Search

In the area of pedestrian trajectory prediction, the hybrid structures of temporal feature extractor or spatial feature extractor have paved the way for the precise prediction model, and they are in larger and larger scale. Learning of specific feature encoding model not only influenced by the structure of the network, but also by the learning manners such as supervised learning and unsupervised learning. Previous works concentrated on more comprehensive encoders and more delicate designs of feature extractors. However, the mutual influence factors from the neighbour pedestrians associate with the distance to the centre pedestrian seldomly noticed. Most of the existed feature extractors in prediction models trained in the way of supervised learning other than unsupervised manners caused the problem that the extracted features are always handcrafted without the natural distinction of obscure situations. The graph contrastive accelerating encoder is proposed, which accelerates the pedestrian trajectory prediction training process of the state of the art method of spatio‐temporal graph transformer networks. Employing the unsupervised contrastive learning process and the graph of neighbours representing distance affection of nearest and farthest pedestrian to the centre pedestrian, the graph contrastive accelerating encoder significantly shrinked the training time. Holding the final performance on to state of the art level, the proposed method let the lowest pedestrian trajectory prediction error show up in the obviously earlier training steps.

Purpose In neonatal brain magnetic resonance image (MRI) segmentation, the model we trained on the training set (source domain) often performs poorly in clinical practice (target domain). As the label of target-domain images is unavailable, this cross-domain segmentation needs unsupervised domain adaptation (UDA) to make the model adapt to the target domain. However, the shape and intensity distribution of neonatal brain MRI images across the domains are largely different from adults'. Current UDA methods aim to make synthesized images similar to the target domain as a whole. But it is impossible to synthesize images with intraclass similarity because of the regional misalignment caused by the cross-domain difference. This will result in generating intraclassly incorrect intensity information from target-domain images. To address this issue, we propose an IAS-NET (joint intraclassly adaptive generative adversarial network (GAN) (IA-NET) and segmentation) framework to bridge the gap between the two domains for intraclass alignment. Methods Our proposed IAS-NET is an elegant learning framework that transfers the appearance of images across the domains from both image and feature perspectives. It consists of the proposed IA-NET and a segmentation network (S-NET). The proposed IA-NET is a GAN-based adaptive network that contains one generator (including two encoders and one shared decoder) and four discriminators for cross-domain transfer. The two encoders are implemented to extract original image, mean, and variance features from source and target domains. The proposed local adaptive instance normalization algorithm is used to perform intraclass feature alignment to the target domain in the feature-map level. S-NET is a U-net structure network that is used to provide semantic constraint by a segmentation loss for the training of IA-NET. Meanwhile, it offers pseudo-label images for calculating intraclass features of the target domain. Source code (in Tensorflow) is available at https://github.com/lb-whu/RAS-NET/. Results Extensive experiments are carried out on two different data sets (NeoBrainS12 and dHCP), respectively. There exist great differences in the shape, size, and intensity distribution of magnetic resonance (MR) images in the two databases. Compared to baseline, we improve the average dice score of all tissues on NeoBrains12 by 6% through adaptive training with unlabeled dHCP images. Besides, we also conduct experiments on dHCP and improved the average dice score by 4%. The quantitative analysis of the mean and variance of the synthesized images shows that the synthesized image by the proposed is closer to the target domain both in the full brain or within each class than that of the compared methods. Conclusions In this paper, the proposed IAS-NET can improve the performance of the S-NET effectively by its intraclass feature alignment in the target domain. Compared to the current UDA methods, the synthesized images by IAS-NET are more intraclassly similar to the target domain for neonatal brain MR images. Therefore, it achieves state-of-the-art results in the compared UDA models for the segmentation task.

This research focuses on frequency‐based lossless new encoding technique with low computation and low memory scheme. This encoding technique doesn't require, any table similar to Huffman and Golomb Rice encoder, and doesn't take high computation time like an arithmetic encoder. For better comparative results, nearly 200 standard images were tested with the proposed encoder and their results were compared with the standard encoders. Based on the analysis of results reported in the experimental part, the proposed scheme in Discrete Cosine Transformation (DCT) achieves a high Compression Ratio (CR) of 97.68 %, Bits Per Pixel (BPP) of 0.19, Peak Signal To Noise Ratio (PSNR) of 27.62 dB for pepper image, CR of 98.65%, BPP of 0.11 and PSNR of 29.24 dB for house image, CR of 98.58 %, BPP of 0.11, PSNR of 30.59 dB for pepper image, CR of 98.37%, BPP of 0.13, PSNR of 31.37dB for house image in Discrete Wavelet Transformation (DWT) than compared to other encoders. It consumes less computation time to attain superior CR with a better quality of reconstructed picture than DCT based Huffman, Golomb‐ Rice (GR) coder and Low Complexity, Low memory Entropy Coder (LLEC), DWT based Arithmetic, GR code, LLEC.

