Impact of light polarization on laser speckle contrast imaging with a custom phantom for microvascular flow | Scientific Reports

Scientific Reports volume 14, Article number: 26652 (2024) Cite this article

Metrics details

Laser speckle contrast imaging (LSCI) is a non-invasive, powerful, and cost-effective imaging technique that has seen widespread adoption across various medical fields, particularly for blood flow imaging. While LSCI provides physicians with valuable insights into changes or occlusions in blood flow, the technique is susceptible to various factors and parameters that can impact measurement sensitivity and signal-to-noise ratio (SNR). These include the scattering of light, which can affect the quality and reliability of the LSCI data. The polarization of light holds significant promise to enhance the performance of LSCI. In this study, we employed polarization manipulation of light to investigate its impact on the performance of LSCI for measuring flow. Focusing on the application of LSCI in microcirculation within capillaries, we examined the effect of polarization control on the technique's flow measurement capabilities using a custom-designed phantom system. This phantom consisted of three tubes with inner diameters of 1.1 mm, 1.6 mm, and 2.8 mm, embedded in a polydimethylsiloxane (PDMS) matrix with optical properties similar to biological tissue. By manipulating the polarization of both the incident and reflected light, alternating between parallel and perpendicular states, we compared the performance of our LSCI system in detecting flow for different tube diameters and depths within the phantom. Our study revealed that while depth is a critical parameter influencing flow detection using LSCI, employing perpendicular polarization (between incident and reflected light) resulted in the lowest measurement error and highest SNR compared to parallel polarization and the absence of polarization control.

Laser speckle contrast imaging (LSCI), as a non-invasive, low-cost, fast, and easy to operate method, has found significant applications in various fields such as medicine1, material evaluation2, thermal characterization3, and force measurement4. In its simplest form the LSCI system can be implemented easily using a coherent laser source, a camera, a lens and a diffuser.

Laser speckle patterns can be classified into two categories: static and dynamic. Static speckle patterns are formed when coherent laser light illuminates a rough surface, and the scattered light from different parts of the surface interferes in the image plane, resulting in a granular pattern that remains stationary5. The speckle size in this pattern is determined by the wavelength of the light, the size of the laser beam, and the distance between the object and the location of the speckle pattern5. Dynamic speckle patterns, on the other hand, are generated by the movement of scattering particles within the medium. The temporal fluctuations in these patterns can be analyzed to generate 2D maps. Dynamic speckle patterns can be influenced by the presence of static scatterers, which can affect the speckle correlation and contrast in LSCI6. These speckle patterns can be exploited to detect motion and measure flow6. From practical point of view, the LSCI provides valuable information about the dynamics of particles scattering light within a random medium. In other word the technique analyzes the temporal fluctuations in speckle patterns caused by the movement of scattering particles and by analyzing these fluctuations, LSCI can generate 2D maps with high spatial and temporal resolution.

Despite considering the speckle patterns as noise in some imaging systems like sonography, biospeckle is widely used to follow biological activities and dynamic changes. There are many application for speckle pattern as LSCI system from agriculture to medicine1,7. Recently, biospeckle applications in medicine as a clinical tool have been increasingly observed. LSCI method is used for optical visualization of blood flow in biological tissues as an indicator to diagnose diseases8. This method has been used as an imaging system to monitor blood perfusion in various tissues, such as the brain9,10,11, gastrointestinal12,13, burns14,15,16, hand, foot17,18, and during surgery19,20,21. For instance, determining the burn depth of a skin based on visual information is one of the challenges of burn surgeons. An incorrect diagnosis of burn depth can lead to unnecessary hospitalization and grafting. The state of microcirculatory blood flow under the skin can be used to detect burn depth. According to Stewart et al. LSCI systems can provide valuable information about burn depth, which is a key factor in determining the appropriate treatment and prognosis for burn patients. Their findings suggest that LSCI could be a useful tool for burn triage, diagnosis, and monitoring22. In a systematic review, Konovalov et al. explored the application of LSCI in neurosurgical interventions, highlighting its potential benefits in real-time monitoring of blood flow during surgical procedures23.

Monitoring cerebral blood flow CBF during neurosurgery is a crucial evaluation criterion for assessing the procedure and mitigating the risk of mortality. Hecht et al. reported the use of LSCI system for real-time visualization of CBF in both the bypass graft and the cortical vasculature during neurovascular surgery24. LSCI is used in rheumatology to monitor changes in peripheral blood perfusion (PBP) in real-time, which can provide valuable insights into the progression of diseases such as systemic sclerosis (SSc) and arthritis. Ruaro et al. used a commercial laser speckle contrast analysis (LSCA) device (Pericam PSI, Perimed, Jarfalla, Sweden) to monitor and analyse blood perfusion in SSc patients. Systemic sclerosis is an autoimmune disease that leads to fibrosis of internal organs, skin, and reduction of peripheral blood perfusion25. In their recent study on the clinical applications of laser speckle, Bi, Renzhe, et al. introduced a new device called portable optical diffuse speckle pulsatile flowmetry (DSPF). This device features a flexible handheld probe designed for measuring deep tissue blood flow in the human foot, facilitating ischemia assessment and aiding in the diagnosis of peripheral artery disease (PAD)26. LSCI has been used in the abdominal cavity during surgeries to identify ischemic areas and reduce anastomotic leakage, which is a major complication within gastrointestinal surgery. Eriksson et al. reported the use of LSCI system for assessing liver microcirculation in undergoing liver resection cases27. By the way, most of the studies have been performed in an open surgical setting, and there is a need for more research in endoscopic or laparoscopic settings. Ophthalmology is one of the most extensive fields of research in the application of LSCI, particularly for assessing retinal and choroidal blood flow dynamics, which are crucial for diagnosing various eye diseases such as glaucoma, retinopathy, and macular degeneration. Tamaki et al. used LSCI system for retinal, optical nerve head (ONH), and choroidal blood flow imaging28.

