Perfusion parameter estimation using neural networks and data augmentation
This addresses the challenge of accurate and efficient perfusion analysis for stroke diagnosis, though it appears incremental as it builds on existing neural network approaches with a specific data augmentation technique.
The paper tackled the problem of perfusion parameter estimation in acute stroke imaging, which traditionally uses ill-conditioned deconvolution methods, by proposing a neural network with data augmentation to predict parameters directly from measurements. The result showed that the neural network provided better estimations for CBF and Tmax than a state-of-the-art deconvolution method across various noise levels, using less than 100 datasets.
Perfusion imaging plays a crucial role in acute stroke diagnosis and treatment decision making. Current perfusion analysis relies on deconvolution of the measured signals, an operation that is mathematically ill-conditioned and requires strong regularization. We propose a neural network and a data augmentation approach to predict perfusion parameters directly from the native measurements. A comparison on simulated CT Perfusion data shows that the neural network provides better estimations for both CBF and Tmax than a state of the art deconvolution method, and this over a wide range of noise levels. The proposed data augmentation enables to achieve these results with less than 100 datasets.