Purpose The developments of PET/CT and PET/MR scanners provide opportunities for improving PET image quality by using anatomical information. In this paper, we propose a novel co-learning three-dimensional (3D) convolutional neural network (CNN) to extract modality-specific features from PET/CT image pairs and integrate complementary features into an iterative reconstruction framework to improve PET image reconstruction. Methods We used a pretrained deep neural network to represent PET images. The network was trained using low-count PET and CT image pairs as inputs and high-count PET images as labels. This network was then incorporated into a constrained maximum likelihood framework to regularize PET image reconstruction. Two different network structures were investigated for the integration of anatomical information from CT images. One was a multichannel CNN, which treated PET and CT volumes as separate channels of the input. The other one was multibranch CNN, which implemented separate encoders for PET and CT images to extract latent features and fed the combined latent features into a decoder. Using computer-based Monte Carlo simulations and two real patient datasets, the proposed method has been compared with existing methods, including the maximum likelihood expectation maximization (MLEM) reconstruction, a kernel-based reconstruction and a CNN-based deep penalty method with and without anatomical guidance. Results Reconstructed images showed that the proposed constrained ML reconstruction approach produced higher quality images than the competing methods. The tumors in the lung region have higher contrast in the proposed constrained ML reconstruction than in the CNN-based deep penalty reconstruction. The image quality was further improved by incorporating the anatomical information. Moreover, the liver standard deviation was lower in the proposed approach than all the competing methods at a matched lesion contrast. Conclusions The supervised co-learning strategy can improve the performance of constrained maximum likelihood reconstruction. Compared with existing techniques, the proposed method produced a better lesion contrast versus background standard deviation trade-off curve, which can potentially improve lesion detection.

PURPOSE Diffuse large B-cell lymphoma (DLBCL) is an aggressive type of lymphoma with high mortality and poor prognosis that especially has a high incidence in Asia. Accurate segmentation of DLBCL lesions is crucial for clinical radiation therapy. However, manual delineation of DLBCL lesions is tedious and time-consuming. Automatic segmentation provides an alternative solution but is difficult for diffuse lesions without the sufficient utilization of multimodality information. Our work is the first study focusing on positron emission tomography and computed tomography (PET-CT) feature fusion for the DLBCL segmentation issue. We aim to improve the fusion performance of complementary information contained in PET-CT imaging with a hybrid learning module in the supervised convolutional neural network. METHODS First, two encoder branches extract single-modality features, respectively. Next, the hybrid learning component utilizes them to generate spatial fusion maps which can quantify the contribution of complementary information. Such feature fusion maps are then concatenated with specific-modality (i.e., PET and CT) feature maps to obtain a representation of the final fused feature maps in different scales. Finally, the reconstruction part of our network creates a prediction map of DLBCL lesions by integrating and up-sampling the final fused feature maps from encoder blocks in different scales. RESULTS The ability of our method was evaluated to detect foreground and segment lesions in three independent body regions (nasopharynx, chest, and abdomen) of a set of 45 PET-CT scans. Extensive ablation experiments compared our method to four baseline techniques for multimodality fusion (input-level (IL) fusion, multichannel (MC) strategy, multibranch (MB) strategy, and quantitative weighting (QW) fusion). The results showed that our method achieved a high detection accuracy (99.63% in the nasopharynx, 99.51% in the chest, and 99.21% in the abdomen) and had the superiority in segmentation performance with the mean dice similarity coefficient (DSC) of 73.03% and the modified Hausdorff distance (MHD) of 4.39 mm, when compared with the baselines (DSC: IL: 53.08%, MC: 63.59%, MB: 69.98%, and QW: 72.19%; MHD: IL: 12.16 mm, MC: 6.46 mm, MB: 4.83 mm, and QW: 4.89 mm). CONCLUSIONS A promising segmentation method has been proposed for the challenging DLBCL lesions in PET-CT images, which improves the understanding of complementary information by feature fusion and may guide clinical radiotherapy. The statistically significant analysis based on P-value calculation has indicated a degree of significant difference between our proposed method and other baselines (almost metrics: P

Purpose Precise segmentation of clinical target volume (CTV) is the key to stomach cancer radiotherapy. We proposed a novel stochastic width-deep neural network (SW-DNN) for better automatically contouring stomach CTV. Methods Stochastic width-deep neural network was an end-to-end approach, of which the core component was a novel SW mechanism that employed shortcut connections between the encoder and decoder in a random manner, and thus the width of the SW-DNN was stochastically adjustable to obtain improved segmentation results. In total, 150 stomach cancer patient computed tomography (CT) cases with the corresponding CTV labels were collected and used to train and evaluate the SW-DNN. Three common quantitative measures: true positive volume fraction (TPVF), positive predictive value (PPV), and Dice similarity coefficient (DSC) were used to evaluate the segmentation accuracy. Results Clinical target volumes calculated by SW-DNN had significant quantitative advantages over three state-of-the-art methods. The average DSC value of SW-DNN was 2.1%, 2.8%, and 3.6% higher than that of three state-of-the-art methods. The average DSC, TPVF, and PPV values of SW-DNN were 2.1%, 4.0%, and 0.3% higher than that of the corresponding constant width DNN. Conclusions Stochastic width-deep neural network provided better performance for contouring stomach cancer CTV accurately and efficiently. It is a promising solution in clinical radiotherapy planning for stomach cancer.

