Semiautomatic neck curve reconstruction for intracranial aneurysm rupture risk assessment based on morphological parameters.

Saalfeld, Sylvia; Berg, Philipp; Niemann, Annika; Luz, Maria; Preim, Bernhard; Beuing, Oliver

doi:10.1007/s11548-018-1848-x

by Sylvia Saalfeld, Philipp Berg, Annika Niemann, Maria Luz, Bernhard Preim, Oliver Beuing

Abstract:

PURPOSE: Morphological parameters of intracranial aneurysms (IAs) are well established for rupture risk assessment. However, a manual measurement is error-prone, not reproducible and cumbersome. For an automatic extraction of morphological parameters, a 3D neck curve reconstruction approach to delineate the aneurysm from the parent vessel is required. METHODS: We present a 3D semiautomatic aneurysm neck curve reconstruction for the automatic extraction of morphological parameters which was developed and evaluated with an experienced neuroradiologist. We calculate common parameters from the literature and include two novel angle-based parameters: the characteristic dome point angle and the angle difference of base points. RESULTS: We applied our method to 100 IAs acquired with rotational angiography in clinical routine. For validation, we compared our approach to manual segmentations yielding highly significant correlations. We analyzed 95 of these datasets regarding rupture state. Statistically significant differences were found in ruptured and unruptured groups for maximum diameter, maximum height, aspect ratio and the characteristic dome point angle. These parameters were also found to statistically significantly correlate with each other. CONCLUSIONS: The new 3D neck curve reconstruction provides robust results for all datasets. The reproducibility depends on the vessel tree centerline and the user input for the initial dome point and parameters characterizing the aneurysm neck region. The characteristic dome point angle as a new metric regarding rupture risk assessment can be extracted. It requires less computational effort than the complete neck curve reconstruction.

Reference:

Semiautomatic neck curve reconstruction for intracranial aneurysm rupture risk assessment based on morphological parameters. (Sylvia Saalfeld, Philipp Berg, Annika Niemann, Maria Luz, Bernhard Preim, Oliver Beuing), In International journal of computer assisted radiology and surgery, 2018.

Bibtex Entry:

@article{saalfeld_semiautomatic_2018,
	title = {Semiautomatic neck curve reconstruction for intracranial aneurysm rupture risk assessment based on morphological parameters.},
	issn = {1861-6429 1861-6410},
	doi = {10.1007/s11548-018-1848-x},
	abstract = {PURPOSE: Morphological parameters of intracranial aneurysms (IAs) are well established for rupture risk assessment. However, a manual measurement is error-prone, not reproducible and cumbersome. For an automatic extraction of morphological parameters, a 3D neck curve reconstruction approach to delineate the aneurysm from the parent vessel is required. METHODS: We present a 3D semiautomatic aneurysm neck curve reconstruction for the automatic extraction of morphological parameters which was developed and evaluated with an experienced neuroradiologist. We calculate common parameters from the literature and include two novel angle-based parameters: the characteristic dome point angle and the angle difference of base points. RESULTS: We applied our method to 100 IAs acquired with rotational angiography in clinical routine. For validation, we compared our approach to manual segmentations yielding highly significant correlations. We analyzed 95 of these datasets regarding rupture state. Statistically significant differences were found in ruptured and unruptured groups for maximum diameter, maximum height, aspect ratio and the characteristic dome point angle. These parameters were also found to statistically significantly correlate with each other. CONCLUSIONS: The new 3D neck curve reconstruction provides robust results for all datasets. The reproducibility depends on the vessel tree centerline and the user input for the initial dome point and parameters characterizing the aneurysm neck region. The characteristic dome point angle as a new metric regarding rupture risk assessment can be extracted. It requires less computational effort than the complete neck curve reconstruction.},
	language = {eng},
	number = {11/2018},
	journal = {International journal of computer assisted radiology and surgery},
	author = {Saalfeld, Sylvia and Berg, Philipp and Niemann, Annika and Luz, Maria and Preim, Bernhard and Beuing, Oliver},
	month = aug,
	year = {2018},
	pmid = {30159832},
	keywords = {Intracranial aneurysm, Morphological parameters, Neck curve, Rupture risk assessment}
}