Time-Dependent Deep Image Prior for Dynamic MRI.

We propose a novel unsupervised deep-learning-based algorithm for dynamic magnetic resonance imaging (MRI) reconstruction. Dynamic MRI requires rapid data acquisition for the study of moving organs such as the heart. We introduce a generalized version of the deep-image-prior approach, which optimizes the weights of a reconstruction network to fit a sequence of sparsely acquired dynamic MRI measurements. Our method needs neither prior training nor additional data. In particular, for cardiac images, it does not require the marking of heartbeats or the reordering of spokes. The key ingredients of our method are threefold: 1) a fixed low-dimensional manifold that encodes the temporal variations of images; 2) a network that maps the manifold into a more expressive latent space; and 3) a convolutional neural network that generates a dynamic series of MRI images from the latent variables and that favors their consistency with the measurements in k -space. Our method outperforms the state-of-the-art methods quantitatively and qualitatively in both retrospective and real fetal cardiac datasets. To the best of our knowledge, this is the first unsupervised deep-learning-based method that can reconstruct the continuous variation of dynamic MRI sequences with high spatial resolution.

Deep Learning Diffuse Optical Tomography.

Diffuse optical tomography (DOT) has been investigated as an alternative imaging modality for breast cancer detection thanks to its excellent contrast to hemoglobin oxidization level. However, due to the complicated non-linear photon scattering physics and ill-posedness, the conventional reconstruction algorithms are sensitive to imaging parameters such as boundary conditions. To address this, here we propose a novel deep learning approach that learns non-linear photon scattering physics and obtains an accurate three dimensional (3D) distribution of optical anomalies. In contrast to the traditional black-box deep learning approaches, our deep network is designed to invert the Lippman-Schwinger integral equation using the recent mathematical theory of deep convolutional framelets. As an example of clinical relevance, we applied the method to our prototype DOT system. We show that our deep neural network, trained with only simulation data, can accurately recover the location of anomalies within biomimetic phantoms and live animals without the use of an exogenous contrast agent.

Deep Residual Learning for Accelerated MRI Using Magnitude and Phase Networks.

OBJECTIVE: Accelerated magnetic resonance (MR) image acquisition with compressed sensing (CS) and parallel imaging is a powerful method to reduce MR imaging scan time. However, many reconstruction algorithms have high computational costs. To address this, we investigate deep residual learning networks to remove aliasing artifacts from artifact corrupted images. METHODS: The deep residual learning networks are composed of magnitude and phase networks that are separately trained. If both phase and magnitude information are available, the proposed algorithm can work as an iterative k-space interpolation algorithm using framelet representation. When only magnitude data are available, the proposed approach works as an image domain postprocessing algorithm. RESULTS: Even with strong coherent aliasing artifacts, the proposed network successfully learned and removed the aliasing artifacts, whereas current parallel and CS reconstruction methods were unable to remove these artifacts. CONCLUSION: Comparisons using single and multiple coil acquisition show that the proposed residual network provides good reconstruction results with orders of magnitude faster computational time than existing CS methods. SIGNIFICANCE: The proposed deep learning framework may have a great potential for accelerated MR reconstruction by generating accurate results immediately.

Deep Convolutional Framelet Denosing for Low-Dose CT via Wavelet Residual Network.

Model-based iterative reconstruction algorithms for low-dose X-ray computed tomography (CT) are computationally expensive. To address this problem, we recently proposed a deep convolutional neural network (CNN) for low-dose X-ray CT and won the second place in 2016 AAPM Low-Dose CT Grand Challenge. However, some of the textures were not fully recovered. To address this problem, here we propose a novel framelet-based denoising algorithm using wavelet residual network which synergistically combines the expressive power of deep learning and the performance guarantee from the framelet-based denoising algorithms. The new algorithms were inspired by the recent interpretation of the deep CNN as a cascaded convolution framelet signal representation. Extensive experimental results confirm that the proposed networks have significantly improved performance and preserve the detail texture of the original images.

Alteration in the Local and Global Functional Connectivity of Resting State Networks in Parkinson's Disease.

OBJECTIVE: Parkinson's disease (PD) is a neurodegenerative disorder that mainly leads to the impairment of patients' motor function, as well as of cognition, as it progresses. This study tried to investigate the impact of PD on the resting state functional connectivity of the default mode network (DMN), as well as of the entire brain. METHODS: Sixty patients with PD were included and compared to 60 matched normal control (NC) subjects. For the local connectivity analysis, the resting state fMRI data were analyzed by seed-based correlation analyses, and then a novel persistent homology analysis was implemented to examine the connectivity from a global perspective. RESULTS: The functional connectivity of the DMN was decreased in the PD group compared to the NC, with a stronger difference in the medial prefrontal cortex. Moreover, the results of the persistent homology analysis indicated that the PD group had a more locally connected and less globally connected network compared to the NC. CONCLUSION: Our findings suggest that the DMN is altered in PD, and persistent homology analysis, as a useful measure of the topological characteristics of the networks from a broader perspective, was able to identify changes in the large-scale functional organization of the patients' brain.

