Purpose: This study evaluates the performance of a kilovoltage x-ray image-guidance system equipped with a novel post-processing optimization algorithm on the newly introduced TAICHI linear accelerator (Linac). Method...
详细信息
Purpose: This study evaluates the performance of a kilovoltage x-ray image-guidance system equipped with a novel post-processing optimization algorithm on the newly introduced TAICHI linear accelerator (Linac). Methods: A comparative study involving image quality tests and radiation dose measurements was conducted across six scanning protocols of the kV-cone beam computed tomography (CBCT) system on the TAICHI Linac. The performance assessment utilized the conventional Feldkamp-Davis-Kress (FDK)algorithm and a novel Non-Local Means denoising and adaptive scattering correction (NLM-ASC) algorithm. Image quality metrics, including spatial resolution, contrast-to-noise ratio (CNR), and signal-to-noise ratio (SNR), were evaluated using a Catphan 604 phantom. Radiation doses for low-dose and standard protocols were measured using a computed tomography dose index (CTDI)phantom, with comparative measurements from the Halcyon Linac's iterative CBCT (iCBCT). Results: The NLM-ASC algorithm significantly improved image quality, achieving a 300%-1000% increase in CNR and SNR over the FDK-only images and it also showed a 100%-200% improvement over the iCBCT images from Halcyon's head protocol. The optimized low-dose protocols yielded higher image quality than the standard FDK protocols, indicating potential for reduced radiation exposure. Clinical implementation confirmed the TAICHI system's utility for precise and adaptive radiotherapy. Conclusion: The kV-IGRT system on the TAICHI Linac, with its novel post-processing algorithm, demonstrated superior image quality suitable for routine clinical use, effectively reducing image noise without compromising other quality metrics.
Ultrasound localization microscopy (ULM) is a super-resolution vascular imaging technique that tracks the position of individual circulating microbubbles over time. Conventional ULM methods fit a gaussian distribution...
详细信息
Ultrasound localization microscopy (ULM) is a super-resolution vascular imaging technique that tracks the position of individual circulating microbubbles over time. Conventional ULM methods fit a gaussian distribution to a coarsely-pixelated brightness image of a microbubble, and assume the microbubble's position is the mean. In this work, we developed alternative algorithms to precisely and accurately estimate the location of a microbubble by considering multiple even and odd receive apodization profiles applied to channel RF data. These receive apodization profiles each yield a point spread function with a unique lateral character, enabling intra-beam determination of a scatterer's location. algorithms were refined and evaluated first for the case of focused-beam transmits over a range of F-numbers, and also for the case of ultrafast, multi-angle plane-wave compounding. The performance of the algorithms was experimentally evaluated on a research ultrasound scanner using a P4-2v phased array probe and a custom wire-target phantom. Results show that the position of the scatterer was resolvable to less than one-fourth of the diffraction limited resolution.
Experimental measurement of two-phase flow in narrow rectangular channels is of great interest for various industrial applications. As it is almost prohibitive to apply the conventional two-phase flow measurement tech...
详细信息
Experimental measurement of two-phase flow in narrow rectangular channels is of great interest for various industrial applications. As it is almost prohibitive to apply the conventional two-phase flow measurement technology in narrow rectangular channels, the novel flexible printed circuit (FPC) sensor has been developed based on the wire-mesh sensor system. The FPC sensor avoids the stretching device which significantly disturbs the flow and achieves the spatial resolution of 1.35 mm and the sampling rate up to 3150 Hz in a narrow channel. The dedicated post-processing algorithms, including that for noise elimination, resolution refinement, interface sharpening and parameter derivation, were developed for bubbly and slug flows. The reliability of the FPC sensor and the accuracy of post-processing algorithms are assessed based on comparison with high-speed camera images. According to the validation, for bubbly flow with the bubble diameter larger than 4 mm, the bubble diameter can be underestimated by 5.2% with the relative standard deviation below 7.8%. When recommended criterion is followed in postprocessing, the bubble velocity is underestimated by 2.0% with the relative standard deviation below 3.74%. For slug flow, Taylor bubble velocity and length can be underestimated by 1.03% and 2.22% with the relative standard deviation of 5.37% and 7.11%.
The concept of local shock strength and a quantitative measure index str of local shock strength are proposed, derived from the oblique shock relation and the monotonic relationship between total pressure loss ratio a...
