# US 11,439,308 B2 · Methods and systems for thermal monitoring of tissue with an ultrasound imaging system > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,439,308 B2 - Document ID: US11439308B2 - Status: Granted - Kind: B2 - Application number: 16/927,685 - Filed: 2020-07-13 - Published: 2022-01-13 - Granted: 2022-09-13 - Earliest claimed: 2020-07-13 - Assignees: GE Precision Healthcare LLC - Applicants: GE Precision Healthcare LLC - Inventors: Jens Ziegle; Alfredo Guillermo Illanes Manríquez; Elmer Jeto Gomes Ataide; Michael Friebe - Web page: https://patentgrep.com/patent/US11439308B2 ## CPC - A61B 5/015: {By temperature mapping of body part} - G06T 7/0012: {Biomedical image inspection} - A61B 8/08: Clinical applications (A61B8/02, A61B8/04, A61B8/06 take precedence) - A61B 8/5223: {for extracting a diagnostic or physiological parameter from medical diagnostic data (for algorithms to analyse biomedical images G06T7/0012)} - G06T 7/0014: {using an image reference approach} - G06V 10/25: Determination of region of interest [ROI] or a volume of interest [VOI] - G06V 10/431: {Frequency domain transformation; Autocorrelation} - G06V 10/52: Scale-space analysis, e.g. wavelet analysis (multi-scale boundary representations G06V10/42) - G06F 2218/20: by applying autoregressive analysis - G06T 2207/10132: Ultrasound image - G06T 2207/20064: Wavelet transform [DWT] - G06T 2207/20104: Interactive definition of region of interest [ROI] - G06T 2207/30096: Tumor; Lesion - G06V 2201/031: of internal organs - A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D) - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract Various methods and systems are provided for thermal monitoring of tissue with an ultrasound imaging system. In one embodiment, a method comprises acquiring, via an ultrasound probe, an ultrasound image, selecting at least one region of interest in the ultrasound image, extracting features from the at least one region of interest, classifying a thermal state of tissue in the at least one region of interest based on the features, and outputting, via a display device, the thermal state of the tissue. In this way, ultrasound imaging may be used for real-time thermal monitoring of tissue during an ablation procedure, thereby improving the accuracy and efficacy of the ablation procedure. ## Claims ### Claim 1 A method, comprising: acquiring, via an ultrasound probe, an ultrasound image; selecting at least one region of interest in the ultrasound image; extracting features from the at least one region of interest, wherein extracting the features from the at least one region of interest comprises: converting the at least one region of interest to at least one signal; decomposing the at least one signal into multiple frequency bands; generating an estimate for each decomposed signal with an autoregressive model; calculating a correlation coefficient for each estimate with an initial estimate in a same frequency band; determining a dissimilarity metric based on the correlation coefficient for each estimate; and extracting the features from the dissimilarity metric; classifying a thermal state of tissue in the at least one region of interest based on the features; and outputting, via a display device, the thermal state of the tissue. ### Claim 2 (depends on 1) The method of claim 1, wherein the features comprise one or more of a median correlation coefficient of the dissimilarity metric, a mean correlation coefficient of the dissimilarity metric, a slope of a linear regression of the dissimilarity metric, and an intercept of a linear regression of the dissimilarity metric. ### Claim 3 (depends on 1) The method of claim 1, wherein converting the at least one region of interest to the at least one signal comprises, for a region of interest of the at least one region of interest, sampling pixel values of the region of interest while traversing pixels of the region of interest along a plurality of pixel trajectories to generate a plurality of signals for the region of interest, each signal of the plurality of signals corresponding to a different pixel trajectory of the plurality of pixel trajectories. ### Claim 4 (depends on 1) The method of claim 1, wherein decomposing the at least one signal into multiple frequency bands comprises decomposing, using a discrete wavelet transform, the at least one signal into the multiple frequency bands. ### Claim 5 (depends on 1) The method of claim 1, wherein generating the estimate for each decomposed signal with the autoregressive model comprises performing autoregressive modeling of each decomposed signal to obtain an autoregressive parameter for each decomposed signal, and calculating a power spectral density for each decomposed signal based on the respective autoregressive parameter, the estimate comprising the power spectral density. ### Claim 6 (depends on 1) The method of claim 1, further comprising calculating the correlation coefficient for each estimate with the initial estimate in the same frequency band and a same image-to-signal conversion type. ### Claim 7 (depends on 1) The method of claim 