A Hybrid Method for Coronary Artery Stenoses Detection and Quantification in CTA Images
This paper describes a hybrid algorithm to detect and quantify the coronary artery stenoses in 3D CT Angiography images. The approach combines the Hessian matrix based “vesselness” filter with three dimensional region growing algorithm for segmentation of the coronary arteries. Centerlines of the segmented arteries are extracted using an in-house fast marching based method. Detection and quantification of the stenoses are then accomplished by estimating the vessel diameter at each centerline location via plane fitting, and applying linear regression analysis on the estimated diameter profile. For the detection stage, a sensitivity of 21% and a PPV of 33% are achieved as compared to QCA, while a sensitivity of 17% and a PPV of 25% are achieved as compared to CTA. However, the stenoses are quantified with an averaged absolute difference of only 44.9% as compared to QCA. The approach allowed evaluation of average data within ten minutes. In conclusion, the algorithm performed relatively well for detection purposes (threshold at 50% diameter reduction); but it remains to be improved for better categorization of severity of stenosis as mild, moderate, and severe.
Detection confusion tables
|Calc. cat.||QCA (per segment)||CTA (per lesion)|
The results of this method are based on the following centerlines: LKEB team manual.
Detection (QCA per segment / CTA per lesion)
| ||%||rank||%||rank||%||rank||%||rank|| |
|0 (0 - 0.1)||0.0||15.0||0.0||15.0||14.3||14.0||16.7||11.0||13.8|
|1 (0.1 - 10)||0.0||16.0||0.0||16.0||0.0||17.0||0.0||17.0||16.5|
|2 (11 - 100)||28.6||13.0||40.0||4.0||18.2||16.0||25.0||8.0||10.2|
|3 (101 - 400)||30.0||15.0||33.3||7.0||21.1||14.0||44.4||5.0||10.2|
These results are based on 30 datasets and 19 submissions.
Avg. Abs. diff.
| ||%||rank||%||rank||Κ||rank|| |
|0 (0 - 0.1)||45.8||12.0||51.9||12.0||0.20||10.0||11.0|
|1 (0.1 - 10)||37.1||10.0||40.1||10.0||0.06||12.0||11.0|
|2 (11 - 100)||46.4||11.0||50.9||10.0||0.24||10.0||10.2|
|3 (101 - 400)||48.8||11.0||55.2||13.0||0.21||9.0||10.5|