详细信息
The concept of local shock strength and a quantitative measure index str of local shock strength are proposed, derived from the oblique shock relation and the monotonic relationship between total pressure loss ratio and normal Mach number. Utilizing the high density gradient characteristic of shock waves and the oblique shock relation, a post-processing algorithm for two-dimensional flow field data is developed. The objective of the post-processing algorithm is to obtain specific shock wave location coordinates and calculate the corresponding str from flow filed data under the calibration of the oblique shock relation. Validation of this post-processing algorithm is conducted using a standard model example that can be solved analytically. Combining the concept of local shock strength with the post-processing algorithm, a local shock strength quantitative mapping approach is established for the first time. This approach enables a quantitative measure and visualization of local shock strength at distinct locations, represented by color mapping on the shock structures. The approach can be applied to post-processing numerical simulation data of two-dimensional flows. Applications to the intersection of two left-running oblique shock waves (straight shock waves), the bow shock in front of a cylinder (curved shock wave), and Mach reflection (mixed straight and curved shock waves) demonstrate the accuracy, and effectiveness of the mapping approach in investigating diverse shock wave phenomena. The quantitative mapping approach of str may be a valuable tool in the design of supersonic/hypersonic vehicles and the exploration of shock wave evolution.
Luminaire detection methods based on deep learning encounter challenges in tunnel luminaire detection due to the complex environment and unfavorable lighting conditions. To overcome these issues, this paper proposes a...
详细信息
Luminaire detection methods based on deep learning encounter challenges in tunnel luminaire detection due to the complex environment and unfavorable lighting conditions. To overcome these issues, this paper proposes an improved tunnel luminaire detection solution by enhancing the Mask R-CNN using brightness balancing and distribution-guided optimization. Leveraging tunnel gray-scale images and the Mask R-CNN object detection framework, a feature fusion network based on ResNet-FPN, trained via transfer learning, which enhances performance in detecting object luminaires. Furthermore, considering the differences in luminaire brightness, their backgrounds, and regular spacing, an object brightness enhancement method based on a fixed threshold and an innovative distribution-guided optimization strategy are introduced to effectively reducing missed detections and false alarms while accurately correcting the detected luminaire positions. To evaluate the performance of the proposed approach, real datasets of tunnel environment are used. Experimental results revealed that the proposed approach achieved precision, recall, and an F1-score of 98.3%, 98.5%, and 0.984, respectively, which improved of 0.5%, 5.1%, and 0.029, respectively, compared to the original model, thus, could be applied to the 3D model construction and intelligent management of tunnels. The code implementing our method is available at https://***/xubintao/OBE-and-DGO.
Tensor-based multi-view clustering (TMVC) has garnered considerable attention for its efficacy in managing data that originate from multiple perspectives. However, the presence of noise in empirical datasets often und...
详细信息
Tensor-based multi-view clustering (TMVC) has garnered considerable attention for its efficacy in managing data that originate from multiple perspectives. However, the presence of noise in empirical datasets often undermines the reliability and robustness of the affinity matrices generated through these methods. To address this challenge, we introduce an innovative approach termed tensor multi-subspace learning (TMSL). Our methodology commences with the employment of a typical TMVC method to produce self-representation matrices for each view. Nevertheless, the affinity matrix derived from these self-representation matrices frequently falls short of the desired levels of dependability and robustness. To uncover the intrinsic architecture of the data within the tensor subspace, we harness the concept of tensor low-rank representation. This enables us to extract a higher-dimensional representation of multi-view data, thereby yielding a multi-subspace representation tensor that is both reliable and robust. These two stages are then seamlessly integrated into a unified framework and are resolved by employing the augmented Lagrangian algorithm. Notably, the TMSL method also serves as an effective post-processing strategy capable of being applied to various TMVC methods to augment their performance. Empirical evidence has established that TMSL outperforms other contemporary methods, and the post-processing strategy has proven to be an effective unified approach that can be extended to other TMVC methods.
Improving tree species classification accuracy often involves complex workflows, constrained by high computational costs, extensive data requirements, and sensitivity to spatiotemporal variations. This study introduce...