Optical coherence tomography (OCT) is a powerful visualization system used to assess retinal perfusion. This imaging technique utilizes low coherence interferometry to produce high-resolution cross-sectional images of the retina and optic nerve, making it a valuable tool for evaluating ophthalmic patients29. With ongoing technological advancements, the application of OCT in neuro-ophthalmology continues to expand. In recent years, LSCI has emerged as a method for measuring blood flow in retinal vessels. For instance, Luft et al. evaluated the validity of retinal perfusion measurements using laser speckle flowgraphy (LSFG) and compared these findings with those obtained from OCT in a healthy Caucasian population30. OCT operates by penetrating tissue to a depth of several hundred microns using near-infrared light interferometry. It measures backscattered light to reconstruct detailed images of the retinal structure. In contrast, LSCI utilizes the speckle pattern created by backscattered light to visualize blood flow dynamics. Another optical imaging technique is the confocal microscope, which employs laser light to generate images. This method involves moving a focused laser spot across the sample and collecting data point by point. The images obtained at various depths are then reconstructed into a three-dimensional representation31. Compared to LSCI, confocal microscopy provides higher optical contrast of the internal components of a sample, allowing for detailed structural imaging. However, LSCI primarily detects blood flow, providing functional images that are essential for assessing hemodynamics. LSCI systems are generally more straightforward to operate, offering a user-friendly method for measuring blood flow dynamics without requiring extensive setup or calibration. This simplicity, however, comes at the cost of the detailed structural information that confocal microscopy and OCT can provide. Sdobnov et al. compared non-invasive imaging systems derived from dynamic light scattering (DLS) techniques for blood flow imaging, such as laser Doppler flowmetry (LDF), diffusing wave spectroscopy (DWS), LSCI, and Doppler optical coherence tomography (DOCT). The study assessed these methods across various characteristics, such as spatial and temporal resolution, penetration depth, and scattering regime32.

Flow velocities in cardiovascular system has a wide range depending on vessels diameters. Red blood cells (RBCs), as most abundant cell type in blood, are key cells that suspended in plasma and has essential role in oxygenation to all body cells. The deformability of RBCs helps their movement in the capillary diameter in the microcirculation system. The microcirculation system is responsible for the oxygen transport of the capillaries, where the oxygen provides energy to the tissue cells. The motion of RBCs in microcirculation system will generate fluctuations in backscattered lights that cause blurring in the speckle pattern on the detector. The fundamentals of the LSCI system are based on the blurring caused by the light fluctuations and the motion of RBCs1,33,34.

The optimizing of the precision in detecting flow with the LSCI method has been performed in different studies to determine the effect of speckle-to-pixel size ratio, exposure time, and polarization on results35,36. The effect of exposure time on laser speckles based on the wavelet approach was investigated and the results showed that increasing exposure time increased the penetration depth37. Singh et al. investigated the effect of polarization on LSCI using a phantom with a flow rate of 2 mm/s. They employed an intensity variation method in the contrast image to evaluate the system's performance at this specific flow rate38.

Compared to laser Doppler flowmetry (LDF), LSCI is an imaging method that provides high spatial and temporal resolution, offering the ability to assess microcirculation over a large area with rapid imaging. In LSCI, the result of the flow measurement is a high-resolution, two-dimensional image, whereas in LDF, the result is a signal representing the flow. One notable limitation of the LSCI method is that it measures flow in relative terms and does not provide absolute flow values. This means that the results are presented in arbitrary units and require calibration, making it challenging to directly compare them with absolute flow values1. One of the significant limitations of the LSCI method is its low sensitivity to tissue depth, effectively rendering it a surface imaging technique. This limitation is significant because it may restrict its ability to accurately assess structures located deeper within tissues, thereby influencing the reliability and applicability of the LSCI in certain contexts.

Polarization of light plays a significant role in various optical techniques, such as fluorescence, Raman spectroscopy, polarization microscopy, birefringence measurements, and circular dichroism spectroscopy39,40,41. Polarization can affect the efficiency, sensitivity, and specificity of these techniques, making it a critical parameter to consider in experimental design and data analysis. Controlling the polarization state of light can enhance the performance of these techniques, improve the signal-to-noise ratio (SNR), and gain insights into the properties of the sample under investigation.

Light polarization can enhance the depth sensitivity of the LSCI method. By changing the polarization state between incident and reflected photons, the influence of specular reflections from the surface can be reduced, improving the method's ability to assess deeper tissue layers. However, as the polarized light penetrates deeper into the tissue, it becomes more scattered and less polarized42,43. Heran et al. used speckle pattern analysis and light polarization to predict the optical absorption and scattering properties of materials44. Tchvialeva et al. used polarization speckle imaging to analyze the depolarization ratio of traditional speckle patterns as a potential tool for skin cancer detection45.

In this study, we utilized polarization states of the incident and reflected light to enhance depth penetration sensitivity and explored the LSCI system's response to vessel diameter variations at different depths. This investigation involved design and fabrication of a phantom vessel containing tubes of three different sizes (simulating different vessel diameters) and different burial depths (simulating of skin layers). The LSCI setup capable of managing the polarization state of both the incident and reflected light was designed and constructed. The images extracted from the raw speckle movie frames were processed by analyzing regions of interest within the phantom tube. Subsequently, new set of images with a color map indicating flow information were generated, which was presented as a dynamic movie. Our study evaluated the system's ability to detect flow at varying depths and tube diameters using a customized phantom setup, while also considering the impact of polarized state of the light on speckle imaging.

The speckle contrast is calculated by Eq. (1) that introduced by Fercher and Briers8:

where K, \(\sigma\) and \(<I>\) are contrast, standard deviation and mean intensity respectively. K is between [0–1], lower value means more blurring, and therefore more flow. In LSCI, the relation between contrast and flow is defined by calculating the relation between the decorrelation time (τ) and the speckle contrast (K). The intensity of the speckle pattern is recorded as an image in the camera. An LSCI image is created by recording the fluctuation of the speckle pattern over time by adjusting the camera's exposure time. The exposure time is a critical parameter in LSCI, as it affects the sensitivity of the technique to different flow rates. By optimizing the exposure time, LSCI can provide accurate and detailed information about blood flow distribution. The relation between contrast and flow is defined by Eq. (2)6:

where K, \({\tau }_{c}\) and T are contrast, decorrelation time and exposure time respectively. Technically, in this paper, the image of K-factor values (derived from the raw speckle images) is referred to as the LSCI image. This equation is derived from the relation between the field autocorrelation and the intensity autocorrelation, assuming a Lorentzian distribution of the field. Finally, the relation between flow rate and contrast can be approximated by Eq. (3)6:

where V represents the flow rate, which can be approximated by the inverse of the decorrelation time (\({\tau }_{c}\)) and the inverse of the squared contrast (\({K}^{2}\)). Therefore, by utilizing Eq. (3), it is possible to estimate the relative flow rate based on the speckle contrast measured in LSCI images (K).

