A machine learning-based method for sensor analysis and vascular tree segmentation.

Abstract

A method for automatically identifying vascular features is described. In some embodiments, one or more machine learning (ML)-based vascular classifiers are used, whose results are combined with the results of at least one other vascular classifier to generate a final result. Potential advantages of this approach include the ability to combine certain strengths of ML classifiers with segmentation approaches based on more classical ("mathematical formula-based") methods. These strengths include, in particular, the identification of anatomically identified targets that are intermixed within images that show visually similar but anatomically distinct targets.

Claims

1. A method for identifying vascular portions in vascular images, The steps include accessing the aforementioned vascular image, The steps include: applying a first ML-based vascular identifier to the vascular image to generate a first data structure that identifies paths extending along at least a portion of the vascular portion depicted in the vascular image; The steps include: applying a second vascular identifier to the vascular image to generate a second data structure that identifies one or more pixel masks containing each blob in the vascular portion; The steps include: combining the first data structure and the second data structure to generate a combined identification of a vascular portion forming a specific blood vessel depicted in the vascular image; Equipped with, The aforementioned path includes a vascular portion not included in the one or more pixel masks, method.

2. The first ML-based vascular identifier is configured to recognize a first anatomically defined vascular type based on the training input used to generate the ML-based vascular identifier. The second vascular identifier is also an ML-based vascular identifier and is configured to recognize a second anatomically defined vascular type based on the training input used to generate the ML-based vascular identifier. The method according to claim 1.

3. The first anatomically defined vascular type and the second anatomically defined vascular type are selected from the group consisting of LAD, other vessels in the LAD subtree, LCX, other vessels in the LCX subtree, RCA, and other vessels in the RCA subtree. The method according to claim 2.

4. The first ML-based vascular identifier and the second ML-based vascular identifier are configured to recognize the first anatomically defined vascular type and the second anatomically defined vascular type, respectively, using training inputs that include images and image location representations of the first and second anatomically defined vascular types. The method according to claim 2.

5. The first ML-based vascular identifier comprises a network in which the output layer is a regression layer. The method according to claim 1.

6. The second vascular identifier is an ML-based vascular identifier, and the output layer is a classification layer. The method according to claim 1.

7. The aforementioned step of combining is, Identifying a first region and a second region identified by both the first ML-based vascular identifier and the second vascular identifier, Based on a third region extending between the first region and the second region, the third region identified by either the first ML-based vascular identifier or the second vascular identifier is selected to be included in the combined identification, Includes, The third region includes the blood vessel portion that is not included in the one or more pixel masks in the path, The method according to claim 1.

8. At least one of the first data structure and the second data structure includes a misidentified vascular portion. The aforementioned method, The process includes a step of determining whether the misidentified vascular portion corresponds to a misidentification tendency representing a known misidentification tendency of either the first ML-based vascular identifier or the second vascular identifier, The combining step includes selectively including identified vascular portions in the combined identification according to the misidentification tendency, The misidentified blood vessel portion includes at least one of the blood vessel portions not included in the one or more pixel masks in the path, or the blood vessel portions not included in the path in the one or more pixel masks. The method according to claim 1.

Images