Purpose Vessel segmentation from volumetric medical images is becoming an essential pre‐step in aiding the diagnosis, guiding the therapy, and patient management for vascular‐related diseases. Deep learning‐based methods have drawn many attentions, but most of them did not fully utilize the multi‐scale spatial information of vessels. To address this shortcoming, we propose a multi‐scale network similar to the well‐known multi‐scale DeepMedic. It also includes a double‐pathway architecture and a class‐balanced loss at the voxel level (MDNet‐Vb) to achieve both the computation efficiency and segmentation accuracy. Methods The proposed network consists two parallel pathways to learn the multi‐scale vessel morphology. Specifically, the pathway with a normal resolution uses three‐dimensional (3D) U‐Net fed with small inputs to learn the local details with relatively small storage and time consumption. The pathway with a low‐resolution employs 3D fully convolutional network (FCN) fed with downsampled large inputs to learn the overall spatial relationships between vessels and adjacent tissues, and the morphological information of large vessels. To cope with the class‐imbalanced issue in vessel segmentation, we propose a class‐balanced loss at the voxel level with uniform sampling strategy. The class‐balanced loss at the voxel level re‐balances the loss function with a coefficient that is inversely proportional to the normalized effective number at the voxel level of each class. The uniform sampling strategy extracts training data by sampling uniformly from two classes in every epoch. Results Our MDNet‐Vb outperforms several state‐of‐the‐art methods including ResNet, DenseNet, 3D U‐Net, V‐Net, and DeepMedic with the highest dice coefficients of 72.91% and 69.32% on cardiac computed tomography angiography (CTA) dataset and cerebral magnetic resonance angiography (MRA) dataset, respectively. Among four different double‐pathway networks, our network (3D U‐Net+3D FCN) not only has the fewest training parameters and shortest training time, but also gets competitive dice coefficients on both the CTA and MRA datasets. Compared with classical losses, our class‐balanced focal loss (FL‐Vb) and dice coefficient loss at the voxel level (Dsc‐Vb) alleviates class imbalanced issue by improving both the sensitivity and dice coefficient on the CTA and MRA datasets. Moreover, simultaneously training on two datasets shows that our method has the highest dice coefficient of 73.06% and 65.40% on CTA and MRA datasets, respectively, outperforming the commonly used methods, such as U‐Net and DeepMedic, which demonstrates the generalization potential of our network for segmenting different blood vessels. Conclusions Our MDNet‐Vb method demonstrates its superiority over other state‐of‐the‐art methods, on both cardiac CTA and cerebral MRA datasets. For the network architecture, the MDNet‐Vb combined the 3D U‐Net and 3D FCN, which dramatically reduces the network parameters yet maintains the segmentation accuracy. The class‐balanced loss at the voxel level further improves accuracy by properly alleviating the class‐imbalanced issue between different classes. In summary, MDNet‐Vb is promising for vessel segmentation from various volumetric medical images.
Purpose Design an efficient CEST scheme for exchange‐dependent images with high contrast‐to‐noise ratio. Theory Reassembled saturation transfer (REST) signals were defined as r.Z = r.Z ref ‐ r.Z CEST and the reassembled exchange‐dependen magnetization transfer ratio r.MTR Rex = r.1/ Z ref ‐ r.1/ Z CEST , utilizing the averages over loosely sampled reference frequency offsets as Z ref and over densely sampled target offsets as Z CEST . Using r.MTR Rex measured under 2 B 1,sat values, exchange rate could be estimated. Methods The REST approach was optimized and assessed quantitatively by simulations for various exchange rates, pool concentration, and water T 1 . In vivo evaluation was performed on ischemic rat brains at 7 Tesla and human brains at 3 Tesla, in comparison with conventional asymmetrical analysis, Lorentzian difference (LD), an MTR Rex_ LD. Results For a broad choice of ranges and numbers, Δr.Z and r.MTR Rex exhibited comparable quantification features with conventional LD and MTR Rex _LD, respectively, when B 1,sat ≤ 1 μT. The subtraction of 2 REST values under distinct B 1,sat values showed linear relationships with exchange rate and obtained immunity to field inhomogeneity and variation in MT and water T 1 . For both rat and human studies, REST images exhibited similar contrast distribution to MTR Rex _LD, with superiority in contrast‐to‐noise ratio and acquisition efficiency. Compared with MTR Rex _LD, 2‐B 1,sat subtraction REST images displayed better resistance to B 1 inhomogeneity, with more specific enhanced regions. They also showed higher signals for amide than for nuclear Overhauser enhancement effect in human brain, presumably reflecting the higher increment from faster‐exchanging species as B 1,sat increased. Conclusion Featuring high contrast‐to‐noise ratio efficiency, REST could be a practical exchange‐dependent approach readily applicable to either retrospective Z‐spectra analysis or perspective 6‐offset acquisition.
