-QC-fNIRS: a deep learning tool for automated quality control in functional near-infrared spectroscopy signals. Neurophotonics, 13(1), 015001. doi:10.1117/1.NPh.13.1.015001