Granted Patent B2

US 12,590,839 · App. 18/022,793 · Granted Mar 31, 2026

Inventor: Wei Lin (Stony Brook, NY)

Assignee: The Research Foundation for The State University of New York

G01J3/457A61B5/0075A61B5/0261

Abstract

Claims (66)

1 . An embedded diffuse correlation spectroscopy (DCS) analyzer to measure blood perfusion in a biological tissue, the analyzer comprising:

a memory, the memory storing autocorrelation values, model parameters used in solving an autocorrelation equation of scattered light, fitting parameters, and simulated correlation values obtained from a theoretical DCS model, the scattered light being detected by a photodetector from the blood perfusion in the biological tissue;

a hardware mean square error (MSE) module, the MSE module configured to compute a mean square error between theoretical autocorrelation values computed from the model parameters and measured autocorrelation values;

a hardware sorting module, the sorting module configured to sort a latest three MSE values obtained from the MSE module and to generate indexes of largest, medium, and smallest MSE values, the indexes being stored in the memory as a subset of the fitting parameters;

a hardware convergence checking module configured to determine whether convergence is reached in solving the autocorrelation equation;

a hardware search module configured to calculate αD B and β values at reflection, extension, contraction, and shrink locations, the αD B and β values being calculated in parallel;

a hardware comparison module configured to compare two latest MSE values and to facilitate finding new αD B and β values to replace those values associated with a largest MSE;

a state controller coupled with the memory and modules comprising the MSE module, the sorting module, the convergence checking module, the search module, and the comparison module to control an operation thereof, and

an output buffer configured to present, as an output of the DCS analyzer, a fitted solution of the autocorrelation equation, facilitating real-time quantification of the blood perfusion in the biological tissue.

2 . The DCS analyzer according to claim 1 , wherein the MSE module is configured to compute the MSE between a theoretical autocorrelation and a measured autocorrelation using a pipeline structure for parallel computing and high throughput.

3 . The DCS analyzer according to claim 1 , further comprising a first buffer coupled between the state controller and the MSE module, the first buffer being configured to send modified correlation values and corresponding modified delay time values (τ n ), generated by the state controller, to the MSE module.

4 . The DCS analyzer according to claim 3 , further comprising a second buffer coupled between the memory and the state controller, the second buffer being configured to output αD B and β values of a smallest MSE to the state controller when prescribed convergence criteria are met.

5 . The DCS analyzer according to claim 4 , wherein at least one of the first buffer and second buffer is a first-in-first-out buffer.

6 . The DCS analyzer according to claim 1 , wherein data paths between the memory and at least one of the modules are bidirectional.

7 . The DCS analyzer according to claim 1 , wherein a start operation of at least one of the modules is controlled by one or more control signals asserted by the state controller.

8 . The DCS analyzer according to claim 1 , wherein the modules are configured to send end of operation and status signals to the state controller.

9 . The DCS analyzer according to claim 1 , wherein the state controller is configured to receive at least one command to load data from an input buffer and to store the data in the memory for DCS analysis.

10 . The DCS analyzer according to claim 1 , further comprising an input buffer, wherein the state controller is configured to receive at least one command from the input buffer to start DCS analysis.

11 . The DCS analyzer according to claim 1 , wherein the state controller is configured to switch operation states based on feedback from a status signal.

12 . The DCS analyzer according to claim 1 , wherein, based on feedback from the comparison module, the state controller is configured to:

select a smallest MSE value (MSE X ) as a first input and MSE at a reflection location (MSR R ) as a second input,

determine whether MSE R is less than MSE X ;

select MSE at an extension location (MSE E ) as an input when MSE R is less than the current MSE value and determine whether MSE E is less than a current MSE R ;

replace αD B and β values related to a largest MSE value (MSE Z ) with αD B and, β values related to MSE E when MSE E is less than MSE R ;

replace αD B and β values related to MSE Z with αD B and β values related to MSE R when MSE E is not less than MSE R ,

select MSE at a contraction location (MSE C ) as an input when MSE R is not less than MSE X and determine whether MSE C is less than a current MSE R ;

replace αD B and β values related to MSE Z with αD B and β values related to MSE C when MSE C is less than MSE R ;

select MSE Z as an input when MSE C is not less than MSE R and determine whether MSE Z is less than a MSE C ;

when MSE Z is less than MSE C , replace αD B and β values related to a medium MSE value (MSE Y ) with αD B and β values related to MSE in a first shrink location (MSE SY ), and replace αD B and β values related to MSE Z with αD R and β values related to MSE in a second shrink location (MSE SZ ); and

when MSE Z is not less than MSE C , replace αD B and β values related to MSE Z with αF B and β values related to MSE C .

13 . The DCS analyzer according to claim 1 , wherein the state controller is configured to start operation of at least one of the modules, control signals based on current states, and end of operation and status of at least one of the modules.

14 . The DCS analyzer according to claim 1 , further comprising an output buffer, wherein the state controller is configured to output data comprising at least one of αD B , β, MSE and a number of correlation values, through the output buffer.

