Recent advances in variational inference that make use of an inference or recognition network for training and evaluating deep directed graphical models have advanced well beyond traditional variational inference and Markov chain Monte Carlo methods. These techniques offer higher flexibility with simpler and faster inference; yet training and evaluation still remains a challenge. We propose a method for improving the per-example approximate posterior by iterative refinement, which can provide notable gains in maximizing the variational lower bound of the log likelihood and works with both continuous and discrete latent variables. We evaluate our approach as a method of training and evaluating directed graphical models. We show that, when used for training, iterative refinement improves the variational lower bound and can also improve the log-likelihood over related methods. We also show that iterative refinement can be used to get a better estimate of the log-likelihood in any directed model trained with mean-field inference.
The simultaneous recording of functional magnetic resonance imaging (fMRI) and electroencephalogram (EEG) provides an efficient signal for the high spatiotemporal brain mapping because each modality provides complementary information. The peak detection in the EEG signal measured in the MR scanner is necessary for removal of the ballistocardiac artifact. Especially, it would be affected by the quality of the EKG signal and the variation of the heart beat rate. Therefore, we propose the peak detection method using a K-teager energy operator (K-TEO) as well as further refinement processes in order to detect precise peaks. We applied this technique to the analysis of simulation waves with random noise and abrupt heat beat changes.
Medical images are acquired through diverse imaging systems, with each system employing specific image reconstruction techniques to transform sensor data into images. In MRI, sensor data (i.e., k-space data) is encoded in the frequency domain, and fully sampled k-space data is transformed into an image using the inverse Fourier Transform. However, in efforts to reduce acquisition time, k-space is often subsampled, necessitating a sophisticated image reconstruction method beyond a simple transform. The proposed approach addresses this challenge by training a model to learn domain transform, generating the final image directly from undersampled k-space input. Significantly, to improve the stability of reconstruction from randomly subsampled k-space data, folded images are incorporated as supplementary inputs in the dual-input ETER-net. Moreover, modifications are made to the formation of inputs for the bi-RNN stages to accommodate non-fixed k-space trajectories. Experimental validation, encompassing both regular and irregular sampling trajectories, validates the method's effectiveness. The results demonstrated superior performance, measured by PSNR, SSIM, and VIF, across acceleration factors of 4 and 8. In summary, the dual-input ETER-net emerges as an effective both regular and irregular sampling trajectories, and accommodating diverse acceleration factors.
Recent advances in deep learning have contributed greatly to the field of parallel MR imaging, where a reduced amount of k-space data are acquired to accelerate imaging time. In our previous work, we have proposed a deep learning method to reconstruct MR images directly from k-space data acquired with Cartesian trajectories. However, MRI utilizes various non-Cartesian trajectories, such as radial trajectories, with various numbers of multi-channel RF coils according to the purpose of an MRI scan. Thus, it is important for a reconstruction network to efficiently unfold aliasing artifacts due to undersampling and to combine multi-channel k-space data into single-channel data. In this work, a neural network named 'ETER-net' is utilized to reconstruct an MR image directly from k-space data acquired with Cartesian and non-Cartesian trajectories and multi-channel RF coils. In the proposed image reconstruction network, the domain transform network converts k-space data into a rough image, which is then refined in the following network to reconstruct a final image. We also analyze loss functions including adversarial and perceptual losses to improve the network performance. For experiments, we acquired k-space data at a 3T MRI scanner with Cartesian and radial trajectories to show the learning mechanism of the direct mapping relationship between the k-space and the corresponding image by the proposed network and to demonstrate the practical applications. According to our experiments, the proposed method showed satisfactory performance in reconstructing images from undersampled single- or multi-channel k-space data with reduced image artifacts. In conclusion, the proposed method is a deep-learning-based MR reconstruction network, which can be used as a unified solution for parallel MRI, where k-space data are acquired with various scanning trajectories.