Topological persistence vineyard for dynamic functional brain connectivity during resting and gaming stages.

BACKGROUND: Recent studies have shown the dynamic functional connectivity (FC) of the brain. Accordingly, new challenges have arisen for analyzing and interpreting this rich information. NEW METHOD: We identified the patterns of coherent FC using a novel method in computational topology called the persistence vineyard. It has been developed to track the characteristic change of the network topology under data perturbations in a threshold-free manner. RESULTS: We showed the relevance of this new approach by examining the dynamic FC in the resting and gaming stages of 26 healthy subjects. Our proposed method revealed stage and band-specific FC states that were topologically robust. COMPARISON WITH EXISTING METHODS: While principal component analysis (PCA) estimated similar patterns to our FC states, it produced spurious connectivity due to its orthogonality assumption. Temporal variations of local and global network properties were examined with graph measures. However, unlike the persistence vineyard approach, their results were affected by the network density and its unknown topology. CONCLUSIONS: Unlike the existing methods, the persistence vineyard provided a more reliable and robust way to estimate FC states. Their extracted network topology changes showed patterns consistent with those of previous studies. Therefore, it may be a potentially powerful tool for studying the dynamic brain network.

Monitoring of bacterial community in a coniferous forest soil after a wildfire.

Changes in the soil bacterial community of a coniferous forest were analyzed to assess microbial responses to wildfire. Soil samples were collected from three different depths in lightly and severely burned areas, as well as a nearby unburned control area. Direct bacterial counts ranged from 3.3-22.6 x 10(8) cells/(g.soil). In surface soil, direct bacterial counts of unburned soil exhibited a great degree of fluctuation. Those in lightly burned soil changed less, but no significant variation was observed in the severely burned soil. The fluctuations of direct bacterial count were less in the middle and deep soil layers. The structure of the bacterial community was analyzed via the fluorescent in situ hybridization method. The number of bacteria detected with the eubacteria-targeted probe out of the direct bacterial count varied from 30.3 to 84.7%, and these ratios were generally higher in the burned soils than in the unburned control soils. In the surface unburned soil, the ratios of alpha-, beta- and gamma-proteobacteria, Cytophaga-Flavobacterium group, and other eubacteria groups to total eubacteria were 9.9, 10.6, 15.5, 9.0, and 55.0%, respectively, and these ratios were relatively stable. The ratios of alpha-, beta- and gamma-proteobacteria, and Cytophaga-Flavobacterium group to total eubacteria increased immediately after the wildfire, and the other eubacterial proportions decreased in the surface and middle layer soils. By way of contrast, the composition of the 5 groups of eubacteria in the subsurface soil exhibited no significant fluctuations during the entire period. The total bacterial population and bacterial community structure disturbed by wildfire soon began to recover, and original levels seemed to be restored 3 months after the wildfire.

Monitoring of soil bacterial community and some inoculated bacteria after prescribed fire in microcosm.

The soil bacterial community and some inoculated bacteria were monitored to assess the microbial responses to prescribed fire in their microcosm. An acridine orange direct count of the bacteria in the unburned control soil were maintained at a relatively stable level (2.0 approximately 2.7 x 10(9) cells/g(-1).soil) during the 180 day study period. The number of bacteria in the surface soil was decreased by fire, but was restored after 3 months. Inoculation of some bacteria increased the number of inoculated bacteria several times and these elevated levels lasted several months. The ratios of eubacteria detected by a fluorescent in situ hybridization (FISH) method to direct bacterial count were in the range of 60 approximately 80% during the study period, with the exception of some lower values at the beginning, but there were no definite differences between the burned and unburned soils or the inoculated and uninoculated soils. In the unburned control soil, the ratios of alpha-, beta- and gamma-subgroups of the proteobacteria, Cytophaga-Flavobacterium and other eubacteria groups to that of the entire eubacteria were 13.7, 31.7, 17.1, 16.8 and 20.8%, respectively, at time 0. The overall change on the patterns of the ratios of the 5 subgroups of eubacteria in the uninoculated burned and inoculated soils were similar to those of the unburned control soil, with the exception of some minor variations during the initial period. The proportions of each group of eubacteria became similar in the different microcosms after 6 months, which may indicate the recovery of the original soil microbial community structure after fire or the inoculation of some bacteria. The populations of Azotobacter vinelandii, Bacillus megaterium and Pseudomonas fluorescens, which had been inoculated to enhance the microbial activities, and monitored by FISH method, showed similar changes in the microcosms, and maintained high levels for several months.