Systems and methods for generating classifying and quantitative analysis reports of aneurysms from medical image data
AssigneeThe Medical College of Wisconsin, Inc.
InventorsAli Bakhshinejad, Kevin M. Koch, Andrew S. Nencka
Life of this patent
Estimating the term…
Abstract
Aneurysms are classified and quantitatively analyzed based on medical image data acquired from a subject. In general, one or more algorithms are implemented to automatically classify, or otherwise diagnose, and measure aneurysms and their change over time. These algorithms make use of artificial intelligence and deep learning to develop quantitative analytics that can be consolidated into diagnostic reports.
Claims46
A method for generating a report that classifies and quantitatively analyzes one or more aneurysms in a subject, the method comprising:
- (a) accessing medical image data with a computer system, wherein the medical image data depict vasculature in a subject;
- (b) generating a binary vasculature mask by segmenting the medical image data using the computer system, wherein the binary vasculature mask has first binary values indicating a presence of vasculature at a voxel location and second binary values indicating an absence of vasculature at a voxel location;
- (c) generating classified feature data with the computer system by inputting the binary vasculature mask to a trained machine learning algorithm, generating output as the classified feature data, wherein the classified feature data classify regions in the vasculature of the subject as being associated with an aneurysm;
- (d) computing quantitative parameters with the computer system by fitting a basis set of geometrical objects to values in the classified feature data, wherein the quantitative parameters quantify aneurysm geometry;
- and
- (e) generating a report from the quantitative parameters using the computer system, wherein the report indicates a quantitative analysis of one or more aneurysms in the vasculature of the subject.
A method for generating a report that classifies and quantitatively analyzes one or more aneurysms in a subject, the method comprising:
- (a) accessing a probability map with a computer system, wherein the probability map indicates a probability of locations in a vasculature of a subject being associated with an aneurysm;
- (b) computing quantitative parameters with the computer system by fitting a basis set of geometrical objects to values in the probability map, wherein the quantitative parameters quantify aneurysm geometry;
- (c) generating a report from the quantitative parameters using the computer system, wherein the report indicates a quantitative analysis of one or more aneurysms in the vasculature of the subject;
- further comprising;
- accessing a binary vasculature mask of the subject with the computer system, wherein the binary vasculature mask has first binary values indicating a presence of vasculature at a voxel location and second binary values indicating an absence of vasculature at a voxel location;
- and
- wherein the report is generated also using the binary vasculature mask, such that the report includes a visual depiction of segmented vasculature in the subject.
A method for generating a report that classifies and quantitatively analyzes one or more aneurysms in a subject, the method comprising:
- (a) accessing a probability map with a computer system, wherein the probability map indicates a probability of locations in a vasculature of a subject being associated with an aneurysm;
- (b) computing quantitative parameters with the computer system by fitting a basis set of geometrical objects to values in the probability map, wherein the quantitative parameters quantify aneurysm geometry;
- (c) generating a report from the quantitative parameters using the computer system, wherein the report indicates a quantitative analysis of one or more aneurysms in the vasculature of the subject;
- and
- further comprising thresholding the probability map using a threshold value to identify regions in the probability map most probable to be associated with an aneurysm, and wherein the quantitative parameters are computed by fitting the basis set of geometrical objects to values in identified regions in the probability map.
Dependencies are inferred from the claim text.