Our LSCI setup, illustrated in Fig. 1, offers the flexibility to manipulate the polarization of both incident and reflected fields to provide the possibility of having different configurations. The laser light (650 nm, 50 mW) passes through a diffuser (D) to ensure uniform illumination. The total laser power used in the experiments was 50 mW. The beam was expanded to cover an area of approximately 64 cm2, resulting in a laser power per unit area of 0.78 mW/cm2. The 650 nm wavelength was chosen to optimize the detection of backscattered light from shallow tissue layers, which is crucial for accurately simulating the microcirculation in the phantom. To refine the shape of laser beam, the diffused light undergoes the aperture (A) before being expanded by a Keplerian beam expander composed of two lenses (L1 with a focal length of 15 mm and L2 with a focal length of 100 mm). Subsequently, a linear polarizer (P1) is employed to manipulate the polarization of light. The polarized light is then directed onto the sample using a mirror (M). The reflected light from the sample, passes through a second linear polarizer (P2) and is captured by a magnifying lens (L3 with a focal length of 100mm), ultimately being recorded by a CMOS camera for analysis. By arranging P1 and P2 in a certain way, it is possible to achieve different polarization states for the incident and reflected light. To control the flow rate of the liquid within the sample, a syringe pump is employed.

Schematic diagram of LSCI setup. D: diffuser, A: aperture, L1–L2: beam expander lenses, P1–P2: polarizer lenses, L3: magnifying lens, and M: mirror.

We have performed LSCI for three configurations: without a polarizer, with parallel polarization between the incident and reflected light path, and with perpendicular polarization. By manipulating the polarization states, it is possible to suppress the reflected light from the surface, leading to a stronger signal from the depth of the sample. In this regard, measurements were conducted within the phantom using tubes of varying diameters at different depths to examine the effects of both depth and tube diameter.

Different phantoms have been used to study the flow of blood46. We have fabricated the phantom using 184 Sylgard PDMS gel (Dow Corning, Midland, MI) based on its optical properties, which are similar to the biological tissues47,48,49,50. In order to make the phantom, we mixed PDMS with TiO2 to create optical properties similar to the tissue51,52,53. Layers of PDMS material were created with thicknesses ranging from from 0.1 to 1 mm to add depth and encase the phantom tubes. The dimensions of phantom were 10 × 50 × 60 mm with three plastic tubes having internal diameters of 2.8, 1.6, and 1.1 mm (T1, T2, and T3) embedded within it. To make the phantom mold, plexiglass material was used, which was designed and cut to the above-mentioned dimensions. Figure 2 shows the schematic and real images of the phantoms used in this study. As shown in part A, there are ten layers (L1:L10) with 0.1 mm thickness used to bury the phantom tubes at depths of 0.1 to 1 mm. We simulated different depths of phantom tubes, which most resemble the burial of capillaries under the skin. In our setup, different depths are considered from zero depth (L0: without any layer) to 1 mm depth to record raw speckle movies. Therefore, we prepared 10 different layers from 0.1 to 1 mm (L1:L10) with 0.1 mm steps to use on top of the phantom.

Designed phantom from different view with embedded tubes shown in each parts (T1: 1.1 mm, T2: 1.6 mm, and T3: 2.8 mm). (A) Schematic side view of the phantom. Phantom layers (L1:L10) were used to create depth ranging from 0.1 mm to 1mm (L0: zero depth, without any layer). The depth was increased incrementally by 0.1 mm. (B) Schematic 3D view of the phantom. (C, D) Photo of the real phantom with embedded tubes (T1: 1.1 mm, T2: 1.6 mm, and T3: 2.8 mm).

In order to control the flow, a syringe pump (B. BRAUN, Perfusor compact S) has been employed. A 50 ml syringe was used to inject the solution to the phantom. The maximum flow rate of pump was 200 ml/h. We have used 10% Lipofundin MCT/LCT intralipid solution (manufactured by B. BRAUN in Germany) to study blood flow. This solution has optical properties similar to those of blood, making it a suitable substitute for LSCI.

The capillary flow velocity was not high enough to allow sufficient time for oxygen and nutrient exchange. In the experimental studies, the average capillary velocity are measured in the range of 0.6–1.5 mm/s for different tissues54. The diameter of most capillaries is on the order of 8–10 µm. Because of the limited tube diameter in our designed phantom, we investigated low-range flow in a capillary tube with a diameter of 1.1 mm. Therefore, we considered flow rates of 0, 0.5, 1 and 1.5 mm/s with a step of 0.5 mm/s for the capillary flow range, and flow rates of 2, 3, 4, 5 and 10 mm/s for larger diameters of the phantom as well.

Figure 3 shows that the epidermis is the outermost layer of the skin with an average thickness of less than 1 mm and a maximum thickness on the soles and palms. The capillaries are located in the papillary region of dermis which is directly beneath the epidermis55,56. Therefore, we designed the phantom and buried the tubes in layers to investigate the effect of depth on the system’s ability to measure capillary flow.

Schematic diagram of skin layers.

The CMOS camera (Canon EOS 4000D),featuring an 18-megapixel CMOS sensor, a pixel size of 4.3 µm, and a bits, depth of 14-bit was used as the detector of the system. The camera used in the study had a zoom lens with a focal length range of 18mm to 55 mm and an aperture range f/5.6 to f/36. The focal length determines the field of view, with a shorter focal length providing a wider view and a longer focal length providing a narrower view and greater magnification. The f/number determines the lens's aperture, with a smaller f/number indicating a larger aperture and more light entering the camera. The f/number also affects the depth of field, with a smaller f/number providing a shallower depth of field and a larger f/number providing a deeper depth of field. The exposure time was adjusted between 1/30 s and 1/4000 s, while the ISO value ranged from 100 and 6400. Calculations and analyzes were performed in MATLAB R2023a on an ASUS ROG STRIX G laptop with a GPU graphics processing unit (NVIDIA GeForce RTX 1660Ti) with 16 GB of RAM. EOS Utility was also used to record and transfer videos from the camera to the computer. Then the recorded videos were imported into the MATLAB environment, contrast analysis was performed on each frame of the recorded movie and a new LSCI movie was reconstructed. Each LSCI frame was reconstructed from the average contrast analysis of 10 raw speckle frames. To draw a graph or signal diagram, we averaged 10 s of LSCI movie frames. Since relative flow is calculated with this method, we normalized the LSCI values obtained from the average of 250 frames (corresponding to 10 s of movie at a frame rate of 25) for different flow rates. Consequently, all graphs and tables presented in the result section use normalized LSCI value between 0 and 1. All analysis, processing and graphing were performed using MATLAB software.