In this study, we proposed a deep learning algorithm (PATCHUNET) to suppress random noise and preserve the coherent seismic signal. The input data are divided into several patches, and each patch is encoded to extract the meaningful features. Following this, the extracted features are decompressed to retrieve the seismic signal. Skip connections are used between the encoder and decoder parts, allowing the proposed algorithm to extract high‐order features without losing important information. Besides, dropout layers are used as regularization layers. The dropout layers preserve the most meaningful features belonging to the seismic signal and discard the remaining features. The proposed algorithm is an unsupervised approach that does not require prior information about the clean signal. The input patches are divided into 80% for training and 20% for testing. However, it is interesting to find that the proposed algorithm can be trained with only 30% of the input patches with an effective denoising performance. Four synthetic and four field examples are used to evaluate the proposed algorithm performance, and compared to the f−x deconvolution and the f−x singular spectrum analysis. The results indicate the ability of the proposed algorithm in attenuating the random noise and preserving the seismic signal effectively despite the existence of a large amount of random noise, for example, when the input signal‐to‐noise ratio is as low as −14.2 dB.

PURPOSE Ultrasound image segmentation is a challenging task due to a low signal-to-noise ratio and poor image quality. Although several approaches based on the convolutional neural network (CNN) have been applied to ultrasound image segmentation, they have weak generalization ability. We propose an end-to-end, multiple-channel and atrous CNN designed to extract a greater amount of semantic information for segmentation of ultrasound images. METHOD A multiple-channel and atrous convolution network is developed, referred to as MA-Net. Similar to U-Net, MA-Net is based on an encoder-decoder architecture and includes five modules: the encoder, atrous convolution, pyramid pooling, decoder, and residual skip pathway modules. In the encoder module, we aim to capture more information with multiple-channel convolution and use large kernel convolution instead of small filters in each convolution operation. In the last layer, atrous convolution and pyramid pooling are used to extract multi-scale features. The architecture of the decoder is similar to that of the encoder module, except that up-sampling is used instead of down-sampling. Furthermore, the residual skip pathway module connects the subnetworks of the encoder and decoder to optimize learning from the deeper layer and improve the accuracy of segmentation. During the learning process, we adopt multi-task learning to enhance segmentation performance. Five types of datasets are used in our experiments. Because the original training data are limited, we apply data augmentation (e.g., horizontal and vertical flipping, random rotations, and random scaling) to our training data. We use the Dice score, precision, recall, Hausdorff distance (HD), average symmetric surface distance (ASD), and root mean square symmetric surface distance (RMSD) as the metrics for segmentation evaluation. Meanwhile, Friedman test was performed as the nonparametric statistical analysis to evaluate the algorithms. RESULTS For the datasets of brachia plexus (BP), fetal head, and lymph node segmentations, MA-Net achieved average Dice scores of 0.776, 0.973, and 0.858, respectively; with average precisions of 0.787, 0.968, and 0.854, respectively; average recalls of 0.788, 0.978, and 0.885, respectively; average HDs (mm) of 13.591, 10.924, and 19.245, respectively; average ASDs (mm) of 4.822, 4.152, and 4.312, respectively; and average RMSDs (mm) of 4.979, 4.161, and 4.930, respectively. Compared with U-Net, U-Net++, M-Net, and Dilated U-Net, the average performance of the MA-Net increased by approximately 5.68%, 2.85%, 6.59%, 36.03%, 23.64%, and 31.71% for Dice, precision, recall, HD, ASD, and RMSD, respectively. Moreover, we verified the generalization of MA-Net segmentation to lower grade brain glioma MRI and lung CT images. In addition, the MA-Net achieved the highest mean rank in the Friedman test. CONCLUSION The proposed MA-Net accurately segments ultrasound images with high generalization, and therefore, it offers a useful tool for diagnostic application in ultrasound images.

In this study, the authors will show that coding the positions of the symbols, instead of their values, can be a good way to implement efficient Slepian–Wolf (SW) coding and can reduce the complexity of both the encoder and the decoder. The authors will also propose a practical distributed video coding (DVC) system that exploits this idea. This system will use binary maps to indicate the positions of the most probable symbols, instead of separating them into bitplanes. Simulations show that this position-based SW coding allows a simple and more efficient DVC system with improved rate-distortion performance, compared to the bitplane-based DVC system that uses the same side information. The memory requirements at the encoder are reduced by about 50% and the number of channel coding operations is also reduced. This DVC system also allows to use an easy way to control quantisation, reduces the decoding latency, and allows fast parallel decoding.

