Minimal Increase In Computational Complexity Research Articles

ACMTR: Attention-guided, combined multi-scale, transformer reasoning-based network for 3D CT pelvic functional bone marrow segmentation

Automatic segmentation of pelvic functional bone marrow (FBM) in computed tomography (CT) plays a critical role in the accurate diagnosis of pelvic malignancies and treatment planning for patients. Pelvic bone marrow near the mid-axis of the body, i.e., the FBM, is the most hematopoietic. Here we propose an Attention-guided, Combined Multi-scale, Transformer Reasoning-based network (ACMTR) for accurate segmentation of pelvic FBM for intensity-modulated radiation therapy. The presented Convolution-Attention Fusion Encoder overcomes the limited receptive field of convolution as well as the class imbalance issue in pelvic FBM segmentation. We also propose a Multi-scale Transformer Global Reasoning Bottleneck module for obtaining multi-scale, global attention features for extraction of small edge tissue of the FBM. We introduce a deformable attention mechanism at the decoding path to focus on key features instead of redundant information with a minimal increase in computational complexity. The pelvic FBM CT images used in this study were from a self-constructed clinical database of 92 cases and showed two main areas of FBM: the most intense FBM area (i.e., most hematopoietic area) and the second most intense FBM area. The proposed ACMTR was compared with the popular deep learning models U-Net, VGG-Net, Attention-UNet, TransUNet, CoTr, nnUNet, and nnFormer. For the most intense and second most intense FBM areas, the proposed ACMTR achieved Dice Scores of 0.799 and 0.772 and Hausdorff Distance of 6.061 and 4.791, respectively, on test data. The implementation code is available at the following Github link:https://github.com/walynlee/ACMTR.

Multi-resolution statistical image reconstruction for mitigation of truncation effects: application to cone-beam CT of the head

A prototype cone-beam CT (CBCT) head scanner featuring model-based iterative reconstruction (MBIR) has been recently developed and demonstrated the potential for reliable detection of acute intracranial hemorrhage (ICH), which is vital to diagnosis of traumatic brain injury and hemorrhagic stroke. However, data truncation (e.g. due to the head holder) can result in artifacts that reduce image uniformity and challenge ICH detection. We propose a multi-resolution MBIR method with an extended reconstruction field of view (RFOV) to mitigate truncation effects in CBCT of the head. The image volume includes a fine voxel size in the (inner) nontruncated region and a coarse voxel size in the (outer) truncated region. This multi-resolution scheme allows extension of the RFOV to mitigate truncation effects while introducing minimal increase in computational complexity. The multi-resolution method was incorporated in a penalized weighted least-squares (PWLS) reconstruction framework previously developed for CBCT of the head. Experiments involving an anthropomorphic head phantom with truncation due to a carbon-fiber holder were shown to result in severe artifacts in conventional single-resolution PWLS, whereas extending the RFOV within the multi-resolution framework strongly reduced truncation artifacts. For the same extended RFOV, the multi-resolution approach reduced computation time compared to the single-resolution approach (viz. time reduced by 40.7%, 83.0%, and over 95% for an image volume of 6003, 8003, 10003 voxels). Algorithm parameters (e.g. regularization strength, the ratio of the fine and coarse voxel size, and RFOV size) were investigated to guide reliable parameter selection. The findings provide a promising method for truncation artifact reduction in CBCT and may be useful for other MBIR methods and applications for which truncation is a challenge.

Video streaming over an Ad Hoc network using data partitioning and path diversity

Multiple description video coding can exploit path diversity to improve error resiliency by generating several independently decodable video descriptions and sending them over different paths. Such paths are readily available in ad hoc wireless networks. To approach the original quality, all descriptions need to be successfully received. However, in a wireless channel, error bursts or link disruption may still affect one or more descriptions, resulting in significant drops in video quality. This article describes a lower complexity yet effective scheme for video streaming over such wireless networks. In order to improve error resiliency, the scheme combines unequal packet importance with the advantages of path diversity. The scheme can be implemented with a minimal increase in computational complexity, as spatial decomposition of video frames is employed to generate the descriptions. Experimental results for an ad hoc wireless network show that a significant objective quality gain can be achieved (up to 3 dB) in the presence of signal shadow fading.

Integration of beamforming and uncertainty-of-observation techniques for robust ASR in multi-source environments

This paper presents a new approach for increasing the robustness of multi-channel automatic speech recognition in noisy and reverberant multi-source environments. The proposed method uses uncertainty propagation techniques to dynamically compensate the speech features and the acoustic models for the observation uncertainty determined at the beamforming stage. We present and analyze two methods that allow integrating classical multi-channel signal processing approaches like delay and sum beamformers or Zelinski-type Wiener filters, with uncertainty-of-observation techniques like uncertainty decoding or modified imputation. An analysis of the results on the PASCAL-CHiME task shows that this approach consistently outperforms conventional beamformers with a minimal increase in computational complexity. The use of dynamic compensation based on observation uncertainty also outperforms conventional static adaptation with no need of adaptation data.

Robust iterative multiuser receiver for coded CDMA systems in non-Gaussian channels

A robust front-end pre-filter incorporating non-quadratic signal processing techniques is proposed. We will demonstrate the effectiveness of this filter in suppressing impulsive noises in a coded direct sequence code-division multiple access (DS-CDMA)communication system. Based on M-Estimation approach, the pre-filter works by minimizing an appropriate penalty function in an iterative fashion with minimal increase in computational complexity of the system. With the use of this pre-filter in the system, performance enhancement in excess of 6dB in non-Gaussian impulsive channels is demonstrated while maintaining comparable performance in Gaussian channel.

Application of degenerate perturbation theory to subspace tracking

First-order perturbation theory has been applied to subspace tracking. This previous work did not solve the problem of signal subspace near-degeneracy. The signal subspace is nearly degenerate when two or more signal subspace eigenvalues are nearly equal, which can occur under a number of circumstances. Failure to correctly handle the near-degeneracy leads to errors in the subspace decomposition. We demonstrate here the application of degenerate perturbation theory to signal subspace tracking under conditions where the eigenvalues may be degenerate. The problems of noise subspace degeneracy and signal subspace near-degeneracy are both solved through diagonalization of the perturbation matrix over the degenerate subspace. The degenerate subspace diagonalization markedly reduces the subspace error with a minimal increase in computational complexity. Similarly, it is shown that using a second-order perturbation update reduces the subspace error by a substantial margin, again with a minimal increase in computational complexity. The effect of increasing the integration time is also examined. By using the second-order perturbation update and the degenerate subspace diagonalization, the resulting subspace decomposition is extremely accurate, robust, and fast.

Nonlinear Model Predictive Control with State and Parameter Estimation

Abstract Among nonlinear predictive control techniques, nonlinear quadratic dynamic matrix control is one of the most attractive mainly for the small additional computational demand with respect to its linear counterpart. However, in order to use a full first principle nonlinear dynamic model in the computation of the control actions, the addition of a state estimator to the algorithm is required. In this paper it is proposed to include parameter estimation in the estimator in order to avoid the problems connected with parametric uncertainty. The main issues connected with the choice of the parameters to be estimated for systems with a large number of uncertain parameters are addressed with reference to a terpolymerization reactor. Applications to this case study show that the modified technique offers a major improvement in performance with a minimal increase in computational complexity.