We describe the neural machine translation system of New York University (NYU) and University of Montreal (MILA) for the translation tasks of WMT'16.The main goal of NYU-MILA submission to WMT'16 is to evaluate a new character-level decoding approach in neural machine translation on various language pairs.The proposed neural machine translation system is an attention-based encoder-decoder with a subword-level encoder and a character-level decoder.The decoder of the neural machine translation system does not require explicit segmentation, when characters are used as tokens.The character-level decoding approach provides benefits especially when translating a source language into other morphologically rich languages.This system description paper summarizes and details the experimental procedure described in Chung et al. (2016)
Sodium is crucial for the maintenance of cell physiology, and its regulation of the sodium‐potassium pump has implications for various neurological conditions. The distribution of sodium concentrations in tissue can be quantitatively evaluated by means of sodium MRI ( 23 Na‐MRI). Despite its usefulness in diagnosing particular disease conditions, tissue sodium concentration (TSC) estimated from 23 Na‐MRI can be strongly biased by partial volume effects (PVEs) that are induced by broad point spread functions (PSFs) as well as tissue fraction effects. In this work, we aimed to propose a robust voxel‐wise partial volume correction (PVC) method for 23 Na‐MRI. The method is based on a linear regression (LR) approach to correct for tissue fraction effects, but it utilizes a 3D kernel combined with a modified least trimmed square (3D‐mLTS) method in order to minimize regression‐induced inherent smoothing effects. We acquired 23 Na‐MRI data with conventional Cartesian sampling at 7 T, and spill‐over effects due to the PSF were considered prior to correcting for tissue fraction effects using 3D‐mLTS. In the simulation, we found that the TSCs of gray matter (GM) and white matter (WM) were underestimated by 20% and 11% respectively without correcting tissue fraction effects, but the differences between ground truth and PVE‐corrected data after the PVC using the 3D‐mLTS method were only approximately 0.6% and 0.4% for GM and WM, respectively. The capability of the 3D‐mLTS method was further demonstrated with in vivo 23 Na‐MRI data, showing significantly lower regression errors (ie root mean squared error) as compared with conventional LR methods ( p < 0.001). The results of simulation and in vivo experiments revealed that 3D‐mLTS is superior for determining under‐ or overestimated TSCs while preserving anatomical details. This suggests that the 3D‐mLTS method is well suited for the accurate determination of TSC, especially in small focal lesions associated with pathological conditions.
In this work, we propose a novel recurrent neural network (RNN) architecture. The proposed RNN, gated-feedback RNN (GF-RNN), extends the existing approach of stacking multiple recurrent layers by allowing and controlling signals flowing from upper recurrent layers to lower layers using a global gating unit for each pair of layers. The recurrent signals exchanged between layers are gated adaptively based on the previous hidden states and the current input. We evaluated the proposed GF-RNN with different types of recurrent units, such as tanh, long short-term memory and gated recurrent units, on the tasks of character-level language modeling and Python program evaluation. Our empirical evaluation of different RNN units, revealed that in both tasks, the GF-RNN outperforms the conventional approaches to build deep stacked RNNs. We suggest that the improvement arises because the GF-RNN can adaptively assign different layers to different timescales and layer-to-layer interactions (including the top-down ones which are not usually present in a stacked RNN) by learning to gate these interactions.
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
'%3e%3cg id='b'%3e%3cpath id='a' stroke='%23000' stroke-width='2' d='M-6-25H6m-12 3H6m-12 3H6'/%3e%3cuse xlink:href='%23a' y='44'/%3e%3c/g%3e%3cpath stroke='%23fff' d='M0 17v10'/%3e%3ccircle r='12' fill='%23cd2e3a'/%3e%3cpath fill='%230047a0' d='M0-12A6 6 0 0 0 0 0a6 6 0 0 1 0 12 12 12 0 0 1 0-24z'/%3e%3c/g%3e%3cg transform='rotate(-123.69)'%3e%3cuse xlink:href='%23b'/%3e%3cpath stroke='%23fff' d='M0-23.5v3M0 17v3.5m0 3v3'/%3e%3c/g%3e%3c/svg%3e)