Purpose Automated segmentation could improve the efficiency of modeling-based pelvic organ prolapse (POP) evaluations. However, segmentation performance is limited by the blurry soft tissue boundaries. In this study, we aimed to present a hybrid solution for uterus, rectum, bladder, and levator ani muscle segmentation by combining a convolutional neural network (CNN) and a level set method. Methods We used 24 sagittal pelvic floor magnetic resonance (MR) series from six anterior vaginal prolapse and six posterior vaginal prolapse subjects (a total 528 MR images). The stress MR images were performed both at rest and at maximal Valsalva. We assigned 264 images for training, 132 images for validation, and 132 images for testing. A CNN was designed by introducing a multi-resolution features pyramid module (MRFP) into an encoder-decoder model. Depth separable convolution and pretraining were used to improve model convergence. Multiclass cross entropy loss and multiclass Dice loss were used for model training. The dice similarity coefficient (DSC) and average surface distance (ASD) were used for evaluating the segmentation results. To prove the effectiveness of our model, we compared it with advanced segmentation methods including Deeplabv3+, U-Net, and FCN-8s. The ablation study was designed to quantify the contributions of MRFP, the encoder network, and pretraining. Besides, we investigated the working mechanism of MRFP in the segmentation network by comparing our model with three of its variants. Finally, the level set method was used to improve the CNN model further. Results Dice loss showed better segmentation performance than multiclass cross entropy loss. MRFP was efficacious for different encoder networks. With MRFP, U-Net and U-Net-X (X represents Xception encoder network) have improved the DSC, on average by 6.8 and 5.3 points. Compared with different CNN models, our model achieved the highest average DSC of 65.6 points and the lowest average ASD of 2.9 mm. With the level set method, the DSC of our model improved to 69.4 points. Conclusions MRFP proved to be effective in addressing the blurry soft tissue boundary problem on pelvic floor MR images. A hybrid solution based on CNN and level set method was presented for pelvic organ segmentation both at rest and at maximal Valsalva; with this method, we achieved state-of-the-art results.

Purpose Accurate and robust segmentation of the prostate from magnetic resonance (MR) images is extensively applied in many clinical applications in prostate cancer diagnosis and treatment. The purpose of this study is the development of a robust interactive segmentation method for accurate segmentation of the prostate from MR images. Methods We propose an interactive segmentation method based on a graph convolutional network (GCN) to refine the automatically segmented results. An atrous multiscale convolutional neural network (CNN) encoder is proposed to learn representative features to obtain accurate segmentations. Based on the multiscale feature, a GCN block is presented to predict the prostate contour in both automatic and interactive manners. To preserve the prostate boundary details and effectively train the GCN, a contour matching loss is proposed. The performance of the proposed algorithm was evaluated on 41 in-house MR subjects and 30 PROMISE12 test subjects. Result The proposed method yields mean Dice similarity coefficients of 93.8 ± 1.2% and 94.4 ± 1.0% on our in-house and PROMISE12 datasets, respectively. The experimental results show that the proposed method outperforms several state-of-the-art segmentation methods. Conclusion The proposed interactive segmentation method based on the GCN can accurately segment the prostate from MR images. Our method has a variety of applications in prostate cancer imaging.

The suppression of motion artefacts from MR images is a challenging task. The purpose of this paper was to develop a standalone novel technique to suppress motion artefacts in MR images using a data-driven deep learning approach. A simulation framework was developed to generate motion-corrupted images from motion-free images using randomly generated motion profiles. An Inception-ResNet deep learning network architecture was used as the encoder and was augmented with a stack of convolution and upsampling layers to form an encoder-decoder network. The network was trained on simulated motion-corrupted images to identify and suppress those artefacts attributable to motion. The network was validated on unseen simulated datasets and real-world experimental motion-corrupted in vivo brain datasets. The trained network was able to suppress the motion artefacts in the reconstructed images, and the mean structural similarity (SSIM) increased from 0.9058 to 0.9338. The network was also able to suppress the motion artefacts from the real-world experimental dataset, and the mean SSIM increased from 0.8671 to 0.9145. The motion correction of the experimental datasets demonstrated the effectiveness of the motion simulation generation process. The proposed method successfully removed motion artefacts and outperformed an iterative entropy minimization method in terms of the SSIM index and normalized root mean squared error, which were 5-10% better for the proposed method. In conclusion, a novel, data-driven motion correction technique has been developed that can suppress motion artefacts from motion-corrupted MR images. The proposed technique is a standalone, post-processing method that does not interfere with data acquisition or reconstruction parameters, thus making it suitable for routine clinical practice.

