All published articles of this journal are available on ScienceDirect.
Comparative Analysis of Computational Models for CBV-based Detection of Ischemic Stroke in DSC-MRI: Influence of Signal-to-Noise Ratio and Tissue Type
Abstract
Introduction
Ischemic stroke remains a leading cause of disability and mortality, making a rapid and reliable diagnosis essential. Dynamic Susceptibility Contrast Magnetic Resonance Imaging (DSC-MRI) is widely used to assess cerebral perfusion, yet its diagnostic accuracy strongly depends on the computational model applied. This study investigates how model selection influences the reliability of CBV-based ischemic stroke detection under varying noise conditions and tissue types.
Methods
Simulated tissue signal curves were generated from clinical reference data and modified to reflect ischemic alterations across multiple noise levels. Cerebral Blood Volume (CBV) was estimated using two established approaches: the modified gamma variate function and a compartmental (triple-exponential) model. Diagnostic performance was evaluated by comparing the accuracy and robustness of CBV estimation.
Results
The compartmental model consistently outperformed the gamma variate function, providing more accurate and stable CBV estimates, particularly under high-noise conditions. In contrast, the gamma variate function demonstrated reduced robustness and greater sensitivity to noise.
Discussion
These findings underscore the importance of computational model selection in DSC-MRI analysis. The performance of the compartmental model suggests its potential for integration into clinical workflows, particularly in acute stroke care, where reliability under challenging conditions is crucial. However, this study has several limitations. Most importantly, the analysis was based on simulated tissue signal curves derived from clinical reference data rather than on in vivo measurements, which may not fully capture the complexity of real patient physiology.
Conclusion
Computational modeling influences the diagnostic value of DSC-MRI in ischemic stroke assessment. The compartmental model offers greater robustness and accuracy, supporting its use in diagnostic systems.