Machine learning segmentation of core and penumbra from acute stroke CT perfusion data

Freda Werdiger; Mark W Parsons; Milanka Visser; Christopher Levi; Neil Spratt; Tim Kleinig; Longting Lin; Andrew Bivard

doi:10.3389/fneur.2023.1098562

Machine learning segmentation of core and penumbra from acute stroke CT perfusion data

Front Neurol. 2023 Feb 23:14:1098562. doi: 10.3389/fneur.2023.1098562. eCollection 2023.

Authors

Freda Werdiger^{1

2}, Mark W Parsons^{3

4

5}, Milanka Visser^{1

2}, Christopher Levi^{6

7}, Neil Spratt^{6

7}, Tim Kleinig⁸, Longting Lin^{6

7}, Andrew Bivard^{1

2}

Affiliations

¹ Melbourne Brain Centre, Department of Neurology, The Royal Melbourne Hospital, Melbourne, VIC, Australia.
² Department of Medicine, University of Melbourne, Melbourne, VIC, Australia.
³ Southwestern Sydney Clinical School, University of New South Wales, Sydney, NSW, Australia.
⁴ Department of Neurology, Liverpool Hospital, Liverpool, NSW, Australia.
⁵ Ingham Institute for Applied Medical Research, Liverpool, NSW, Australia.
⁶ Hunter Medical Research Institution, University of Newcastle, Newcastle, NSW, Australia.
⁷ Department of Neurology, John Hunter Hospital, University of Newcastle, Newcastle, NSW, Australia.
⁸ Department of Neurology, Royal Adelaide Hospital, Adelaide, SA, Australia.

Abstract

Introduction: Computed tomography perfusion (CTP) imaging is widely used in cases of suspected acute ischemic stroke to positively identify ischemia and assess suitability for treatment through identification of reversible and irreversible tissue injury. Traditionally, this has been done via setting single perfusion thresholds on two or four CTP parameter maps. We present an alternative model for the estimation of tissue fate using multiple perfusion measures simultaneously.

Methods: We used machine learning (ML) models based on four different algorithms, combining four CTP measures (cerebral blood flow, cerebral blood volume, mean transit time and delay time) plus 3D-neighborhood (patch) analysis to predict the acute ischemic core and perfusion lesion volumes. The model was developed using 86 patient images, and then tested further on 22 images.

Results: XGBoost was the highest-performing algorithm. With standard threshold-based core and penumbra measures as the reference, the model demonstrated moderate agreement in segmenting core and penumbra on test images. Dice similarity coefficients for core and penumbra were 0.38 ± 0.26 and 0.50 ± 0.21, respectively, demonstrating moderate agreement. Skull-related image artefacts contributed to lower accuracy.

Discussion: Further development may enable us to move beyond the current overly simplistic core and penumbra definitions using single thresholds where a single error or artefact may lead to substantial error.

Keywords: CT perfusion imaging; acute ischemic stroke; ischemic core; machine learning; penumbra.