This chapter is devoted to the use of various neural networks and related learning algorithms in the channel coding (encoder and decoder) of wireless communications and networking. Due to its powerful nonlinear mapping and distributed processing capability, neural network–based machine learning technology could offer a more powerful channel coding solution than conventional approaches in many aspects including coding performance, computational complexity, power consumption, and processing latency. The neural networks discussed in the chapter mainly include deep neural networks (DNNs), convolutional neural networks (CNNs), and recurrent neural networks (RNNs). The chapter first focuses on the background information of channel coding and deep learning, together with the motivation for the use of machine learning in channel coding. It introduces the channel coding schemes with DNN, CNN, and RNN networks, respectively, and finally discusses potential coding/decoding performance, computational complexity, power consumption, and processing delay.

Purpose To automatically and efficiently segment the lesion area of the colonoscopy polyp image, a polyp segmentation method has been presented. Methods An ensemble model of pretrained convolutional neural networks was proposed, using Unet-VGG, SegNet-VGG, and PSPNet. Firstly, the Unet-VGG is obtained by the first 10 layers of VGG16 as the contraction path of the left half of the Unet. Then, the SegNet-VGG is acquired by fine-tuned transfer learning VGG16, using the first 13 layers of VGG16 as the encoder of the SegNet and combined the original decoder of the SegNet. By adjusting the input size of the Unet-VGG, SegNet-VGG, and PSPNet, the preprocessed data can be correctly fed to the three network models. The three models are used as the basic trainer to train and segment the datasets. Based on the ensemble learning algorithm, the weight voting method is used to ensemble the segmentation results corresponding to single basic trainer. Results Both IoU and DICE similarity score were used to evaluate the segmentation quality for cvc300 with 300 images, CVC-ClinicDB with 612 images, and ETIS-LaribPolypDB with 196 images. From the experimental results, the IoU and DICE obtained by the proposed method for the cvc300 datasets can reach up to 96.16% and 98.04%, respectively, the IoU and DICE for the CVC-ClinicDB datasets can reach up to 96.66% and 98.30%, respectively, whereas the IoU and DICE for the ETIS-LaribPolypDB datasets can reach up to 96.95% and 98.45%, respectively. Evaluation of the IoU and DICE in our methods shows higher accuracy than previous methods. Conclusions The experimental results show that the proposed method improved correspondingly in IoU and DICE compared to a single basic trainer. The range of improvement is 1.98%-6.38%. The proposed ensemble learning succeeds in automatic polyp segmentation, which potentially helps to establish more polyp datasets.

Purpose The superior soft-tissue contrast achieved using magnetic resonance imaging (MRI) compared to x-ray computed tomography (CT) has led to the popularization of MRI-guided radiation therapy (MR-IGRT), especially in recent years with the advent of first and second generation MRI-based therapy delivery systems for MR-IGRT. The expanding use of these systems is driving interest in MRI-only RT workflows in which MRI is the sole imaging modality used for treatment planning and dose calculations. To enable such a workflow, synthetic CT (sCT) data must be generated based on a patient's MRI data so that dose calculations may be performed using the electron density information derived from CT images. In this study, we propose a novel deep spatial pyramid convolutional framework for the MRI-to-CT image-to-image translation task and compare its performance to the well established U-Net architecture in a generative adversarial network (GAN) framework. Methods Our proposed framework utilizes atrous convolution in a method named atrous spatial pyramid pooling (ASPP) to significantly reduce the total number of parameters required to describe the model while effectively capturing rich, multi-scale structural information in a manner that is not possible in the conventional framework. The proposed framework consists of a generative model composed of stacked encoders and decoders separated by the ASPP module, where atrous convolution is applied at increasing rates in parallel to encode large-scale features. The performance of the proposed method is compared to that of the conventional GAN framework in terms of the time required to train the model and the image quality of the generated sCT as measured by the root mean square error (RMSE), structural similarity index (SSIM), and peak signal-to-noise ratio (PSNR) depending on the size of the training data set. Dose calculations based on sCT data generated using the proposed architecture are also compared to clinical plans to evaluate the dosimetric accuracy of the method. Results Significant reductions in training time and improvements in image quality are observed at every training data set size when the proposed framework is adopted instead of the conventional framework. Over 1042 test images, values of 17.7 ± 4.3 HU, 0.9995 ± 0.0003, and 71.7 ± 2.3 are observed for the RMSE, SSIM, and PSNR metrics, respectively. Dose distributions calculated based on sCT data generated using the proposed framework demonstrate passing rates equal to or greater than 98% using the 3D gamma index with a 2%/2 mm criterion. Conclusions The deep spatial pyramid convolutional framework proposed here demonstrates improved performance compared to the conventional GAN framework that has been applied to the image-to-image translation task of sCT generation. Adopting the method is a first step toward an MRI-only RT workflow that enables widespread clinical applications for MR-IGRT including online adaptive therapy.