The SNR of the laser speckle contrast image is considered as criteria. The SNR is mathematically defined in Eq. (4)35,57.

where the M and Var are the mean and variance of the image, respectively. Enhancing the SNR performance is crucial because it directly affects measurement accuracy. To ensure high accuracy and SNR, we investigated the effects of light polarization on speckle contrast and SNR in reducing vessel size diameter and increasing vessel burial depth for LSCI. A high SNR indicates that the contrast imaging has a small variation compared to the mean, whereas a low SNR indicates significant variation in the mean.

Mean error is another factor that we also considered as a criterion. is calculated using Eq. (5)35.

where <K> is the mean LSCI that is calculated from Eq. (1), and x, y are the neighborhoods pixels of the images, and N is the number of frames.

Camera parameters (exposure time, ISO, f/number, frame rate) are effective parameters in improving the quality of speckle contrast images. These parameters were adjusted and the optimal values were used in this study.

Increasing the exposure time (T) brightens the image, as more photons hit the sensor, but makes it difficult to record fast movements and can result in blurring. Decreasing T beyond an optimal point worsens the SNR of the image. However, low values of T are necessary for low-level light applications, such as LSCI, because speckle patterns appear in the image when the amount of scattered light reaching the camera sensor is sufficient. A shorter exposure time allows for the capture of faster dynamics in the speckle pattern, providing more detailed information about the movement and flow within the imaged object. Therefore, the precise setting of T is important for capturing details in LSCI. This is because the flow can be detected by reducing the contrast and amount of blurring in the speckle pattern of the image. In video recording, T refers to the amount of time each individual frame is exposed to light, which can significantly influence the quality and clarity of the captured footage. In video recording, if the frame rate of the camera is 25 frame per second (fps), the T is typically set to 1/50 of a second, which is the reciprocal of the frame rate. On the other hand, setting the exposure time also depends on other factors such as tube diameter and flow rate. To maintain a consistent intensity range across all movie frames, we employed a multi-exposure time approach. By utilizing a range of exposure times, we ensured that the intensity remained within an optimal range, preventing both over- exposure and under-exposure. This approach allowed us to maintain a constant intensity while still benefiting from the increased penetration depth provided by the perpendicular polarizers.

We have tailored T according to our setup for different light polarization states as shown in Table 1. This customization allows us to optimize the imaging conditions for varying experimental parameters, ensuring accurate and detailed results in LSCI.

As can be observed, the lowest T was used for the largest diameter tube, while the highest T was used for the narrowest tube in the cross polarization state. This adjustment was necessary because the placement of the polarizer in the light path reduced the intensity of the scattered light returning to the camera. By increasing T, we were able to obtain better results in this situation. This highlights the importance of tailoring the exposure time to the specific experimental conditions to achieve accurate and detailed imaging in LSCI.

Generally, in movie recording, the shutter speed is set to approximately double the frame rate to minimize motion blur. In this study, we utilized a frame rate of 25 fps, resulting in an appropriate exposure time of about 1/50 s, which was applied to both the perpendicular and parallel polarization conditions. For the condition without a polarizer, we opted for a shorter exposure time to prevent saturation of the camera sensor and to maintain a consistent intensity range across all movie frames. Additionally, the shortest exposure time was employed for the largest diameter tube, while longer exposure times were used for narrower tubes to ensure uniform intensity levels. This approach allowed us to maintain the mean laser speckle contrast within a specific range of 0.165 ± 0.02 throughout the experiments.

The f-number of camera was set to f#8 to prevent image saturation from excessive light. This setting provided an adequate balance between the amount of light and the depth of field required for the setup. If fewer photons had entered the camera, it would have been more challenging to obtain clear speckle patterns. The ISO setting of the camera, which indicated its sensitivity to light, was set to 400. This value was chosen to balance the need for a bright image with the need to minimize the noise that can result from higher ISO values. After setting the parameters of the camera and receiving the raw speckle image, a color map of the flow was obtained.

In Fig. 4, the raw speckle images and selected region of interest (ROI) are displayed. We employed the Global Thresholding method to segment the tube in each frame of the recorded images. This approach allowed us to effectively distinguish the tube from the background, ensuring accurate analysis of the speckle patterns. Each image represent a frame extracted from the recorded movie. After processing and implementing algorithms, we transformed raw images to display flow information through the contrast. Finally, the flow images obtained from LSCI analysis at various flow rates (0, 0.5, 1, 2 mm/s) are depicted in Fig. 4. As the flow rate increases, the color map of the tube shifts toward red, visually indicating the heightened flow rate. Contrast and flow analysis were conducted for all images. To generate the speckle contrast image, we utilized the average of 10 frames extracted from the movie.

(A, B) The selected ROI of raw speckle image (RGB and grayscale images), (C–F) the flow (1/k2) images related to (LSCI) with colormap.

Figure 5 show the flow images obtained from LSCI analysis at various flow rates (0, 0.5, 1, 1.5, 2, 3, 4, 5 mm/s). As expected, when the flow rate increases, the color map of the tube shifts towards red, providing a visual indication of the heightened flow rate. Similar to the Fig. 4, the flow analysis in the Fig. 5 is extracted from one frame of the LSCI result, which is produced from the average of 10 raw speckle frames from the recorded movie. The images are related to the case with no polarization adjustment and zero depth.

Flow (1/k2) images with colormap of (A) 0 mm/s, (B) 0.5 mm/s, (C) 1 mm/s, (D) 1.5 mm/s, (E) 2 mm/s, (F) 3 mm/s, (G) 4 mm/s, (H) 5 mm/s.