Fluorescence Optical Diffusion Tomography (FODT) is considered as one of the most promising ways for non-invasive molecular-based imaging. Many reconstructed approaches to FODT utilize iterative methods for data inversion. However, they are regarded as being time-consuming and far from meeting the real-time imaging requests. In this work, a fast pre-iteration algorithm based on the generalized inverse is established, which divides the image reconstruction into two steps that are off-line pre-iteration and on-line one-step reconstruction. In the pre-iteration step for obtaining the approximation of generalized inverse, a second order iterative format is employed to accelerate the convergence. Simulation based on the linear diffusion model shows that the distribution of fluorescent yield can be well estimated by this algorithm with second-order iteration. And the reconstructed speed is remarkably increased. Time-efficiency of this method will potentially promote the development of real-time imaging and the dynamic monitoring of molecular activity.
We introduce a novel technique for seismic wave extrapolation in time. The technique involves cascading a Fourier transform operator and a finite-difference operator to form a chain operator: Fourier finite differences (FFD). We derive the FFD operator from a pseudoanalytical solution of the acoustic wave equation. Two-dimensional synthetic examples demonstrate that the FFD operator can have high accuracy and stability in complex-velocity media. Applying the FFD method to the anisotropic case overcomes some disadvantages of other methods, such as the coupling of qP-waves and qSV-waves. The FFD method can be applied to enhance accuracy and stability of seismic imaging by reverse time migration.
Conjugate gradient method is verified to be efficient for nonlinear optimization problems of large-dimension data. In this paper, a penalized linear and nonlinear combined conjugate gradient method for the reconstruction of fluorescence molecular tomography (FMT) is presented. The algorithm combines the linear conjugate gradient method and the nonlinear conjugate gradient method together based on a restart strategy, in order to take advantage of the two kinds of conjugate gradient methods and compensate for the disadvantages. A quadratic penalty method is adopted to gain a nonnegative constraint and reduce the illposedness of the problem. Simulation studies show that the presented algorithm is accurate, stable, and fast. It has a better performance than the conventional conjugate gradient-based reconstruction algorithms. It offers an effective approach to reconstruct fluorochrome information for FMT.
Genetically engineered reporters have revolutionized the understanding of many biological processes. MRI-based reporter genes can dramatically improve our ability to monitor dynamic gene expression and allow coregistration of subcellular genetic information with high-resolution anatomical images. We have developed a biocompatible MRI reporter gene based on a human gene, the human protamine-1 (hPRM1). The arginine-rich hPRM1 (47% arginine residues) generates high MRI contrast based on the chemical exchange saturation transfer (CEST) contrast mechanism. The 51 amino acid-long hPRM1 protein was fully synthesized using microwave-assisted technology, and the CEST characteristics of this protein were compared to other CEST-based contrast agents. Both bacterial and human cells were engineered to express an optimized hPRM1 gene and showed higher CEST contrast compared to controls. Live cells expressing the hPRM1 reporter gene, and embedded in three-dimensional culture, also generated higher CEST contrast compared to wild-type live cells.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)