Respiratory‐gated volumetric modulated arc therapy (gated VMAT) involves further complexities to the dose delivery process because the gantry rotation must repeatedly stop and restart according to the gating signals. In previous studies, the gantry rotation performances were evaluated by the difference between the plan and the machine log. However, several reports pointed out that log analysis does not sufficiently replicate the machine performance. In this report, a measurement‐based quality assurance of the relation between the gantry angle and gate‐on or gate‐off using triggered kilovoltage imaging and a cylinder phantom with 16 ball bearings is proposed. For the analysis, an in‐house program that estimates and corrects the phantom offset was developed. The gantry angle in static and gated arc delivery was compared between the machine log and the proposed method. The gantry was set every 5 deg through its full motion range in static delivery, and rotated at three speeds (2, 4 and 6 deg s‐1) with different gating intervals (1.5 or 3.0 s) in gated arc delivery. The mean and standard deviation of the angular differences between the log and the proposed method was −0.05 deg ± 0.12 deg in static delivery. The mean of the angular difference was within ±0.10 deg and the largest difference was 0.41 deg in gated arc delivery. The log records the output of the encoder so that miscalibration and mechanical sagging will be disregarded. However, the proposed method will help the users to detect the mechanical issues due to the repeated gantry stops and restarts in gated VMAT.

The two‐user multiple‐access relay channel (MARC) is investigated in the presence of side information (SI) available only at the informed transmitter in the present article. First, for the...

OBJECTIVE: To demonstrate the potential of machine learning and capability of natural language processing (NLP) to predict disposition of patients based on triage notes in the ED. METHODS: A retrospective cohort of ED triage notes from St Vincent's Hospital (Melbourne) was used to develop a deep-learning algorithm that predicts patient disposition. Bidirectional Encoder Representations from Transformers, a recent language representation model developed by Google, was utilised for NLP. Eighty percent of the dataset was used for training the model and 20% was used to test the algorithm performance. Ktrain library, a wrapper for TensorFlow Keras, was employed to develop the model. RESULTS: The accuracy of the algorithm was 83% and the area under the curve was 0.88. Sensitivity, specificity, precision and F1-score of the algorithm were 72%, 86%, 56% and 63%, respectively. CONCLUSION: Machine learning and NLP can be together applied to the ED triage note to predict patient disposition with a high level of accuracy. The algorithm can potentially assist ED clinicians in early identification of patients requiring admission by mitigating the cognitive load, thus optimises resource allocation in EDs.

A new polar coding scheme for higher order modulation is presented. The proposed scheme is based on multi-level coding (MLC) with natural labeling, where the bit-level corresponding to the sign-bit is generated in dependence on the previous bit-levels, such that the modulated symbols are distributed according to a target non-uniform distribution resulting in a shaping gain. This is realized by replacing the polar encoder in the sign-bit level by a successive cancellation (SC) decoder, such that the overall complexity is not increased compared to a conventional MLC scheme with polar codes. Numerical simulations show significant performance improvements of the proposed approach compared to conventional transmission schemes with uniform symbol distribution. By using more complex decoders (e.g. SC list decoder), the proposed scheme outperforms Gallager's random coding bound.

This study focuses on a method for sequential data augmentation in order to alleviate data sparseness problems. Specifically, we present corpus expansion techniques for enhancing the coverage of a language model. Recent recurrent neural network studies show that a seq2seq model can be applied for addressing language generation issues; it has the ability to generate new sentences from given input sentences. We present a method of corpus expansion using a sentence-chain based seq2seq model. For training the seq2seq model, sentence chains are used as triples. The first two sentences in a triple are used for the encoder of the seq2seq model, while the last sentence becomes a target sequence for the decoder. Using only internal resources, evaluation results show an improvement of approximately 7.6% relative perplexity over a baseline language model of Korean text. Additionally, from a comparison with a previous study, the sentence chain approach reduces the size of the training data by 38.4% while generating 1.4-times the number of n-grams with superior performance for English text.

The false data injection attack can tamper with the measurement information collected by the Supervisory Control and Data Acquisition, threat the security of state estimation. Therefore, the analysing methods and detection methods of false data injection attacks have important theoretical and practical significance for ensuring the safe operation of smart grids. This study uses the invertible automatic encoder to reduce the dimension of the original data and uses the long-short-term memory to detect false data injection attacks. This method overcomes the shortcomings of shallow algorithm and traditional machine learning algorithm for power big data training and avoiding the problems of gradient explosion and gradient disappearing during training. Finally, the authors perform a large number of experiments in the IEEE 118-node test system and the 300-node test system and verify the effectiveness of the proposed method.

