An explainable multicentric analysis for understanding the aetiology of intracranial aneurysm disease

Delucchi, Matteo; Spinner, Georg R.; Bijlenga, Philippe; Morel, Sandrine; Hostettler, Isabel; Werring, David; Wostrack, Maria; Meyer, Bernhard; Bourcier, Romain; Lindgren, Antti; Bakker, Mark K.; Ruigrok, Ynte M.; Furrer, Reinhard; Hirsch, Sven

doi:10.3390/ctn7040039

Publication type:	Conference poster
Type of review:	Peer review (publication)
Title:	An explainable multicentric analysis for understanding the aetiology of intracranial aneurysm disease
Authors:	Delucchi, Matteo Spinner, Georg R. Bijlenga, Philippe Morel, Sandrine Hostettler, Isabel Werring, David Wostrack, Maria Meyer, Bernhard Bourcier, Romain Lindgren, Antti Bakker, Mark K. Ruigrok, Ynte M. Furrer, Reinhard Hirsch, Sven
et. al:	No
DOI:	10.3390/ctn7040039
Published in:	Clinical and Translational Neuroscience
Volume(Issue):	7
Issue:	4
Page(s):	46
Pages to:	47
Issue Date:	15-Nov-2023
Publisher / Ed. Institution:	MDPI
ISSN:	2514-183X
Language:	English
Subject (DDC):	616: Internal medicine and diseases
Abstract:	Aims: Intracranial aneurysms (IAs) are present in approximately 3% of the population [Vlak et al., 2011]. Rupture of IA leads to an aneurysmal subarachnoid haemorrhage with often poor functional outcomes [Lawton and Vates, 2017]. Unruptured IA (UIA) detection rates increase with advances in imaging technologies [Bijlenga et al., 2017]. The complexity of UIA treatment decision making is compounded by the difficulty of accurately predicting the risk of rupture and the lack of understanding of how modifiable factors affect IA rupture [Lognon et al., 2022]. Here, we show an explainable model for IA rupture based on easily accessible phenotypic risk factors. Methods: This model development study was validated on IA patient‑level registry data in a multicenter (n = 7) retrospective case‑control design with 9 phenotypic risk factors. Data were analysed using discrete and additive Bayesian network (BN) models. Expert knowledge a priori restricted the model search space, leading to a sparse network representing clinical expertise [Delucchi et al., 2022]. Results: We included 8604 patients with IA (median age 54y, IQR 45−63, 67% female), of whom 4254 (49%) patients had IA at the time of diagnosis. The point prevalence of recommended follow‑up patients with UIA [Bijlenga et al., 2013] was estimated to be approximately 43%. The joint probability distribution estimates patient‑specific disease management recommendations. Preliminary results indicate, for example, that older women with an IA in a low‑risk location are unlikely to experience a rupture (OR_{rupture} = 0.05), and patients who are active smokers at the time of IA diagnosis generally have a higher likelihood to be diagnosed with a ruptured IA (OR_{rupture} = 1.46). Conclusions: This study shows that mixed‑effect additive BNs can help clinicians understand the aetiology of IA rupture and may have potential for providing personalised guidance for UIA management. Our findings anticipate the starting point for IA disease models that encompass the entire evolution of the disease and could be refined in a more extensive prospective cohort study to develop a user‑friendly bedside decision support application.
URI:	https://digitalcollection.zhaw.ch/handle/11475/29370
Fulltext version:	Published version
License (according to publishing contract):	Licence according to publishing contract
Departement:	Life Sciences and Facility Management
Organisational Unit:	Institute of Computational Life Sciences (ICLS)
Published as part of the ZHAW project:	Modellierung multizentrischer und dynamischer Schlaganfall Gesundheitsdaten Stroke DynamiX
Appears in collections:	Publikationen Life Sciences und Facility Management

Files in This Item:

There are no files associated with this item.

Show full item record

Delucchi, M., Spinner, G. R., Bijlenga, P., Morel, S., Hostettler, I., Werring, D., Wostrack, M., Meyer, B., Bourcier, R., Lindgren, A., Bakker, M. K., Ruigrok, Y. M., Furrer, R., & Hirsch, S. (2023). An explainable multicentric analysis for understanding the aetiology of intracranial aneurysm disease [Conference poster]. Clinical and Translational Neuroscience, 7(4), 46–47. https://doi.org/10.3390/ctn7040039

Delucchi, M. et al. (2023) ‘An explainable multicentric analysis for understanding the aetiology of intracranial aneurysm disease’, in Clinical and Translational Neuroscience. MDPI, pp. 46–47. Available at: https://doi.org/10.3390/ctn7040039.

M. Delucchi et al., “An explainable multicentric analysis for understanding the aetiology of intracranial aneurysm disease,” in Clinical and Translational Neuroscience, Nov. 2023, vol. 7, no. 4, pp. 46–47. doi: 10.3390/ctn7040039.

DELUCCHI, Matteo, Georg R. SPINNER, Philippe BIJLENGA, Sandrine MOREL, Isabel HOSTETTLER, David WERRING, Maria WOSTRACK, Bernhard MEYER, Romain BOURCIER, Antti LINDGREN, Mark K. BAKKER, Ynte M. RUIGROK, Reinhard FURRER und Sven HIRSCH, 2023. An explainable multicentric analysis for understanding the aetiology of intracranial aneurysm disease. In: Clinical and Translational Neuroscience. Conference poster. MDPI. 15 November 2023. S. 46–47

Delucchi, Matteo, Georg R. Spinner, Philippe Bijlenga, Sandrine Morel, Isabel Hostettler, David Werring, Maria Wostrack, et al. 2023. “An Explainable Multicentric Analysis for Understanding the Aetiology of Intracranial Aneurysm Disease.” Conference poster. In Clinical and Translational Neuroscience, 7:46–47. MDPI. https://doi.org/10.3390/ctn7040039.

Delucchi, Matteo, et al. “An Explainable Multicentric Analysis for Understanding the Aetiology of Intracranial Aneurysm Disease.” Clinical and Translational Neuroscience, vol. 7, no. 4, MDPI, 2023, pp. 46–47, https://doi.org/10.3390/ctn7040039.