Abstract:Abstract Background . Computed tomography perfusion (CTP) imaging plays a critical role in the acute stroke syndrome assessment due to its widespread availability, speed of image acquisition, and relatively low cost. However, due to its repeated scanning protocol, CTP imaging involves a substantial radiation dose, which might increase potential cancer risks. Methods . In this work, we present a novel deep learning model called non-local perfusion texture learning network (NPTN) for high-quality CTP imaging at low-dose cases. Specifically, considering abundant similarities in the CTP images, i.e. latent self-similarities within the non-local region in the CTP images, we firstly search the most similar pixels from the adjacent frames within a fixed search window to obtain the non-local similarities and to construct non-local textures vector. Then, both the low-dose frame and these non-local textures from adjacent frames are fed into a convolution neural network to predict high-quality CTP images, which can help better characterize the structure details and contrast variants in the targeted CTP image rather than simply utilizing the targeted frame itself. The residual learning strategy and batch normalization are utilized to boost the performance of the convolution neural network. In the experiment, the CTP images of 31 patients with suspected stroke disease are collected to demonstrate the performance of the presented NPTN method. Results . The results show the presented NPTN method obtains superior performance compared with the competing methods. From numerical value, at all dose levels, the presented NPTN method has achieved around 3.0 dB improvement of average PSNR, an increase of around 1.4% of average SSIM, and a decrease of around 4.8% of average RMSE in the low-dose CTP reconstruction task, and also has achieved an increase of around 3.4% of average SSIM and a decrease of around 61.1% of average RMSE in the cerebral blood flow (CBF) estimation task. Conclusions . The presented NPTN method can obtain high-quality CTP images and estimate high-accuracy CBF map by characterizing more structure details and contrast variants in the CTP image and outperform the competing methods at low-dose cases.

RTN: Reinforced Transformer Network for Coronary CT Angiography Vessel-level Image Quality Assessment

Contrast-Medium Anisotropy-Aware Tensor Total Variation Model for Robust Cerebral Perfusion CT Reconstruction with Low-Dose Scans

Transformer Network for Significant Stenosis Detection in CCTA of Coronary Arteries

An Efficient Iterative Cerebral Perfusion CT Reconstruction Via Low-Rank Tensor Decomposition with Spatial-Temporal Total Variation Regularization.

TCU-Net: Transformer Embedded in Convolutional U-Shaped Network for Retinal Vessel Segmentation

TransCC: Transformer Network for Coronary Artery CCTA Segmentation

Multi-view Contour-constrained Transformer Network for Thin-cap Fibroatheroma Identification

Transformer-based Dual-domain Network for Few-view Dedicated Cardiac SPECT Image Reconstructions

DistilIQA: Distilling vision transformers for no-reference perceptual CT image quality assessment

CPCT-LRTDTV: Cerebral Perfusion CT Image Restoration Via a Low Rank Tensor Decomposition with Total Variation Regularization

Dual-domain sparse-view CT reconstruction with Transformers

Hybrid-Domain Integrative Transformer Iterative Network for Spectral CT Imaging

Predicting Thrombectomy Recanalization from CT Imaging Using Deep Learning Models

Learning non-local perfusion textures for high-quality computed tomography perfusion imaging

RTNet: Relation Transformer Network for Diabetic Retinopathy Multi-lesion Segmentation

A transformer-based pyramid network for coronary calcified plaque segmentation in intravascular optical coherence tomography images

Pure Vision Transformer (CT-ViT) with Noise2Neighbors Interpolation for Low-Dose CT Image Denoising

Application of Vision-Series Transformer in Screening for Coronary Heart Diseases Using Coronary CT Angiography.

STIR-Net: Deep Spatial-Temporal Image Restoration Net for Radiation Reduction in CT Perfusion

CT perfusion parameter estimation in stroke using neural network with transformer and physical model priors

TransCT: Dual-path Transformer for Low Dose Computed Tomography