SignificanceDiffuse correlation spectroscopy (DCS) is a powerful, noninvasive optical technique for measuring blood flow. Traditionally the blood flow index (BFi) is derived through nonlinear least-square fitting the measured intensity autocorrelation function (ACF). However, the fitting process is computationally intensive, susceptible to measurement noise, and easily influenced by optical properties (absorption coefficient μa and reduced scattering coefficient μs′) and scalp and skull thicknesses.AimWe aim to develop a data-driven method that enables rapid and robust analysis of multiple-scattered light’s temporal ACFs. Moreover, the proposed method can be applied to a range of source–detector distances instead of being limited to a specific source–detector distance.ApproachWe present a deep learning architecture with one-dimensional convolution neural networks, called DCS neural network (DCS-NET), for BFi and coherent factor (β) estimation. This DCS-NET was performed using simulated DCS data based on a three-layer brain model. We quantified the impact from physiologically relevant optical property variations, layer thicknesses, realistic noise levels, and multiple source–detector distances (5, 10, 15, 20, 25, and 30 mm) on BFi and β estimations among DCS-NET, semi-infinite, and three-layer fitting models.ResultsDCS-NET shows a much faster analysis speed, around 17,000-fold and 32-fold faster than the traditional three-layer and semi-infinite models, respectively. It offers higher intrinsic sensitivity to deep tissues compared with fitting methods. DCS-NET shows excellent anti-noise features and is less sensitive to variations of μa and μs′ at a source–detector separation of 30 mm. Also, we have demonstrated that relative BFi (rBFi) can be extracted by DCS-NET with a much lower error of 8.35%. By contrast, the semi-infinite and three-layer fitting models result in significant errors in rBFi of 43.76% and 19.66%, respectively.ConclusionsDCS-NET can robustly quantify blood flow measurements at considerable source–detector distances, corresponding to much deeper biological tissues. It has excellent potential for hardware implementation, promising continuous real-time blood flow measurements.

This study introduces a fast and accurate online training method for blood flow index (BFI) and relative BFI (rBFI) reconstruction in diffuse correlation spectroscopy (DCS). We implement rigorous mathematical models to simulate the auto-correlation functions (g ₂) for semi-infinite homogeneous and three-layer human brain models. We implemented a fast online training algorithm known as random vector functional link (RVFL) to reconstruct BFI from noisy g ₂. We extensively evaluated RVFL regarding both speed and accuracy for training and inference. Moreover, we compared RVFL with extreme learning machine (ELM) architecture, a conventional convolutional neural network (CNN), and three fitting algorithms. Results from semi-infinite and three-layer models indicate that RVFL achieves higher accuracy than the other algorithms, as evidenced by comprehensive metrics. While RVFL offers comparable accuracy to CNNs, it boosts training speeds that are 3900-fold faster and inference speeds that are 19.8-fold faster, enhancing its generalizability across different experimental settings. We also used g ₂ from one- and three-layer Monte Carlo (MC)-based in-silico simulations, as well as from analytical models, to compare the accuracy and consistency of the results obtained from RVFL and ELM. Furthermore, we discuss how RVFL is more suitable for embedded hardware due to its lower computational complexity than ELM and CNN for training and inference.

Cerebral blood flow (CBF) is a critical biomarker for brain health and function, supporting cognitive and neurological processes.¹ Existing technique, such as transcranial Doppler ultrasound, is noninvasive but relies on experienced operators and is limited to larger arteries.² Functional magnetic resonance imaging (fMRI) and positron emission tomography (PET), on the other hand, are bulky and provide only “snapshot” observations; they are not suitable for bedside applications.^3,4 By contrast, diffuse correlation spectroscopy (DCS) offers noninvasive, continuous, high temporal resolution CBF index (CBFi) measurements at the bedside.⁵ DCS analyzes light intensity fluctuations caused by red blood cell movement using the autocorrelation function (ACF), given by $g_2=⟨I(t)I(t+\tau )⟩/{⟨I(t)⟩}^2$, to assess blood flow dynamics.^6,7

Diffuse speckle contrast analysis (DSCA), also called speckle contrast optical spectroscopy (SCOS), has emerged as a groundbreaking optical imaging technique for tracking dynamic biological processes, including blood flow and tissue perfusion. Recent advancements in single-photon avalanche diode (SPAD) cameras have unlocked exceptional sensitivity, time resolution, and high frame-rate imaging capabilities. Despite this, the application of large-format SPAD arrays in speckle contrast analysis is still relatively uncommon. This study introduces a pioneering use of a large-format SPAD camera for DSCA. By harnessing the camera’s high temporal resolution and photon-detection efficiency, we significantly enhance the accuracy and robustness of speckle contrast measurements. Our experimental results demonstrate the system’s remarkable ability to capture rapid temporal variations over a broad field of view, enabling detailed spatiotemporal analysis. Through simulations, phantom experiments, and in vivo studies, we validated the proposed approach’s potential for cerebral blood flow and functional tissue monitoring. This work highlights the transformative impact of large SPAD cameras on DSCA, setting the stage for breakthroughs in optical imaging.

This study proposes a compact deep learning (DL) architecture and a highly parallelized computing hardware platform to reconstruct the blood flow index (BFi) in diffuse correlation spectroscopy (DCS). We leveraged a rigorous analytical model to generate autocorrelation functions (ACFs) to train the DL network. We assessed the accuracy of the proposed DL using simulated and milk phantom data. Compared to convolutional neural networks (CNN), our lightweight DL architecture achieves 66.7% and 18.5% improvement in MSE for BFi and the coherence factor /3, using synthetic data evaluation. The accuracy of rBFi over different algorithms was also investigated. We further simplified the DL computing primitives using subtraction for feature extraction, considering further hardware implementation. We extensively explored computing parallelism and fixed-point quantization within the DL architecture. With the DL model's compact size, we employed unrolling and pipe- lining optimizations for computation-intensive for-loops in the DL model while storing all learned parameters in on-chip BRAMs. We also achieved pixel-wise parallelism, enabling simultaneous, real-time processing of 10 and 15 autocorrelation functions on Zynq-7000 and Zynq-UltraScale+ field programmable gate array (FPGA), respectively. Unlike existing FPGA accelerators that produce BFi and the /3 from autocorrelation functions on standalone hardware, our approach is an encapsulated, end-to-end on-chip conversion process from intensity photon data to the temporal intensity ACF and subsequently reconstructing BFi and /3. This hardware platform achieves an on-chip solution to replace post-processing and miniaturize modern DCS systems that use single- photon cameras. We also comprehensively compared the computational efficiency of our FPGA accelerator to CPU and GPU solutions.

Diffuse correlation spectroscopy (DCS) is a widely used noninvasive optical technique for measuring tissue blood flow. Accurate blood flow estimation with DCS requires a high signal-to-noise ratio (SNR), but achieving high SNR is often limited by safety constraints on the optical irradiance (maximum permissible exposure) that can be delivered to tissue. To overcome this limitation, we investigated the possibility of replacing the conventional multi-mode fiber (MMF) with a liquid light guide (LLG) for illumination. The LLG provides a more uniform illumination profile and higher photon throughput to the tissue under the same irradiance limit, resulting in a significantly increased detected photon count rate and enhanced SNR. In experiments under identical power-density conditions, the LLG-based system achieved approximately a three-fold increase in SNR compared to the traditional MMF configuration. This improvement arises from the uniform beam profile and efficient light delivery of the LLG, which permits safe use of higher total power. These results indicate that LLG illumination effectively enhances DCS sensitivity without exceeding safety limits, potentially enabling more sensitive and accurate blood flow monitoring in biomedical applications.