Journal: bioRxiv
Article Title: Laser Speckle Particle Sizer (SPARSE) Informs the Size Distribution of Tissue Granularities
doi: 10.1101/2024.09.26.615148
Figure Lengend Snippet: a Optical setup of SPARSE. A coherent laser beam is linearly polarized and focused on the luminal surface of the sample. Parallel and perpendicular linearly polarized speckle patterns emanating from the sample are collected by a pair of high-speed CMOS cameras that detect the co-polarized and cross-polarized channels. b Co-polarized and c Cross-polarized speckle frame series simultaneously collected from a sample. d Intensity envelope of the co-polarized speckle frame series, Î ∥ . The insets display the top view and the contour of normalized Î ∥ at 30% of the peak intensity. The long axis of the contour is aligned with the polarization axis. To identify the inner and outer circles, the centroid of the contour is determined from the coordinates of the individual points on the contour. The inner and outer circles are then drawn to have their centers at the centroid and pass through the points on the contour that are closest and farthest from the centroid, respectively. Subsequently, Î ∥ is radially averaged in the annular region between the inner and outer circles of the contour, yielding an angular profile, from which is calculated. e Intensity envelope of the cross-polarized speckle frame series, Î ⊥ . The insets display the top view and the contour of normalized Î ⊥ at 30% of the peak intensity, from which is calculated, where A and P are the area and perimeter of the contour. In addition, X 4 =√ (A/π) represents the average radius of the contour. f Differential decorrelation rates of co- and cross-polarized speckle time series, defined as is obtained from the intensity autocorrelation, g 2 (t) , of the acquired speckle time series (inset) at a time instant where the gap is the widest. As seen later, a Monte-Carlo ray tracing algorithm is used to create a synthetic library of [X 1 , X 2 , X 3 , X 4 ] metrics for numerous possible combinations of particle size and optical properties. Cluster analysis of the synthetic [X 1 , X 2 , X 3 , X 4 ] followed by multiple regression analysis provides the particle size estimation equation for each cluster. G To estimate the particle size from the experimentally evaluated [ X 1 , X 2 , X 3 , X 4 ] metrics, first the Euclidean distance from the cluster centers is evaluated to identify the cluster associated with the experimentally evaluated [X 1 ,X 2 ,X 3 , X 4 ]. Subsequently, the particle size estimation equation corresponding to this nearest cluster center is used to return the average particle size, a , using the [X 1 , X 2 , X 3 , X 4 ] metrics.
Article Snippet: The backscattered light rays are collected and split by a second beam splitter in two different light paths and detected by two different high-speed CMOS cameras (Basler, ACa 2000-340 km, Germany).
Techniques:
Basler is a verified supplier 
Basler manufactures this product 
![a Optical setup of SPARSE. A coherent laser beam is linearly polarized and focused on the luminal surface of the sample. Parallel and perpendicular linearly polarized speckle patterns emanating from the sample are collected by a pair of high-speed <t>CMOS</t> cameras that detect the co-polarized and cross-polarized channels. b Co-polarized and c Cross-polarized speckle frame series simultaneously collected from a sample. d Intensity envelope of the co-polarized speckle frame series, Î ∥ . The insets display the top view and the contour of normalized Î ∥ at 30% of the peak intensity. The long axis of the contour is aligned with the polarization axis. To identify the inner and outer circles, the centroid of the contour is determined from the coordinates of the individual points on the contour. The inner and outer circles are then drawn to have their centers at the centroid and pass through the points on the contour that are closest and farthest from the centroid, respectively. Subsequently, Î ∥ is radially averaged in the annular region between the inner and outer circles of the contour, yielding an angular profile, from which is calculated. e Intensity envelope of the cross-polarized speckle frame series, Î ⊥ . The insets display the top view and the contour of normalized Î ⊥ at 30% of the peak intensity, from which is calculated, where A and P are the area and perimeter of the contour. In addition, X 4 =√ (A/π) represents the average radius of the contour. f Differential decorrelation rates of co- and cross-polarized speckle time series, defined as is obtained from the intensity autocorrelation, g 2 (t) , of the acquired speckle time series (inset) at a time instant where the gap is the widest. As seen later, a Monte-Carlo ray tracing algorithm is used to create a synthetic library of [X 1 , X 2 , X 3 , X 4 ] metrics for numerous possible combinations of particle size and optical properties. Cluster analysis of the synthetic [X 1 , X 2 , X 3 , X 4 ] followed by multiple regression analysis provides the particle size estimation equation for each cluster. G To estimate the particle size from the experimentally evaluated [ X 1 , X 2 , X 3 , X 4 ] metrics, first the Euclidean distance from the cluster centers is evaluated to identify the cluster associated with the experimentally evaluated [X 1 ,X 2 ,X 3 , X 4 ]. Subsequently, the particle size estimation equation corresponding to this nearest cluster center is used to return the average particle size, a , using the [X 1 , X 2 , X 3 , X 4 ] metrics.](https://bio-rxiv-images-cdn.bioz.com/dois_ending_with_48/10__1101_slash_2024__09__26__615148/10__1101_slash_2024__09__26__615148___F1.large.jpg)