Three speed estimation methodologies: pulse count, elapsed time, and constant sample-time digital tachometer (CSDT) methods, used with optical incremental encoders, are implemented and compared in a closed-loop system setup using a specialised Analog Devices CM403f DSP. Dedicated hardware for quadrature encoders on the DSP facilitates the implementation of advanced speed estimation methods, such as the CSDT, by minimising the delay resulting from software intervention. The error of each methodology is evaluated theoretically and experimentally. The CSDT has limited error over all its speed range. The error of the elapsed time method increases as the speed increases. On the contrary, the relative error of the pulse count method (which is often substantial) decreases with increased speed. Several filtering techniques can be used to reduce the error of the elapsed time method; here an oversampling technique is implemented with that aim.

This paper presents a backlight control algorithm for liquid crystal display devices, which considers the human visual properties that we are usually attracted to the saliencies in a scene. The saliency means regions or objects that have different contrast or color from the surrounding, and thus attract our attention, which can be measured in various ways from each pixel value and its relation with the surrounding ones. Hence, by keeping or boosting the backlight of salient regions while suppressing in others, the quality of salient regions and overall contrast are enhanced. In addition, power can be saved by backlight dimming in non-salient regions, without loss of overall quality in terms of human visual perception. In this backlight control process, the amount of energy consumption is regulated so that the proposed method consumes less than or equal energy as before, by developing a power management algorithm based on the bit rate control strategies of MPEG2 video encoders. Precisely, the energy consumption in each backlight unit is controlled by a parameter, so that the sum of energies of overall backlight units is kept within a limit.

This paper presents a framework that encodes saliency information in a video sequence at the encoder and uses this information for backlight scaling at the decoder. The saliency region in a scene is extracted and encoded as a supplementary information of H.264/AVC codec in our scheme. The saliency information decoded at the receiver is used to highlight the saliencies in the scene, which enhances the dynamic range of output images.

Computational spectroscopic instruments with Broadband Encoding Stochastic (BEST) filters allow the reconstruction of the spectrum at high precision with only a few filters. However, conventional design manners of BEST filters are often heuristic and may fail to fully explore the encoding potential of BEST filters. The Parameter Constrained Spectral Encoder and Decoder (PCSED) - a neural network-based framework is presented for the design of BEST filters in spectroscopic instruments. By incorporating the target spectral response definition and the optical design procedures comprehensively, PCSED links the mathematical optimum and practical limits confined by available fabrication techniques. Benefiting from this, the BEST-filter-based spectral camera present a higher reconstruction accuracy with up to 30 times' enhancement and a better tolerance on fabrication errors. The generalizability of PCSED is validated in designing metasurface- and interference-thin-film-based BEST filters.

In this paper, a novel rate control scheme for low delay video coding of High Efficiency Video Coding (HEVC) is proposed. The proposed scheme is developed with the consideration of the new temporal prediction structure adopted into HEVC. In the proposed scheme, the relationship between bit rate and quantization step is exploited firstly to formulate an accurate quadratic ratequantization (R-Q) model. Secondly, a method of determining the quantization parameters (QPs) for the first frames within group of pictures is proposed. Thirdly, an accurate frame level bit allocation method is proposed for HEVC. Finally, based on the proposed R-Q model and the target bit allocated for the frame, the QPs are predicted for coding tree units by using rate-distortion optimization. Compared to three state-of-the-art rate control schemes, experimental results show that the proposed rate control scheme can increase Bjontegaard delta peak signal-to-noise ratio (BD-PSNR) by 0.65 dB and 0.09 dB on average than JCTVC-I0094 and JCTVC-M0036 which have been implemented in HEVC test model encoder, respectively, and achieve a similar rate-distortion performance to Wang’s scheme, and furthermore obtain the smallest bit rate mismatch error among all the schemes.

As light emitting diodes (LEDs) of fast response time are wide-spreading in a variety of applications such as displays, lighting, traffic signals, and automobiles, many researches on transferring the data with the visible lights of LEDs have been conducted. This paper describes such a visible light communication system that supports the audio applications at the short distance up to 80 cm and the data bandwidth of 32 kHz. Especially, a simple clock data recovery (CDR) is used to achieve the optimum sampling and compensate for the frequency offset between transmitter and receiver. The proposed CDR obtains three samples per symbol and then adjusts the sampling point properly with the pattern of three samples over the frequency offset range of ±1.5% without changing the operating frequency of a receiver side. In the prototype system, a transmitter is composed of audio encoder, field programmable gate array, LED driver, and LEDs and a receiver consists of photodiode, comparator, field programmable gate array, and audio decoder. At the distance of 50 cm and the data rate of 32 kbps, the measured bit error rates are 0.101 and less than 10−8 without and with CDR, respectively.

