Purpose: To develop a deep learning‐based Bayesian estimation for MRI reconstruction.Methods: We modeled the MRI reconstruction problem with Bayes’s theorem, following the recently proposed PixelCNN++ method. The image reconstruction from incomplete k‐space measurement was obtained by maximizing the posterior possibility. A generative network was utilized as the image prior, which was computationally tractable, and the k‐space data fidelity was enforced by using an equality constraint. The stochastic backpropagation was utilized to calculate the descent gradient in the process of maximum a posterior, and a projected subgradient method was used to impose the equality constraint. In contrast to the other deep learning reconstruction methods, the proposed one used the likelihood of prior as the training loss and the objective function in reconstruction to improve the image quality.Results: The proposed method showed an improved performance in preserving image details and reducing aliasing artifacts, compared with GRAPPA, ‐ESPRiT, model‐based deep learning architecture for inverse problems (MODL), and variational network (VN), last two were state‐of‐the‐art deep learning reconstruction methods. The proposed method generally achieved more than 3 dB peak signal‐to‐noise ratio improvement for compressed sensing and parallel imaging reconstructions compared with the other methods.Conclusions: The Bayesian estimation significantly improved the reconstruction performance, compared with the conventional ‐sparsity prior in compressed sensing reconstruction tasks. More importantly, the proposed reconstruction framework can be generalized for most MRI reconstruction scenarios.

A computational framework is presented for relating the kurtosis tensor for water diffusion in brain to tissue models of brain microstructure. The tissue models are assumed to be comprised of non-exchanging compartments that may be associated with various microstructural spaces separated by cell membranes. Within each compartment the water diffusion is regarded as Gaussian, although the diffusion for the full system would typically be non-Gaussian. The model parameters are determined so as to minimize the Frobenius norm of the difference between the measured kurtosis tensor and the model kurtosis tensor. This framework, referred to as kurtosis analysis of neural diffusion organization (KANDO), may be used to help provide a biophysical interpretation to the information provided by the kurtosis tensor. In addition, KANDO combined with diffusional kurtosis imaging can furnish a practical approach for developing candidate biomarkers for neuropathologies that involve alterations in tissue microstructure. KANDO is illustrated for simple tissue models of white and gray matter using data obtained from healthy human subjects.

Background: We proposed a population graph with Transformer-generated and clinical features for the purpose of predicting overall survival (OS) and recurrence-free survival (RFS) for patients with early stage non-small cell lung carcinomas and to compare this model with traditional models.Methods: The study included 1705 patients with lung cancer (stages I and II), and a public data set for external validation (n=127). We proposed a graph with edges representing non-imaging patient characteristics and nodes representing imaging tumour region characteristics generated by a pretrained Vision Transformer. The model was compared with a TNM model and a ResNet-Graph model. To evaluate the models' performance, the area under the receiver operator characteristic curve (ROC-AUC) was calculated for both OS and RFS prediction. The Kaplan–Meier method was used to generate prognostic and survival estimates for low- and high-risk groups, along with net reclassification improvement (NRI), integrated discrimination improvement (IDI), and decision curve analysis. An additional subanalysis was conducted to examine the relationship between clinical data and imaging features associated with risk prediction.Results: Our model achieved AUC values of 0.785 (95% confidence interval [CI]: 0.716–0.855) and 0.695 (95% CI: 0.603–0.787) on the testing and external data sets for OS prediction, and 0.726 (95% CI: 0.653–0.800) and 0.700 (95% CI: 0.615–0.785) for RFS prediction. Additional survival analyses indicated that our model outperformed the present TNM and ResNet-Graph models in terms of net benefit for survival prediction.Conclusions: Our Transformer-Graph model was effective at predicting survival in patients with early stage lung cancer, which was constructed using both imaging and non-imaging clinical features. Some high-risk patients were distinguishable by using a similarity score function defined by non-imaging characteristics such as age, gender, histology type, and tumour location, while Transformer-generated features demonstrated additional benefits for patients whose non-imaging characteristics were non-discriminatory for survival outcomes.Funding: The study was supported by the National Natural Science Foundation of China (91959126, 8210071009), and Science and Technology Commission of Shanghai Municipality (20XD1403000, 21YF1438200).