Figure 6 shows the decrease in contrast caused by the increase in the movement of the solution in the phantom tubes. This decrease was expected in LSCI due to the decrease in contrast of the raw speckle image. Contrast analysis (LSCI) is shown for three tube diameters (at zero depth, and no polarization adjustment). LSCI in each flow rate was analyzed from the average of 10 s of raw movie. LSCI analysis is normalized in different flow rates between 0 and 1.

The average laser speckle contrast image (K) plot for different tube diameters (1.1, 1.6, 2.8 mm) of phantom and different flow rates from 0 to 10 mm/s.

Figure 7 shows the bar graph of normalized LSCI at zero depth (L0, without any layers on the phantom) for three polarization states. The results indicate the decrease in normalized LSCI for increasing the flow rate. We calculated the LSCI for three polarization states between two polarizers in incident and reflected branch of the system: without adjustment, parallel and perpendicular (cross) polarization. The graphs of parts A, B, and C correspond to the tube diameter of 1.1, 1.6, and 2.8 mm, respectively. The reduced values of the normalized LSCI are obsereved with increasing flow rates in three polarization states. Therefore, the system was able to detect the flow rate (even very low flows) in all different polarizations at the zero depth (L0, without any layers on phantom). Values have been calculated for different flow rates (0, 0.5, 1, 1.5, 2, 3, 4, and 5 mm/s). We selected low flow rates, approximately 0.5 mm/s to resemble the capillary flow in the human body and then increased the flow by enlarging the tube diameter. The results of our LSCI system demonstrated its effectiveness in detecting flow at zero depth within the phantom. Notably, the LSCI signal decreased as the flow rate increased, which is a favorable outcome.

Bar graph of normalized LSCI calculated at zero depth (L0, without any layers on the phantom) for three polarization (no polarization adjustment, parallel, and cross) states and different flow rates from 0, 0.5, 1, 1.5, 2 ,3, 4, 5, and 10 mm/s for tube diameters (A) 1.1 mm, (B) 1.6 mm, and (C) 2.8 mm.

The primary objective of our analysis was to investigate the sensitivity of the system to various factors, including the burial depth of the tubes containing the flow, the effect of tube diameter, and the polarization of the light. To achieve this, we compared the results and made efforts to minimize errors as much as possible. By adjusting the polarization, we aim to reduce errors in LSCI analysis and enhance depth sensitivity in cases where the tube’s burial depth is increased and its diameter is reduced.

Figure 8 shows the bar graph of the normalized LSCI results for different depths of phantom tubes from 0.1 to 1 mm (L1 to L10). Our goal was to investigate the system’s sensitivity to different burial depths and the effects of polarizations. The bar graph shows normalized LSCI for three polarization (without adjustment, parallel, cross) and different flow rates from 0 to 10 mm/s for different tube diameters. The bar graph displays the normalized LSCI values for different tube diameters and depths. The leftmost column corresponds to a tube diameter of 0.1 mm, the middle column to 1.6 mm, and the rightmost column to 2.8 mm. The depths of the tubes are indicated next to each row in the figure. The first row (L1) shows the LSCI results for a tube depth of 0.1 mm, the second row (L2) corresponds to a depth of 0.2 mm, the tenth row (L10) represents a depth of 1 mm. The normalized LSCI values are plotted for each combination of tube diameter and depth. In first and second rows (depths of 0.1 and 0.2 mm), results show the reduced values of normalized LSCI with increasing flow rate in all three polarization configurations. Therefore, the system can detect flow rates, including very low flows, up to a depth of 0.2 mm without any issues. In the third and fourth rows (depth of 0.3 and 0.4 mm) with perpendicular polarization, our LSCI system detected increases in flow rate without any errors. However, for other polarization configurations the normalized LSCI values in some cases did not decrease with increasing flow rates and were considered as errors. In the fifth and sixth rows (depth of 0.5 and 0.6 mm) in all polarization configurations, we encountered errors in detecting increase in flow rate. In general, the failure detection of flow rates inour LSCI system occured at higher flow rates (around 5 mm/s) for the thinnest tube, while for the thicker tube, the flow rate was close to zero (0.5 mm/s). In the seventh, eighth and ninth rows (depths of 0.7, 0.8, and 0.9 mm), our LSCI system operated without error for the thickest tube in cross polarization, however, it encountered errors with others. For the last layer (1 mm), the system had errors for all configurations and tube diameters.

Bar graph of normalized LSCI calculated for three polarization (no polarization adjustment, parallel, cross) and different flow rates from 0 to 10 mm/s for different tube diameters. The left, middle, and the right column correspond to the tube diameter of 1.1 mm, 1.6 mm, and the tube diameter of 2.8 mm, respectively. The burial depth of the phantom tubes is written next to each row of the figure.

In general, the errors were less in perpendicular polarization, especially in the thickest tube. It indicates that at depths greater than 0.9 mm, the system experienced errors and was unable to detect the increase in flow. Additionally, the error percentage increased, and the expected decreasing pattern associated with rising flow rate was not observed. Practically, our system in cross polarization can detect flow at a depth 1 mm in the thickest tube diameter. However, for a tube diameter of 1.1 mm at shallow depths, maximum of 0.7 mm can be detected with error. Additionally, because capillaries form of a network under the skin rather than existing as a single capillary, it is feasible to detect decreasing patterns in real tissue at greater depths. Therefore, by polarizing the light, we were able to reduce errors and enhance the LSCI reduction pattern.

Figure 9 shows the percentage of error in LSCI that occurred in each polarization configuration. This error was calculated by dividing the number of instances where LSCI increased (instead of decreased) by the total number of measurement for each tube diameter (number of errors/total number of measurements)✕100. The results indicate that the lowest error rate belongs to the cross polarization, particularly for the thickest tube diameter (2.8 mm).

Bar graph of error (percentage of LSCI increasing values) for all polarization and different tube diameters.

Figure 10 shows the bar graph of the mean error and the mean SNR averaged across all polarization states (A, C) and all tube diameters (B, D). As seen in part A of Fig. 10, the 2.8 mm tube diameter has the lowest error, while cross polarization exhibits the lowest error in part B. These findings are consistent with the parts C and D, which indicate that the 2.8 tube diameter and the cross polarization yield the highest SNR.

Bar graph of normalized LSCI calculated for three polarization (no polarization adjustment, parallel, and cross) and different flow rates from 0 to 10 mm/s for different tube diameters.