Variable optical signal delay devices composed of cascaded double ring add/drop filters and delay waveguides are designed and fabricated. In the measurement, variable time delay is observed in proportion to the number of delay lines. The experimental results exhibit the potential for the fabricated delay device to be applied to an optical code division multiple access encoder/decoder. © 2015 Wiley Periodicals, Inc. Microwave Opt Technol Lett 57:1508–1511, 2015

The adaptive multi-rate wideband (AMR-WB) speech codec with a sampling rate of 16 kHz is one of the speech codecs applied to 4G mobile communication systems as a way to remarkably improve the speech quality of a smartphone. However, a major drawback is that the vector quantization of the immittance spectral frequency (ISF) coefficients takes up the second largest of the total computational load of the AMR-WB encoder. In other words, the speech quality is improved at the cost of high battery power consumption. Accordingly, this letter presents a triangular inequality elimination algorithm equipped with a dynamic and an intersection mechanism, named DI-TIE, as a means to considerably improve the performance of ISF coefficient quantization in AMR-WB speech codec. Both mechanisms are designed in a way that recursively enhances the performance of TIE algorithm. This proposal is demonstrated by experimental results as a superior search algorithm relative to a conventional TIE and a multiple TIE (MTIE) approaches. With a full search algorithm as a benchmark for search load comparison, this work provides a search load reduction of more than 77%, a figure far beyond the 36% in the TIE and 49% in the MTIE approach. © 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

Three dimensional 3D medical images possess some specific characteristics that should be utilized by an efficient compression scheme. In this article, one such compression scheme for volumetric 3D medical image data is presented. Two processes involved in this scheme are decorrelation and encoding. Decorrelation of the 3D data is realized through 3D multiwavelet transform with apt prefiltering so as to give good representation of the image which could be exploited by the encoder. Encoding is done through proposed Block Coding Algorithm, which is embedded, block based, and wavelet transform coding algorithm without maintaining any list structures. The idea behind this algorithm is to sort the 3D transform coefficients in to a 1D array with respect to declining thresholds and to use state table to keep track of the blocks and coefficients that has been coded. In the experiment conducted on various 3D magnetic resonance and computed tomography images of human brain with multiwavelets such as Geronimo-Hardin-Massopust, Chui-Lian, and orthogonal symmetric/antisymmetric SA4, efficiency of the proposed scheme was weighed against the state of art encoders such as 3D Set Partitioning in Hierarchical Trees, 2D Set Partitioned Embedded BloCK Coder, and No List SPIHT. Attributes used for performance measurements are peak signal to noise ratio, bit rate, and structural similarity index of reconstructed image with respect to original image. © 2014 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 24, 182-192, 2014

Summary Distributed dynamic average consensus is investigated under quantized communication data. We use a uniform quantizer with constant quantization step-size to deal with the saturation caused by the dynamic consensus error and propose a communication feedback-based distributed consensus protocol suitable for directed time-varying topologies to make the internal state of each agent's encoder consistent with the output of its neighbors' decoder. For the case where the communication topology is directed, balanced and periodically connected, it is shown that if the difference of the reference inputs satisfies some boundedness condition, then the designed quantized dynamic consensus protocol can ensure the states of all the agents achieve dynamic average consensus with arbitrarily small steady state error by properly choosing system parameters. The lower bound of the required quantization levels and the method to choose the system parameters are also presented. Copyright © 2014 John Wiley & Sons, Ltd.

A new all-optical optical code division multiple access encoder that simultaneously regenerates input signals is proposed and experimentally investigated. The device is based on a fiber four-wave mixing and experimental results show that eye penalties may be decreased up to 5.1 dB. © 2014 Wiley Periodicals, Inc. Microwave Opt Technol Lett 56:1024–1028, 2014

This paper presents a high speed intra panel interface which applies point to point topology, phase locked loop (PLL) type clock data recovery (CDR) and DC self-adjusting data encoder. The preliminary measurement result on Altera stratix5 FPGA platform shows that the maximum data rate can be operated up to 5Gbps/lane.

This paper describes a technique related to the design of a trellis encoder, combined with the continuous phase modulation (CPM). Based on the idea of decomposing CPM into continuous phase encoder (CPE) and memoryless modulator (MM), superior trellis encoders are designed following Ungerboeck's set partitioning rules. Combinations are optimized with the squared minimum Euclidean distance based on the super trellis structure constructed by the trellis structure of trellis coded modulation (TCM) and CPE. Simulation results show that this new coding scheme consistently obtains better performance than previous schemes. © 2013 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

This paper proposes a multimedia streaming architecture that combines network and video technologies to handle video traffic over multi-hop access networks. In this context, resource overprovision typical of current quality of service approaches will become a limiting factor because of the increasing spread of bandwidth-intensive multimedia applications. The proposed scheme adopts a time-driven priority scheduling at network nodes and exploits slice classification at the video encoder to differentiate packets. The service guarantees offered by time-driven priority, together with the packet classification, significantly reduce congestion and increase the video quality at the receiver with respect to the traditional differentiated services DiffServ approach, also achieving high resource utilization. This is an important result, as current DiffServ-based architectures are far from obtaining such resource utilization in conjunction with reasonable delays and jitters. Copyright © 2013 John Wiley & Sons, Ltd.