Computer-Aided Detection and Quantification of Intracranial Aneurysms
Early detection, assessment and treatment of intracranial aneurysms is important to prevent rupture, which may cause death. We propose a framework for detection and quantification of morphology of the aneurysms. A novel detector using decision forests, wh
- PDF / 3,088,627 Bytes
- 8 Pages / 439.363 x 666.131 pts Page_size
- 17 Downloads / 197 Views
ract. Early detection, assessment and treatment of intracranial aneurysms is important to prevent rupture, which may cause death. We propose a framework for detection and quantification of morphology of the aneurysms. A novel detector using decision forests, which employ responses of blobness and vesselness filters encoded in rotation invariant and scale normalized frequency components of spherical harmonics representation is proposed. Aneurysm location is used to seed growcut segmentation, followed by improved neck extraction based on intravascular ray-casting and robust closed-curve fit to the segmentation. Aneurysm segmentation and neck curve are used to compute three morphologic metrics: neck width, dome height and aspect ratio. The proposed framework was evaluated on ten cerebral 3D-DSA images containing saccular aneurysms. Sensitivity of aneurysm detection was 100% at 0.4 false positives per image. Compared to measurements of two expert raters, the values of metrics obtained by the proposed framework were accurate and, thus, suitable for assessing the risk of rupture. Keywords: intracranial aneurysm, rupture, detection, multiscale enhancement, random forests, segmentation, neck extraction, morphology.
1
Introduction
The prevalence of intracranial aneurysms is between 2%-5% of the world population and, although aneurysm rupture is a rather rare event, the majority of patients that experience the rupture die of subarachnoid hemorrhage [3]. To prevent such fatal events, aneurysms should be detected, assessed and treated as early as possible. After an aneurysm is detected, the risk of rupture is assessed through quantitative studies of its morphology and hemodynamics so as to prioritize the treatment of patients. Quantifying the morphology of aneurysms can be performed manually using a 3D angiographic image like 3D digital subtraction angiography (3D-DSA). However, tasks like manual detection, segmentation, isolation and measurement of the aneurysm in a complex 3D vascular tree, usually based on observing 2D cross-sections of the 3D image, are tedious and time consuming to perform for a clinician. To assist the clinician, but also to improve the accuracy, reliability and reproducibility of the outcome [1], there is a need for computer-aided detection and quantification of the aneurysms. c Springer International Publishing Switzerland 2015 N. Navab et al. (Eds.): MICCAI 2015, Part II, LNCS 9350, pp. 3–10, 2015. DOI: 10.1007/978-3-319-24571-3_1
4
T. Jerman et al.
Fig. 1. Image analysis framework for aneurysm detection, segmentation and neck extraction based on a 3D angiogram.
Although several methods for detecting intracranial aneurysms [6,4], segmenting vascular structures [7] and extracting the aneurysm’s neck curve [1], were proposed in the past, to the best of our knowledge, these methods still need to be further improved before being incorporated into a computer-aided detection and quantification framework. Of the aneurysm detection, segmentation, and quantification tasks, detection seems to be the most d
Data Loading...