15 . The DCS analyzer according to claim 1 , wherein the sorting module is configured to determine an order of three latest input MSE values and to generate corresponding indexes to large, medium, and small MSE values simultaneously.

16 . The DCS analyzer according to claim 1 , wherein the search module is configured to compute MSE αD B and β values at reflection, extension, construction, and shrink locations simultaneously, and to store results of the computation in the memory.

17 . The DCS analyzer according to claim 1 , wherein the convergence module is configured to check whether a relative change of a small MSE is within a prescribed limit, or whether a number of iterations has reached a prescribed limit, and to decide when to end a curve fitting iteration.

18 . The DCS analyzer according to claim 1 , wherein the DCS analyzer is implemented in a field-programmable gate array (FPGA).

19 . An apparatus to perform diffuse correlation spectroscopy (DCS) to measure blood perfusion in a biological tissue, the apparatus comprising:

a hardware DCS correlator; and

a hardware DCS analyzer coupled with the DCS correlator and embedded therewith on an integrated circuit chip, the DCS analyzer comprising:

a memory, the memory storing autocorrelation values, model parameters used in solving an autocorrelation equation of scattered light, fitting parameters, and simulated correlation values obtained from a theoretical DCS model, the scattered light being detected by a photodetector from the blood perfusion in the biological tissue;

a hardware mean square error (MSE) module configured to compute a mean square error between theoretical autocorrelation values computed from the model parameters and measured autocorrelation values;

a hardware sorting module configured to sort a latest three MSE values obtained from the MSE module and to generate indexes of largest, medium, and smallest MSE values, the indexes being stored in the memory as a subset of the fitting parameters;

a hardware convergence checking module configured to determine whether convergence is reached in solving the autocorrelation equation;

a hardware search module configured to calculate αD B and β values at reflection, extension, contraction, and shrink locations, the αD B and β values being calculated in parallel;

a hardware comparison module configured to compare two latest MSE values and facilitate finding new αD B and β values to replace those values associated with a largest MSE;

a state controller coupled with the memory and modules comprising the MSE module, the sorting module, the convergence checking module, the search module, and the comparison module to control an operation thereof; and

an output buffer to present, as an output of the DCS analyzer, a fitted solution of the autocorrelation equation, facilitating real-time quantification of the blood perfusion in the biological tissue.

20 . An embedded diffuse correlation spectroscopy (DCS) analyzer to measure blood perfusion in a biological tissue, the analyzer comprising:

a memory, the memory storing autocorrelation values, model parameters used in solving an autocorrelation equation of scattered light, fitting parameters, and simulated correlation values obtained from a theoretical DCS model, the scattered light being detected by a photodetector from the blood perfusion in the biological tissue;

a hardware mean square error (MSE) module configured to compute a mean square error between theoretical autocorrelation values computed from the model parameters and measured autocorrelation values;

a hardware sorting module configured to sort a latest three MSE values obtained from the MSE module and to generate indexes of largest, medium, and smallest MSE values, the indexes being stored in the memory as a subset of the fitting parameters;

a hardware convergence checking module configured to determine whether convergence is reached in solving the autocorrelation equation;

a hardware search module configured to calculate αD R and β values at reflection, extension, contraction, and shrink locations, the αD B and β values being calculated in parallel;

a hardware comparison module configured to compare two latest MSE values and to facilitate finding new αD B and β values to replace those values associated with a largest MSE;

a state controller coupled with the memory and modules comprising the MSE module, the sorting module, the convergence checking module, the search module, and the comparison module to control an operation thereof; and

an output buffer configured to present, as an output of the DCS analyzer, a fitted solution of the autocorrelation equation, facilitating real-time quantification of the blood perfusion in the biological tissue, wherein based on feedback from the comparison module, the state controller is configured to:

select a smallest MSE value (MSE x ) as a first input and MSE at a reflection location (MSR R ) as a second input;

determine whether MSE R is less than MSE X ;

select MSE at an extension location (MSE E ) as an input when MSE R is less than the current MSE value and determine whether MSE E is less than a current MSE R ;

replace αD B and β values related to a largest MSE value (MSE Z ) with αD B and β values related to MSE E when MSE E is less than MSE R ;

replace αD B and β values related to MSE Z with αD B and β values related to MSE R when MSE E is not less than MSE R ;

select MSE at a contraction location (MSE C ) as an input when MSE R is not less than MSE X and determine whether MSE C is less than a current MSE R ;

replace αD B and, β values related to MSE Z with αD B and, β values related to MSE C when MSE C is less than MSE R ;

select MSE Z as an input when MSE C is not less than MSE R and determine whether MSE Z is less than a MSE C ;

when MSE Z is less than MSE C , replace αD B and β values related to a medium MSE value (MSE Y ) with αD B and β values related to MSE in a first shrink location (MSE SY ), and replace αD B and β values related to MSE Z with αD B and β values related to MSE in a second shrink location (MSE SZ ); and

when MSE Z is not less than MSE C , replace αD B and β values related to MSE Z with αD B and β values related to MSE C .

Continuity (2)

References Cited (77)