In general, the results showed that as the diameter of the tube (vessel) increases, the LSCI system detects the flow rate with less error and higher SNR. Additionally, perpendicular polarization can enhance penetration depth. It is essential to note that the cross configuration provides better SNR and less error, whereas in the absence of polarizers and with parallel configuration, SNR decreases and error increases. Our result indicate that polarization has a significant effect on both SNR and error. This effect may be attributed to the filtering of the unwanted scattering components through cross polarization.

Circularly polarized light (CPL) presents unique advantages for LSCI, particularly in turbid media such as biological tissues. Its superior polarization memory allows for deeper penetration and greater interaction with scattering media compared to linear polarization. This enhanced interaction may improve depth sensitivity by reducing the effects of scattering, which is particularly important in biological tissues where multiple scattering events can obscure imaging signals. However, implementing CPL in LSCI systems can be complex. The generation and maintenance of circular polarization require precise optical components, such as wave plates and polarizers, which can introduce additional challenges in alignment and calibration. This complexity may increase the cost and time required for system setup and maintenance. While CPL is less susceptible to cross-polarization interference, allowing for more reliable signal detection from various orientations of scattering particles, this benefit can be offset by the need for sophisticated optical configurations. Furthermore, the robustness of CPL against multi-path interference—where signals bounce off surfaces and create complex interference patterns—can lead to clearer imaging results. However, this same robustness may require careful consideration of the optical path and environmental conditions to avoid artifacts that could compromise image quality. Additionally, while CPL can provide a more uniform illumination across the imaging area, the intricacies of achieving and maintaining this uniformity in practice can pose challenges. Variations in the optical properties of the biological tissues being imaged may also affect the performance of CPL, potentially leading to inconsistencies in SNR across different imaging scenarios.

It is important to note that the LSCI method measures relative blood flow rather than providing absolute flow values. The results are presented in arbitrary units, necessitating calibration for meaningful comparisons with absolute flow metrics. To address this limitation, careful design of suitable phantoms and consideration of various parameters prior to device application are essential. However, it is worth mentioning that relative measurements can be highly valuable in clinical follow-up applications, particularly for patients with burns or wounds, where monitoring changes in blood flow is critical. Another significant limitation of the LSCI technique is its reduced sensitivity to tissue depth, effectively making it a surface imaging method. This limitation may hinder its ability to accurately assess deeper structures within opaque tissues. Nevertheless, leveraging the properties of light polarization may help mitigate this issue and enhance depth sensitivity.

One key advantage of using light polarization in LSCI is its potential to enhance depth sensitivity. By manipulating the polarization state between incident and reflected photons, the influence of specular reflections from the surface can be reduced, improving the method's ability to assess deeper tissue layers. Additionally, controlling the polarization state of light can enhance the performance of these techniques, boost the SNR, and provide valuable insights into the properties of the sample under investigation. However, there are notable disadvantages associated with the use of polarization in this system. The addition of a polarizer introduces complexity to the LSCI setup, requiring precise optical components that can increase both the cost and the time needed for system calibration and maintenance. Furthermore, the use of polarization can lead to a reduction in the light intensity reaching the camera sensor, which decreases the average image intensity. This reduction can be mitigated by adjusting the exposure time, but care must be taken to avoid saturation or underexposure of the image, which could compromise the quality of the speckle contrast measurements.

In comparison to previous studies examining the effects of polarization on LSCI, this study offers a more comprehensive quantitative analysis of how polarization influences LSCI performance. We systematically evaluated measurement error and SNR across various polarization states, tube diameters, and depths. This thorough analysis enables us to draw more precise conclusions regarding the optimal polarization configurations for enhancing LSCI in microvascular flow imaging.

The results of this study could have significant implications for various practical applications. By leveraging polarization, LSCI can provide more accurate measurements of blood flow in capillaries, enabling better evaluation of tissue health and healing in conditions like diabetes. Polarization-enhanced LSCI can offer deeper insights into the wound bed, allowing clinicians to better assess the healing process and make more informed treatment decisions. In burn patients, polarization can help LSCI systems penetrate through the eschar and provide a more comprehensive assessment of tissue perfusion, guiding surgical interventions. In neuro-ophthalmology, polarization-sensitive LSCI can offer improved visualization of retinal blood flow, potentially leading to earlier detection and better management of conditions like glaucoma and diabetic retinopathy. During surgical procedures, polarization can enhance the ability of LSCI to detect subtle changes in blood flow, helping surgeons make more informed decisions about tissue viability and surgical planning.

To further optimize the LSCI parameters for different clinical scenarios, exploring the following strategies is recommended. Experimenting with various polarization configurationscan help enhance depth sensitivity and reduce specular reflections in specific applications. Fine-tuning the exposure time based on factors like tissue type, depth, and blood flow dynamics can help maintain an optimal speckle contrast range for improved SNR and accuracy. Balancing the spatial and temporal resolution of the LSCI system can be crucial for capturing relevant blood flow changes in different clinical contexts, such as microcirculation monitoring or large vessel assessment. Establishing well-defined protocols for parameter optimization, calibration, and data analysis can facilitate the adoption of LSCI in clinical settings and enable meaningful comparisons across studies and institutions.

To study the effect of polarization on the performance of the LSCI system for flow detection, we fabricated a phantom with three tubes of varying diameters (1.1 mm, 1.6 mm, and 2.8 mm) buried at different depths. We then set up an LSCI system capable of controlling the polarization of both the incident and reflected light. By alternating between parallel and perpendicular polarization configurations, we evaluated the system's performance in detecting flow within the phantom. The results demonstrate that light polarization, as a powerful intrinsic property of electromagnetic waves, can effectively mitigate the impact of unwanted scattering components that degrade the SNR in LSCI. Specifically, the perpendicular configuration of the incident and reflected light polarizers yielded the best performance, outperforming both the parallel polarization and the absence of any polarization control.

All data generated or analysed during this study are available from the corresponding author on request.

Heeman, W. et al. Clinical applications of laser speckle contrast imaging: A review. J. Biomed. Opt. 24(8), 080901–080901 (2019).

Article ADS PubMed PubMed Central Google Scholar

Cikalova, U., Bendjus, B. & Schreiber, J. Laser-Speckle-photometry—A method for non-contact evaluation of material damage, hardness and porosity. Mater. Testing 54(2), 80–84 (2012).