1, wherein determining the dissimilarity metric based on the correlation coefficient for each estimate comprises sorting the correlation coefficients for the estimates into a vector, and wherein extracting the features from the at least one region of interest comprises calculating the features from the vector. ### Claim 8 (depends on 1) The method of claim 1, wherein classifying the thermal state of the tissue in the at least one region of interest based on the features comprises inputting the features into one or more machine learning algorithms, and receiving, from the one or more machine learning algorithms, a classification of the thermal state of the tissue. ### Claim 9 (depends on 8) The method of claim 8, wherein the classification comprises one or more of a classification of temperature or a classification of tissue damage reversibility. ### Claim 10 (depends on 1) The method of claim 1, wherein selecting the at least one region of interest in the ultrasound image comprises selecting a first region of interest at an ablation target in the ultrasound image, and selecting a plurality of additional regions of interest in the ultrasound image positioned away from the ablation target. ### Claim 11 A method, comprising: acquiring, with an ultrasound probe, an ultrasound image; converting a region of interest in the ultrasound image to a plurality of signals; decomposing each signal of the plurality of signals into a plurality of frequency bands; generating an estimate for each decomposed signal with an autoregressive model for each frequency band of the plurality of frequency bands; measuring a correlation of the estimate for each decomposed signal with a corresponding estimate obtained from a previously-acquired ultrasound image; calculating a plurality of features based on the correlation for each decomposed signal; classifying, with a machine learning algorithm based on the plurality of features, a thermal state of tissue in the region of interest; and outputting, via a display device, the thermal state of the tissue. ### Claim 12 (depends on 11) The method of claim 11, wherein converting the region of interest to the plurality of signals comprises sampling pixel values of the region of interest with a plurality of pixel-traversing trajectories to generate the plurality of signals. ### Claim 13 (depends on 11) The method of claim 11, wherein decomposing each signal of the plurality of signals into the plurality of frequency bands comprises applying a discrete wavelet transform to each signal to decompose each signal into a low-frequency band signal, a middle-frequency band signal, a high-frequency band signal, and a total frequency band signal. ### Claim 14 (depends on 11) The method of claim 11, wherein the plurality of features includes one or more of a median correlation coefficient, a mean correlation coefficient, a slope of a linear regression of a vector including the correlation of each decomposed signal, and an intercept of the linear regression of the vector. ### Claim 15 (depends on 11) The method of claim 11, wherein classifying, with the machine learning algorithm based on the plurality of features, the thermal state of the tissue in the region of interest comprises classifying a temperature of the tissue relative to a critical ablation temperature. ### Claim 16 A system, comprising: an ultrasound probe; a display device; and a processor configured with executable instructions in non-transitory memory that when executed cause the processor to: acquire, via the ultrasound probe, an ultrasound image; select at least one region of interest in the ultrasound image; extract features from the at least one region of interest; convert the at least one region of interest to a plurality of signals; decompose each signal of the plurality of signals into a plurality of frequency bands; generate an estimate for each decomposed signal with an autoregressive model; calculate a correlation coefficient for each estimate with an initial estimate obtained from a previously-acquired ultrasound image in a same frequency band; determine a dissimilarity metric based on the correlation coefficient for each estimate; extract the features from the dissimilarity metric; classify a thermal state of tissue in the at least one region of interest based on the features; and output, via the display device, the thermal state of the tissue. ### Claim 17 (depends on 16) The system of claim 16, wherein the features comprise one or more of a mean correlation coefficient, a median correlation coefficient, a slope of a linear regression of the dissimilarity metric, and an intercept of the linear regression of the dissimilarity metric. ### Claim 18 (depends on 16) The system of claim 16, wherein the processor is further configured with executable instructions in the non-transitory memory that when executed cause the processor to: classify, with a machine learning algorithm based on the plurality of features, the thermal state of the tissue in the region of interest, the thermal state comprising a temperature of the tissue relative to a critical ablation temperature.