详细信息
Improving tree species classification accuracy often involves complex workflows, constrained by high computational costs, extensive data requirements, and sensitivity to spatiotemporal variations. This study introduces the Change Resistance Filter (CR-Filter), inspired by the stable growth patterns of the Climax Community. The CR-Filter, applied as a post-processing tool, integrates Change Resistance on Timelines and Change Resistance on Spatial Neighboring into a unified framework, enhancing classification precision by mitigating spatiotemporal fluctuations. Liupan Mountain Nature Reserve was selected as the study area for its ecological stability. Multi-temporal Sentinel-2 data spanning several years were used to extract and correct phenological indices, which were combined with Sentinel-1 and terrain data to generate interannual tree species classification maps. These maps were subsequently refined using the CR-Filter. Compared to traditional methods, robustness in highly heterogeneous regions was improved by leveraging interannual map integration, yielding species distribution maps with greater spatial consistency and temporal stability. Overall accuracy increased by 8.44%, from 85.85% to 93.10%, effectively reducing misclassification from noise or transient changes. This approach highlights the CR-Filter's efficacy with limited samples and medium-to-low resolution, providing strong technical support for remote sensing-based species mapping and ecological research.
Word embeddings are widely deployed in a tremendous range of fundamental natural language processing applications and are also useful for generating representations of paragraphs, sentences, and documents. In some con...
详细信息
Word embeddings are widely deployed in a tremendous range of fundamental natural language processing applications and are also useful for generating representations of paragraphs, sentences, and documents. In some contexts involving constrained memory, it may be beneficial to reduce the size of word embeddings since they represent a core component of several natural language processing tasks. By reducing the dimensionality of word embeddings, their usefulness in memory-limited devices can be significantly improved, yielding gains in many real-world applications. This article aims to provide a comparative study of different dimensionality reduction techniques to generate efficient lower-dimensional word vectors. Based on empirical experiments carried out on the Arabic machine translation task, we found that the post-processing algorithm combined with independent component analysis provides optimal performance over the considered dimensionality reduction techniques. Therefore, we arrive at a new combination of the post-processing algorithm and dimensionality reduction (independent component analysis) techniques, which has not been investigated before. The latter was applied to both contextual and non-contextual word embeddings to reduce the size of the vectors while achieving a better translation quality than the original ones.
The concept of "form recognition" has had a lot of applications from different times, such as art, training sports, and movement analysis. Nonetheless, bicycle configuration recognition is one of the areas t...
详细信息
ISBN:
(纸本)9798331530372;9798331530365
The concept of "form recognition" has had a lot of applications from different times, such as art, training sports, and movement analysis. Nonetheless, bicycle configuration recognition is one of the areas that led to the development of a model that only earlier didn't receive much attention up to recently. The posture of cycling is also crucial for human health, in order to better analyze and correct posture, in our study, we present a bike form detection model which is a product designed for the use of leisure cyclists. Cyclists will be able to utilize this model to inspect whether they have a good sitting position, affect their acceleration, and improve their own well-being. We can now evaluate the position of joints and develop a system of grades with just a phone camera being used yet still achieving top level accuracy. In addition, our 3D pose estimation has been brought to cycling through our unique database that has shown remarkable results in terms of rider's posture improvement.
The vibration signals produced by rotating machinery are mostly non-stationary, and there are numerous methods for dealing with them. People's expectations for time-frequency analysis (TFA) results are increasing ...
详细信息
The vibration signals produced by rotating machinery are mostly non-stationary, and there are numerous methods for dealing with them. People's expectations for time-frequency analysis (TFA) results are increasing all the time. The emergence of post-processing algorithms based on the short-time Fourier transform (STFT) provides scholars with new ideas, but such algorithms heavily rely on the window length selected by STFT and have significant uncertainty. To address this issue, we propose the generalized S-synchroextracting transform, a new time-frequency post-processing algorithm (GS-SET). The algorithm extracts the coefficients on the TF ridge of the generalized S-transform (GST) to remove the majority of the dispersed TF energy, allowing the time-frequency representation (TFR) to achieve optimal TF resolution. The results of the analog signal processing show that the method can characterize the signal clearly and accurately, and it has good noise robustness. To process the fault signals of the three groups of rolling bearings, we use different TFA methods. The results show that the method can more precisely determine the characteristic frequency of the faulty bearing. Finally, the superiority of this method is demonstrated further by processing data from Case Western Reserve University's faulty bearing database.
暂无评论