Article ADS Google Scholar

Mukhurov, N., et al. A speckle-photometry method of measurement of thermal diffusion coefficient of porous anodic alumina structures intended for optical sensing. in Optical Sensors 2011; and Photonic Crystal Fibers V. 2011. SPIE.

Pei, S., et al. Forcesight: Non-contact force sensing with laser speckle imaging. in Proceedings of the 35th Annual ACM Symposium on User Interface Software and Technology. (2022).

Goodman, J.W. Speckle Phenomena in Optics: Theory and Applications. (Roberts and Company Publishers, 2007).

Briers, J.D. Time-varying laser speckle for measuring motion and flow. in Saratov Fall Meeting 2000: Coherent Optics of Ordered and Random Media. 2001. SPIE.

Zdunek, A. et al. The biospeckle method for the investigation of agricultural crops: A review. Opt. Lasers Eng. 52, 276–285 (2014).

Article Google Scholar

Briers, J. D. Speckle fluctuations and biomedical optics: Implications and applications. Opt. Eng. 32(2), 277–283 (1993).

Article ADS Google Scholar

Chen, M. et al. Laser speckle contrast imaging of blood flow in the deep brain using microendoscopy. Opt. Lett. 43(22), 5627–5630 (2018).

Article ADS PubMed Google Scholar

Mangraviti, A. et al. Intraoperative laser speckle contrast imaging for real-time visualization of cerebral blood flow in cerebrovascular surgery: Results from pre-clinical studies. Sci. Rep. 10(1), 7614 (2020).

Article ADS PubMed PubMed Central Google Scholar

Liu, B. et al. Dynamic light scattering and laser speckle contrast imaging of the brain: Theory of the spatial and temporal statistics of speckle pattern evolution. Biomed. Opt. Express 15(2), 579–593 (2024).

Article PubMed PubMed Central Google Scholar

Kaneko, T. et al. Noninvasive assessment of bowel blood perfusion using intraoperative laser speckle flowgraphy. Langenbecks Arch. Surg. 405, 817–826 (2020).

Article PubMed Google Scholar

Skinner, G.C., et al. Clinical utility of laser speckle contrast imaging and real-time quantification of bowel perfusion in minimally invasive left-sided colorectal resections. Diseases Colon Rectum. 10.1097 (2024).

Mirdell, R. et al. Microvascular blood flow in scalds in children and its relation to duration of wound healing: A study using laser speckle contrast imaging. Burns 42(3), 648–654 (2016).

Article PubMed Google Scholar

Lindahl, F., Tesselaar, E. & Sjöberg, F. Assessing paediatric scald injuries using laser speckle contrast imaging. Burns 39(4), 662–666 (2013).

Article PubMed Google Scholar

Li, H. et al. Non-invasive medical imaging technology for the diagnosis of burn depth. Int. Wound J. 21(1), e14681 (2024).

Article PubMed PubMed Central Google Scholar

Mennes, O. A. et al. Semi-automatic tracking of laser speckle contrast images of microcirculation in diabetic foot ulcers. Diagnostics 10(12), 1054 (2020).

Article PubMed PubMed Central Google Scholar

Friedrich, S. et al. Disturbed microcirculation in the hands of patients with systemic sclerosis detected by fluorescence optical imaging: A pilot study. Arthritis Res. Therapy 19(1), 1–13 (2017).

Article Google Scholar

Heeman, W. et al. Application of laser speckle contrast imaging in laparoscopic surgery. Biomed. Opt. Express 10(4), 2010–2019 (2019).

Article PubMed PubMed Central Google Scholar

Potapova, E. V. et al. Laser speckle contrast imaging of blood microcirculation in pancreatic tissues during laparoscopic interventions. Quant. Electron. 50(1), 33 (2020).

Article ADS Google Scholar

Markwalder, L., et al. In vivo laser speckle contrast imaging of microvascular blood perfusion using a chip-on-tip camera. Iscience. 27(3) (2024).

Stewart, C. et al. A comparison of two laser-based methods for determination of burn scar perfusion: Laser Doppler versus laser speckle imaging. Burns 31(6), 744–752 (2005).

Article MathSciNet PubMed Google Scholar

Konovalov, A. et al. Laser speckle contrast imaging in neurosurgery: A systematic review. World Neurosurg. 171, 35–40 (2023).

Article PubMed Google Scholar

Hecht, N. et al. Intraoperative monitoring of cerebral blood flow by laser speckle contrast analysis. Neurosurg. Focus 27(4), E11 (2009).

Article PubMed Google Scholar

Ruaro, B. et al. Laser speckle contrast analysis: A new method to evaluate peripheral blood perfusion in systemic sclerosis patients. Ann. Rheumatic Diseases 73(6), 1181–1185 (2014).

Article Google Scholar

Bi, R., et al. A portable optical pulsatile flowmetry demonstrates strong clinical relevance for diabetic foot perfusion assessment. APL Bioeng. 8(1) (2024).

Eriksson, S., et al. Laser speckle contrast imaging for intraoperative assessment of liver microcirculation: A clinical pilot study. Med. Dev. Evid. Res. 257–261 (2014).

Tamaki, Y. et al. Noncontact, two-dimensional measurement of retinal microcirculation using laser speckle phenomenon. Investig. Ophthalmol. Visual Sci. 35(11), 3825–3834 (1994).

Google Scholar

Minakaran, N. et al. Optical coherence tomography (OCT) in neuro-ophthalmology. Eye 35(1), 17–32 (2021).

Article PubMed Google Scholar

Luft, N. et al. Measurements of retinal perfusion using laser speckle flowgraphy and doppler optical coherence tomography. Investig. Ophthalmol. Visual Sci. 57(13), 5417–5425 (2016).

Article Google Scholar

Elliott, A. D. Confocal microscopy: Principles and modern practices. Curr. Protocols Cytometry 92(1), e68 (2020).

Article Google Scholar

Sdobnov, A. et al. Advances in dynamic light scattering imaging of blood flow. Laser Photon. Rev. 18(2), 2300494 (2024).

Article ADS Google Scholar

Secomb, T. W. Blood flow in the microcirculation. Annu. Rev. Fluid Mech. 49, 443–461 (2017).