Aging primarily affects memory and executive functions, a relationship that may be underpinned by the fact that almost all adults over 60 years old develop small vessel disease (SVD). The fact that a wide range of neuropathologies could only explain up to 43% of the variation in age-related cognitive impairment suggests that other factors, such as cognitive reserve, may play a role in the brain’s resilience against aging-related cognitive decline. This study aims to examine the relationship between structural–functional-connectivity coupling (SFC), and aging, cognitive abilities and reserve, and SVD-related neuropathologies using a cohort of n = 176 healthy elders from the Harvard Aging Brain Study. The SFC is a recently proposed biomarker that reflects the extent to which anatomical brain connections can predict coordinated neural activity. After controlling for the effect of age, sex, and years of education, global SFC, as well as the intra-network SFC of the dorsolateral somatomotor and dorsal attention networks, and the inter-network SFC between dorsolateral somatomotor and frontoparietal networks decreased with age. The global SFC decreased with total cognitive score. There were significant interaction effects between years of education versus white matter hyperintensities and between years of education versus cerebral microbleeds on inter-network SFC. Enlarged perivascular space in basal ganglia was associated with higher inter-network SFC. Our results suggest that cognitive ability is associated with brain coupling at the global level and cognitive reserve with brain coupling at the inter-functional-brain-cluster level with interaction effect from white matter hyperintensities and cerebral microbleed in a cohort of healthy elderlies.

Different cerebral small vessel disease (SVD) lesion types have been shown to disrupt structural brain network individually. Considering that they often coexist, we investigated the relation between their collective effect using the recently proposed total SVD score and structural brain network on MRI in 95 patients with first transient ischemic attack (TIA) or ischemic stroke. Fifty-nine patients with and 36 without any SVD lesions were included. The total SVD score was recorded. Diffusion tensor imaging was performed to estimate structural brain connections for subsequent brain connectivity analysis. The global efficiency and characteristic path length of the structural brain network are respectively lower and higher due to SVD. Lower nodal efficiency is also found in the insular, precuneus, supplementary motor area, paracentral lobule, putamen and hippocampus. The total SVD score is correlated with global network measures, the local clustering coefficient and nodal efficiency of hippocampus, and the nodal efficiency of paracentral lobule. We have successfully demonstrated that the disruption of global and local structural brain networks are associated with the increase in the overall SVD severity or burden of patients with TIA or first-time stroke.

Purpose: This study aims to develop a novel time‐resolved magnetic resonance fingerprinting (TR‐MRF) technique for respiratory motion imaging applications.Materials and methods: The TR‐MRF technique consists of repeated MRF acquisitions using an unbalanced steady‐state free precession sequence with spiral‐in–spiral‐out trajectory. Time‐resolved magnetic resonance fingerprinting was first tested via computer simulation using a four‐dimensional (4D) extended cardiac‐torso (XCAT) phantom for both regular and irregular breathing profiles, and was tested in three healthy volunteers. Parametric TR‐MRF maps at different respiratory phases were subsequently estimated using our TR‐MRF sorting and reconstruction techniques. The resulting TR‐MRF maps were evaluated using a set of metrices related to radiotherapy applications, including absolute difference in motion amplitude, error in the amplitude of diaphragm motion (ADM), tumor volume error (TVE), signal‐to‐noise ratio (SNR), and tumor contrast.Results: TR‐MRF maps with regular and irregular breathing were successfully generated in XCAT phantom. Numerical simulations showed that the TVE were 1.6 ± 2.7% and 1.3 ± 2.2%, the average absolute differences in tumor motion amplitude were 0.3 ± 0.7 mm and 0.3 ± 0.6 mm, and the ADM were 4.1 ± 0.9% and 3.5 ± 0.9% for irregular and regular breathing, respectively. The SNR of the T₁ and T₂ maps of the liver and the tumor were generally higher for regular breathing compared to irregular breathing, whereas tumor‐to‐liver contrast is similar between the two breathing patterns. The proposed technique was successfully implemented on the healthy volunteers.Conclusion: We have successfully demonstrated in both digital phantom and healthy subjects a novel TR‐MRF technique capable of imaging respiratory motions with simultaneous quantification of MR multiparametric maps.