Method for segmentation of IVUS image sequences

A method for performing segmentation of an interior vessel within the body of a patient to detect the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel wherein images in said sequence of IVUS images are spatially associated with one another;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• scaling down said sequence of IVUS images, at least before said procedure of extracting an outer boundary;
• and
• scaling up said extracted outer boundary.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence of IVUS images to counter various distortions;
• performing statistical analysis and spatial integration on said sequence of IVUS images to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• scaling down said sequence of IVUS images, at least before said procedure of performing statistical analysis and spatial integration;
• and
• scaling up said extracted outer boundary and said extracted inner boundary.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• wherein said procedure of detecting uninformative regions in said sequence of IVUS images, comprises the procedures of;
• detecting the location of said guide wire in said sequence of IVUS images;
• identifying regions of calcified plaque in said sequence of IVUS images;
• and
• integrating said guide wire data with said regions of calcified plaque data.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• wherein said sequence of IVUS images is divided into batches;
• detecting stents in each of said batches;
• determining borders of said stents in said sequence of IVUS images;
• and
• wherein said procedure of detecting stents in each of said batches, comprises the procedures of:
• applying a morphological closing operation to every image in each of said batches;
• partitioning each image into sub-regions and locating pixels with outlying intensities in each of said sub-regions;
• assigning tentative stent pixels based on pixel intensity characteristics, resulting in a binary image of true and false pixels;
• applying a morphological dilation operation to said resulting binary image;
• eliminating falsely identified stent pixels from said binary image;
• determining correlation value between consecutive images in each of said batches;
• and
• determining if a stent exists in a batch for each of said batches.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• wherein said sequence of IVUS images is divided into batches;
• estimating roll offsets between each pair of consecutive images in each of said batches;
• and
• accumulating said roll offsets in said sequence of IVUS images.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• where said procedure of performing statistical analysis and spatial integration involves calculating statistical properties relating to the intensities of a pixel over a series of said images;
• wherein said statistical properties includes properties selected from the list consisting of;
• average intensity;
• total variation;
• number of extrema;
• variance;
• standard deviation;
• skewness;
• and
• kurtosis.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions:
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• where said procedure of performing statistical analysis and spatial integration involves calculating statistical properties relating to the intensities of a pixel over a series of said images;
• wherein two histograms are constructed based on said manually traced images, one of said histograms indicating the number of pixels in said manually traced images that have a given value of said statistical properties in the inner layer of said interior vessel, the other of said histograms indicating the number of pixels in said manually traced images that have a given value of said statistical properties in the outer layer of said interior vessel.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• wherein said sequence of IVUS images is divided into batches;
• identifying bifurcation sectors in each of said batches;
• computing bifurcation regions in said sequence of IVUS images;
• and
• wherein said procedure of identifying bifurcation sectors in each of said batches is based on blood speckles detected linearly along a curve at a constant offset beyond said extracted outer boundary.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• wherein said sequence of IVUS images is divided into batches;
• identifying bifurcation sectors in each of said batches;
• computing bifurcation regions in said sequence of IVUS images;
• and
• wherein said procedure of identifying bifurcation sectors in each of said batches is performed excluding said detected uninformative regions arising from an acoustic shadow cast by guide wire and by calcified plaque within said interior vessel.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• wherein said sequence of IVUS images is divided into batches;
• identifying bifurcation sectors in each of said batches;
• computing bifurcation regions in said sequence of IVUS images;
• and
• wherein said procedure of computing said bifurcation regions in said sequence of IVUS images is assisted by bifurcation regions identified via at least one medical imaging modality.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• manually editing output of said segmentation in selected images by a user;
• and
• wherein said procedure of manually editing output of said segmentation involves integrating manual editing on the detected boundaries in a given image by said user, while adjusting the boundaries in adjacent images to seamlessly integrate said manual editing.

A method for performing segmentation of an interior vessel within the body of a patient to detect the inner boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• manually editing output of said segmentation in selected images by a user;
• and
• wherein said procedure of manually editing output of said segmentation serves as feedback for an additional execution of said segmentation method, by forcing the output at desired locations to be identical to anchor points specified by said user in said manual edited images.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• displaying output of said segmentation with respect to level of confidence measures;
• and
• wherein said level of confidence measures is represented as a confidence interval about the computed value of said output of said segmentation, such that the actual value of said output of said segmentation lies within said interval with a predetermined probability.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming:
• displaying output of said segmentation with respect to level of confidence measures;
• and
• wherein said level of confidence measures is represented as the probability that the actual value of said output of said segmentation lies within some predetermined interval about the computed value of said output of said segmentation.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• detecting uninformative regions in said sequence of IVUS images, said uninformative regions arising from an acoustic shadow cast by guide wire and by calcified plaque within said interior vessel;
• wherein said procedure of detecting uninformative regions in said sequence of IVUS images, comprises the procedures of:
• detecting the location of said guide wire in said sequence of IVUS images;
• identifying regions of calcified plaque in said sequence of IVUS images;
• and
• integrating said guide wire data with said regions of calcified plaque data.