Article ADS MathSciNet Google Scholar

Guven, G., Hilty, M. P. & Ince, C. Microcirculation: Physiology, pathophysiology, and clinical application. Blood Purif. 49(1–2), 143–150 (2020).

Article PubMed Google Scholar

González Olmos, A. et al. Optimizing the precision of laser speckle contrast imaging. Sci. Rep. 13(1), 17970 (2023).

Article ADS PubMed PubMed Central Google Scholar

Liu, C. et al. Choosing a model for laser speckle contrast imaging. Biomed. Opt. Express 12(6), 3571–3583 (2021).

Article ADS PubMed PubMed Central Google Scholar

Lopez-Tiro, F., et al. Effect of the exposure time in laser speckle imaging for improving blood vessels localization: A wavelet approach. in 2020 IEEE International Instrumentation and Measurement Technology Conference (I2MTC). 2020. IEEE.

Singh, M.S. Optical polarization technique—For enhancement of image quality—In speckle contrast-based perfusion imaging: A characterization study. AIP Adv. 9(7) (2019).

Badieyan, S. et al. Detection and discrimination of bacterial colonies with mueller matrix imaging. Sci. Rep. 8(1), 10815 (2018).

Article ADS PubMed PubMed Central Google Scholar

Amiri, S. et al. Viral infected cells reveal distinct polarization behavior; a polarimetric microscopy analysis on HSV infected vero and hela cells. J. Quant. Spectrosc. Radiative Transfer 262, 107484 (2021).

Article Google Scholar

Badieyan, S. et al. Mueller matrix imaging of prostate bulk tissues; polarization parameters as a discriminating benchmark. Photodiagnosis Photodyn. Ther. 26, 90–96 (2019).

Article PubMed Google Scholar

Young, A.M. Investigation of Laser Speckle Contrast Imaging's Sensitivity to Flow. (Miami University, 2018).

Akther, S., Mikkelsen, M. B. & Postnov, D. D. Choosing a polarisation configuration for dynamic light scattering and laser speckle contrast imaging. Biomed. Opt. Express 15(1), 336–345 (2024).

Article PubMed Google Scholar

Héran, D. et al. Combining light polarization and speckle measurements with multivariate analysis to predict bulk optical properties of turbid media. Appl. Opt. 58(30), 8247–8256 (2019).

Article ADS PubMed Google Scholar

Tchvialeva, L. et al. Polarization speckle imaging as a potential technique for in vivo skin cancer detection. J. Biomed. Opt. 18(6), 061211–061211 (2013).

Article ADS PubMed Google Scholar

Bi, R. et al. Fast pulsatile blood flow measurement in deep tissue through a multimode detection fiber. J. Biomed. Optics 25(5), 055003–055003 (2020).

Article ADS Google Scholar

Mazhar, A. et al. Laser speckle imaging in the spatial frequency domain. Biomed. Optics Express 2(6), 1553–1563 (2011).

Article ADS Google Scholar

Ayers, F., et al. Fabrication and characterization of silicone-based tissue phantoms with tunable optical properties in the visible and near infrared domain. in Design and Performance Validation of Phantoms Used in Conjunction with Optical Measurements of Tissue. (SPIE, 2008).

Park, H., K. Seo, & K.B. Crozier. Adding colors to polydimethylsiloxane by embedding vertical silicon nanowires. Appl. Phys. Lett. 101(19) (2012).

Kono, T. & Yamada, J. In vivo measurement of optical properties of human skin for 450–800 nm and 950–1600 nm wavelengths. Int. J. Thermophys. 40, 1–14 (2019).

ADS Google Scholar

Bosschaart, N. et al. A literature review and novel theoretical approach on the optical properties of whole blood. Lasers Med. Sci. 29, 453–479 (2014).

Article PubMed Google Scholar

Van Staveren, H. J. et al. Light scattering in lntralipid-10% in the wavelength range of 400–1100 nm. Appl. Opt. 30(31), 4507–4514 (1991).

Article ADS PubMed Google Scholar

Flock, S. T. et al. Optical properties of Intralipid: A phantom medium for light propagation studies. Lasers Surg. Med. 12(5), 510–519 (1992).

Article PubMed Google Scholar

Ivanov, K., Kalinina, M. & Levkovich, Y. I. Blood flow velocity in capillaries of brain and muscles and its physiological significance. Microvasc. Res. 22(2), 143–155 (1981).

Article PubMed Google Scholar

Neerken, S. et al. Characterization of age-related effects in human skin: A comparative study that applies confocal laser scanning microscopy and optical coherence tomography. J. Biomed. Opt. 9(2), 274–281 (2004).

Article ADS PubMed Google Scholar

Fredriksson, I. in Skin Capillary Ensemble Visualisation and Computation. (Master’s thesis, Linköpings Universitet, 2004).

Yang, Z., Zhao, M. & Lu, P. How to improve the signal-to-noise ratio for circular polarizers consisting of helical metamaterials?. Opt. Express 19(5), 4255–4260 (2011).

Article ADS PubMed Google Scholar

Download references

This paper has been extracted from the thesis written by Mrs. Nasrin Amini in School of Medicine, Shahid Beheshti University of Medical Sciences (Registration No. 31800).

Department of Medical Physics & Biomedical Engineering, School of Medicine, Shahid Beheshti University of Medical Sciences, Tehran, Iran

Nasrin Amini, Ali Esteki, Mohsen Ahmadi & Pezhman Sasanpour

School of Nanoscience, Institute for Research in Fundamental Sciences (IPM), P.O. Box 19395-5531, Tehran, Iran

Pezhman Sasanpour

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

P.S. proposed the original idea and with the contribution of N.A. the experimental setup and the phantom was implemented. P.S., A.E. and M.A. provided advice in analyzing the results and discussion. All authors contributed to writing and editing of the manuscript.

Correspondence to Pezhman Sasanpour.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Amini, N., Esteki, A., Ahmadi, M. et al. Impact of light polarization on laser speckle contrast imaging with a custom phantom for microvascular flow. Sci Rep 14, 26652 (2024). https://doi.org/10.1038/s41598-024-73757-2

Download citation

Received: 28 July 2024

Accepted: 20 September 2024

Published: 04 November 2024

DOI: https://doi.org/10.1038/s41598-024-73757-2

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative