Information Propagation Scheme Research Articles

A Curriculum-style Self-training Approach for Source-Free Semantic Segmentation.

Source-free domain adaptation has developed rapidly in recent years, where the well-trained source model is adapted to the target domain instead of the source data, offering the potential for privacy concerns and intellectual property protection. However, a number of feature alignment techniques in prior domain adaptation methods are not feasible in this challenging problem setting. Thereby, we resort to probing inherent domain-invariant feature learning and propose a curriculum-style self-training approach for source-free domain adaptive semantic segmentation. In particular, we introduce a curriculum-style entropy minimization method to explore the implicit knowledge from the source model, which fits the trained source model to the target data using certain information from easy-to-hard predictions. We then train the segmentation network by the proposed complementary curriculum-style self-training, which utilizes the negative and positive pseudo labels following the curriculum-learning manner. Although negative pseudo-labels with high uncertainty cannot be identified with the correct labels, they can definitely indicate absent classes. Moreover, we employ an information propagation scheme to further reduce the intra-domain discrepancy within the target domain, which could act as a standard post-processing method for the domain adaptation field. Furthermore, we extend the proposed method to a more challenging black-box source model scenario where only the source model's predictions are available. Extensive experiments validate that our method yields state-of-the-art performance on source-free semantic segmentation tasks for both synthetic-to-real and adverse conditions datasets. The code and corresponding trained models are released at https://github.com/yxiwang/ATP.

Meta-path guided graph attention network for explainable herb recommendation.

Traditional Chinese Medicine (TCM) has been widely adopted in clinical practice by Eastern Asia people for thousands of years. Nowadays, TCM still plays a critical role in Chinese society and receives increasing attention worldwide. The existing herb recommenders learn the complex relations between symptoms and herbs by mining the TCM prescriptions. Given a set of symptoms, they will provide a set of herbs and explanations from the TCM theory. However, the foundation of TCM is Yinyangism (i.e. the combination of Five Phases theory with Yin-yang theory), which is very different from modern medicine philosophy. Only recommending herbs from the TCM theory aspect largely prevents TCM from modern medical treatment. As TCM and modern medicine share a common view at the molecular level, it is necessary to integrate the ancient practice of TCM and standards of modern medicine. In this paper, we explore the underlying action mechanisms of herbs from both TCM and modern medicine, and propose a Meta-path guided Graph Attention Network (MGAT) to provide the explainable herb recommendations. Technically, to translate TCM from an experience-based medicine to an evidence-based medicine system, we incorporate the pharmacology knowledge of modern Chinese medicine with the TCM knowledge. We design a meta-path guided information propagation scheme based on the extended knowledge graph, which combines information propagation and decision process. This scheme adopts meta-paths (predefined relation sequences) to guide neighbor selection in the propagation process. Furthermore, the attention mechanism is utilized in aggregation to help distinguish the salience of different paths connecting a symptom with a herb. In this way, our model can distill the long-range semantics along meta-paths and generate fine-grained explanations. We conduct extensive experiments on a public TCM dataset, demonstrating comparable performance to the state-of-the-art herb recommendation models and the strong explainability.

Stateless Node Failure Information Propagation Scheme for Stable Overlay Networks

A structured overlay technology has the advantages for fault tolerance and computation resource (i.e., node) discovery in distributed data storage and its computation platform, however, these strengths are only guaranteed on stable environment that node failures do not occur frequently. To deal with the environment, many advanced schemes based on the well-known node failure information propagation scheme are proposed, which stabilizes the platform by quickly handling node failures. In the existing scheme, a computation node propagates a node-failure information when the node detect its failure. However, the existing scheme needs stateful maintenance against propagation targets; in other words, it must maintain the network connections of both the propagation target nodes and the nodes held on the general overlay. The nodes then exhaust the machine resources (e.g., CPU, memory, network bandwidth) for the connection management and cannot concentrates on their own tasks, such as data analysis or its storage application. To resolve this problem, I propose a stateless node-failure information propagation scheme, which propagates a node failure at the speed of the existing scheme but without requiring maintenance of the propagation target connections. In the proposed scheme, each computational node can effectively utilize the machine resources for its own task. Instead of retaining the propagation targets, my scheme estimates the propagation targets after detecting a node failure. I analyzed the estimation accuracy of a simple propagation model, which guarantees effective propagation. The accuracy was found to depend on the overlay distance between the failed node and the propagator node. Based on this observation, my scheme adjusts the keep-alive interval to bias the detection of closer node failures. In a simulation evaluation, the detection delay of the proposed stateless propagation was similar to that of the stateful propagation scheme, but delivered superior maintenance cost and scalability.

Consideration on definitions of part-whole relation and its efficient information propagation scheme on cellular-automata universe

Consideration on definitions of part-whole relation and its efficient information propagation scheme on cellular-automata universe

Iterative framework for the joint segmentation and CT synthesis of MR images: application to MRI-only radiotherapy treatment planning

To tackle the problem of magnetic resonance imaging (MRI)-only radiotherapy treatment planning (RTP), we propose a multi-atlas information propagation scheme that jointly segments organs and generates pseudo x-ray computed tomography (CT) data from structural MR images (T1-weighted and T2-weighted). As the performance of the method strongly depends on the quality of the atlas database composed of multiple sets of aligned MR, CT and segmented images, we also propose a robust way of registering atlas MR and CT images, which combines structure-guided registration, and CT and MR image synthesis.We first evaluated the proposed framework in terms of segmentation and CT synthesis accuracy on 15 subjects with prostate cancer. The segmentations obtained with the proposed method were compared using the Dice score coefficient (DSC) to the manual segmentations. Mean DSCs of 0.73, 0.90, 0.77 and 0.90 were obtained for the prostate, bladder, rectum and femur heads, respectively. The mean absolute error (MAE) and the mean error (ME) were computed between the reference CTs (non-rigidly aligned to the MRs) and the pseudo CTs generated with the proposed method. The MAE was on average HU and the ME HU. We then performed a dosimetric evaluation by re-calculating plans on the pseudo CTs and comparing them to the plans optimised on the reference CTs. We compared the cumulative dose volume histograms (DVH) obtained for the pseudo CTs to the DVH obtained for the reference CTs in the planning target volume (PTV) located in the prostate, and in the organs at risk at different DVH points. We obtained average differences of in the PTV for , and between and 0.05% in the PTV, bladder, rectum and femur heads for Dmean and .Overall, we demonstrate that the proposed framework is able to automatically generate accurate pseudo CT images and segmentations in the pelvic region, potentially bypassing the need for CT scan for accurate RTP.

Virtual-MIMO-Boosted Information Propagation on Highways

In vehicular communications, traffic-related information should be spread over the network as quickly as possible to maintain a safer transportation system. This motivates us to develop more efficient information propagation schemes. In this paper, we propose a novel virtual-MIMO-enabled information dissemination scheme, in which the vehicles opportunistically form virtual antenna arrays to boost the transmission range and therefore accelerate information propagation along the highway. We model the information propagation process as a renewal reward process and investigate in detail the \emph{Information Propagation Speed} (IPS) of the proposed scheme. The corresponding closed-form IPS is derived, which shows that the IPS increases cubically with the vehicle density but will ultimately converge to a constant upper bound. Moreover, increased mobility also facilitates the information spreading by offering more communication opportunities. However, the limited network density essentially determines the bottleneck in information spreading. Extensive simulations are carried out to verify our analysis. We also show that the proposed scheme exhibits a significant IPS gain over its conventional counterpart.

CT synthesis in the head & neck region for PET/MR attenuation correction: an iterative multi-atlas approach.

In this work, we propose to tackle the problem of attenuation correction in the head and neck by synthesising CT from MR images using an iterative multi-atlas approach. The proposed method relies on pre-acquired T2-weighted MRI and CT images of the neck. For each subject, the MRI is non-rigidly mapped to the CT. To synthesise a pseudo CT, all the MRIs in the database are first registered to the target MRI. This registration consists of a robust affine followed by a non-rigid registration. The pseudo CT is obtained by fusing the mapped atlases according to their morphological similarity to the target. In contrast to CTs, T2 images do not provide a good estimate of the bone location. Combining multiple modalities at both the registration and image similarity stages is expected to provide more realistic mappings and to reduce the bias. An initial pseudo CT (pCT) is combined with the target MRI to form a MRI-pCT pair. The MRI-pCT pair is registered to all the MRI-CT pairs from the database. An improved pseudo CT is obtained by fusing the mapped MRI-CT pairs according to their morphological similarity to the target MRI-pCT pair. Results showed that the proposed CT synthesis algorithm based on a multi-atlas information propagation scheme and iterative process is able to synthesise pseudo CT images in a region challenging for registration algorithms. The results also demonstrate that the robust affine decreases the absolute error compared to the classic approach and that the bone refinement process reduces the bias in the bone region. The proposed method could be used to correct for attenuation PET/MR data, but also for dosimetry calculations in the context of MR-based radiotherapy treatment planning.

Attenuation correction synthesis for hybrid PET-MR scanners: application to brain studies.

Attenuation correction is an essential requirement for quantification of positron emission tomography (PET) data. In PET/CT acquisition systems, attenuation maps are derived from computed tomography (CT) images. However, in hybrid PET/MR scanners, magnetic resonance imaging (MRI) images do not directly provide a patient-specific attenuation map. The aim of the proposed work is to improve attenuation correction for PET/MR scanners by generating synthetic CTs and attenuation maps. The synthetic images are generated through a multi-atlas information propagation scheme, locally matching the MRI-derived patient's morphology to a database of MRI/CT pairs, using a local image similarity measure. Results show significant improvements in CT synthesis and PET reconstruction accuracy when compared to a segmentation method using an ultrashort-echo-time MRI sequence and to a simplified atlas-based method.