A method for performing segmentation of an interior vessel within the body of a patient to detect the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel wherein images in said sequence of IVUS images are spatially associated with one another,
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• detecting uninformative regions in said sequence of IVUS images, said uninformative regions arising from an acoustic shadow cast by guide wire and by calcified plaque within said interior vessel;
• and
• wherein said procedure of detecting uninformative regions in said sequence of IVUS images, comprises the procedures of:
• detecting the location of said guide wire in said sequence of IVUS images;
• identifying regions of calcified plaque in said sequence of IVUS images;
• and
• integrating said guide wire data with said regions of calcified plaque data.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• wherein said sequence of IVUS images is divided into batches;
• detecting stents in each of said batches;
• determining borders of said stents in said sequence of IVUS images;
• wherein said procedure of detecting stents in each of said batches, comprises the procedures of:
• applying a morphological closing operation to every image in each of said batches;
• partitioning each image into sub-regions and locating pixels with outlying intensities in each of said sub-regions;
• assigning tentative stent pixels based on pixel intensity characteristics, resulting in a binary image of true and false pixels;
• applying a morphological dilation operation to said resulting binary image;
• eliminating falsely identified stent pixels from said binary image;
• determining correlation value between consecutive images in each of said batches;
• and
• determining if a stent exists in a batch for each of said batches.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence to counter various distortions;
• performing statistical analysis and spatial integration on said sequence to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• wherein said sequence of IVUS images is divided into batches;
• detecting stents in each of said batches;
• determining borders of said stents in said sequence of IVUS images;
• and
• estimating roll offsets between each pair of consecutive images in each of said batches;
• and
• accumulating said roll offsets in said sequence of IVUS images.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence of IVUS images to counter various distortions;
• performing statistical analysis and spatial integration on said sequence of IVUS images to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• wherein said procedure of performing statistical analysis and spatial integration involves calculating statistical properties relating to the intensities of a pixel over a series of said images;
• wherein said statistical properties includes properties selected from the list consisting of;
• average intensity;
• total variation;
• number of extrema;
• variance;
• standard deviation;
• skewness;
• and
• kurtosis.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence of IVUS images to counter various distortions;
• performing statistical analysis and spatial integration on said sequence of IVUS images to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• wherein said procedure of performing statistical analysis and spatial integration involves calculating statistical properties relating to the intensities of a pixel over a series of said images;
• and
• wherein two histograms are constructed based on said manually traced images, one of said histograms indicating the number of pixels in said manually traced images that have a given value of said statistical properties in the inner layer of said interior vessel, the other of said histograms indicating the number of pixels in said manually traced images that have a given value of said statistical properties in the outer layer of said interior vessel.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence of IVUS images to counter various distortions;
• performing statistical analysis and spatial integration on said sequence of IVUS images to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• wherein said sequence of IVUS images is divided into batches;
• and
• wherein said procedure of identifying bifurcation sectors in each of said batches is based on blood speckles detected linearly along a curve at a constant offset beyond said extracted outer boundary.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence of IVUS images to counter various distortions;
• performing statistical analysis and spatial integration on said sequence of IVUS images to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• wherein said sequence of IVUS images is divided into batches;
• and
• wherein said procedure of identifying bifurcation sectors in each of said batches is performed excluding said detected uninformative regions arising from an acoustic shadow cast by guide wire and by calcified plaque within said interior vessel.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence of IVUS images to counter various distortions;
• performing statistical analysis and spatial integration on said sequence of IVUS images to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• wherein said sequence of IVUS images is divided into batches;
• and
• wherein said procedure of computing said bifurcation regions in said sequence of IVUS images is assisted by bifurcation regions identified via at least one medical imaging modality.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel wherein images in said sequence of IVUS images are spatially associated with one another;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• manually editing output of said segmentation in selected images by a user;
• and
• wherein said procedure of manually editing output of said segmentation involves integrating manual editing on the detected boundaries in a given image by said user, while adjusting the boundaries in adjacent images to seamlessly integrate said manual editing.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence of IVUS images to counter various distortions;
• performing statistical analysis and spatial integration on said sequence of IVUS images to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• manually editing output of said segmentation in selected images by a user;
• and
• wherein said procedure of manually editing output of said segmentation involves integrating manual editing on the detected boundaries in a given image by said user, while adjusting the boundaries in adjacent images to seamlessly integrate said manual editing.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel wherein images in said sequence of IVUS images are spatially associated with one another;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• manually editing output of said segmentation in selected images by a user;
• and
• wherein said procedure of manually editing output of said segmentation serves as feedback for an additional execution of said segmentation method, by forcing the output at desired locations to be identical to anchor points specified by said user in said manual edited images.

A method for performing segmentation of an interior vessel within the body of a patient to detect both the inner boundary and the outer boundary of said vessel, the method comprising the procedures of:

• obtaining a sequence of intravascular ultrasound (IVUS) images of an interior vessel;
• tracking images in said sequence of IVUS images to counter various distortions;
• performing statistical analysis and spatial integration on said sequence of IVUS images to obtain a classification of blood and tissue regions;
• extracting inner boundary of said interior vessel based on said classification of blood and tissue regions;
• extracting an outer boundary of said interior vessel utilizing dynamic programming;
• manually editing output of said segmentation in selected images by a user;
• and
• wherein said procedure of manually editing output of said segmentation serves as feedback for an additional execution of said segmentation method, by forcing the output at desired locations to be identical to anchor points specified by said user in said manual edited images.