Back to Search Start Over

Uncertainty-aware deep-learning model for prediction of supratentorial hematoma expansion from admission non-contrast head computed tomography scan.

Authors :
Tran AT
Zeevi T
Haider SP
Abou Karam G
Berson ER
Tharmaseelan H
Qureshi AI
Sanelli PC
Werring DJ
Malhotra A
Petersen NH
de Havenon A
Falcone GJ
Sheth KN
Payabvash S
Source :
NPJ digital medicine [NPJ Digit Med] 2024 Feb 06; Vol. 7 (1), pp. 26. Date of Electronic Publication: 2024 Feb 06.
Publication Year :
2024

Abstract

Hematoma expansion (HE) is a modifiable risk factor and a potential treatment target in patients with intracerebral hemorrhage (ICH). We aimed to train and validate deep-learning models for high-confidence prediction of supratentorial ICH expansion, based on admission non-contrast head Computed Tomography (CT). Applying Monte Carlo dropout and entropy of deep-learning model predictions, we estimated the model uncertainty and identified patients at high risk of HE with high confidence. Using the receiver operating characteristics area under the curve (AUC), we compared the deep-learning model prediction performance with multivariable models based on visual markers of HE determined by expert reviewers. We randomly split a multicentric dataset of patients (4-to-1) into training/cross-validation (n = 634) versus test (n = 159) cohorts. We trained and tested separate models for prediction of ≥6 mL and ≥3 mL ICH expansion. The deep-learning models achieved an AUC = 0.81 for high-confidence prediction of HE <subscript>≥6 mL</subscript> and AUC = 0.80 for prediction of HE <subscript>≥3 mL</subscript> , which were higher than visual maker models AUC = 0.69 for HE <subscript>≥6 mL</subscript> (p = 0.036) and AUC = 0.68 for HE <subscript>≥3 mL</subscript> (p = 0.043). Our results show that fully automated deep-learning models can identify patients at risk of supratentorial ICH expansion based on admission non-contrast head CT, with high confidence, and more accurately than benchmark visual markers.<br /> (© 2024. The Author(s).)

Details

Language :
English
ISSN :
2398-6352
Volume :
7
Issue :
1
Database :
MEDLINE
Journal :
NPJ digital medicine
Publication Type :
Academic Journal
Accession number :
38321131
Full Text :
https://doi.org/10.1038/s41746-024-01007-w