Abstract
Functional near-infrared spectroscopy (fNIRS) and its advanced offshoot - diffuse optical tomography (DOT) are promising non-invasive neuroimaging techniques. The advancement of next-generation high-density fNIRS/DOT systems, particularly high-density wearable systems, requires compact light source arrays with high wavelength tuning precision and fine modulation capabilities. Vertical-cavity surface-emitting lasers (VCSELs) have emerged as a strong candidate for this purpose. However, VCSELs' performance is highly sensitive to temperature variations, where heating effects induce wavelength shifts and output power fluctuations, leading to measurement drift and reduced accuracy in fNIRS/DOT data. Conventional multi-channel VCSEL temperature control methods face constraints due to limited computational resources and poor scalability. To address these limitations, we propose a reconfigurable hardware-accelerated temperature control platform based on the heterogeneous ZYNQ-7000 Field-programmable Gate Array (FPGA). By integrating a real-time proportional-integral-derivative (PID) algorithm into the programmable logic (PL), the platform achieves precise temperature regulation with an error margin of ±0.01 °C. Experimental validation demonstrates the encouraging capability of this proposed platform to regulate the temperature of over 100 VCSELs simultaneously while maintaining low resource utilization, ensuring efficient parallel control with large channel counts in real-time. The proposed reconfigurable architecture significantly enhances the reliability and scalability of VCSEL-driven fNIRS/DOT systems while maintaining sufficient resources for future implementations of extra functions. This platform not only improves the thermal stability of VCSELs-based wearable high-density fNIRS/DOT devices but also establishes a robust thermal-control framework for broader applications requiring high-density, thermally stable light source configurations.