Indirect Shear Wave Excitation for Brain Magnetic Resonance Elastography with Minimal Cerebral Blood Flow Alteration

Magnetic resonance elastography (MRE) of brain relies on inducing and measuring shear waves in the brain. However, studies have shown vibration could induce changes in cerebral blood flow (CBF), which has a modulation effect and can affect the biomechanical properties measured.

OBJECTIVE

This work demonstrates the initial prototype of the indirect excitation method, which can generate shear waves in the brain with minimal changes in CBF.

METHODS

A simple system was designed to produce stable vibrations underneath the neck. Instead of directly stimulating the skull, shear waves were indirectly transmitted to the brain through the spine and brainstem.

RESULTS

Phantom results showed that the proposed actuator did not interfere with the routine imaging sequence and successfully generated multifrequency shear waves. When compared with the conventional direct head stimulation method, brain MRE results from the proposed actuator showed no significant differences in terms of intraclass correlation coefficients (ICC) and coefficients of variation (CV). Moreover, the octahedral shear strain (OSS) generated by the indirect excitation in the frontal and parietal lobes decreased by 25.96% and 16.73% respectively. Evaluation of CBF in healthy volunteers revealed no significant changes for the indirect excitation method, whereas significant decreases in CBF were observed in four subregions when employing direct excitation.

CONCLUSION

The proposed actuator offers a more accurate and comfortable approach to MRE measurements while causing minimal CBF alterations.

SIGNIFICANCE

This work presents the first demonstration of an indirect excitation brain MRE system that minimizes CBF changes, thus holding potential for future applications of brain MRE.

A Comparative Study of Three Systems for Liver Magnetic Resonance Elastography

BACKGROUND

Different MR elastography (MRE) systems may produce different stiffness measurements, making direct comparison difficult in multi-center investigations.

PURPOSE

To assess the repeatability and reproducibility of liver stiffness measured by three typical MRE systems.

STUDY TYPE

Prospective.

POPULATION/PHANTOMS

Thirty volunteers without liver disease history (20 males, aged 21-28)/5 gel phantoms.

FIELD STRENGTH/SEQUENCE

3.0 T United Imaging Healthcare (UIH), 1.5 T Siemens Healthcare, 3.0 T General Electric Healthcare (GE)/Echo planar imaging-based MRE sequence.

ASSESSMENT

Wave images of volunteers and phantoms were acquired by three MRE systems. Tissue stiffness was evaluated by two observers, while phantom stiffness was assessed automatically by code. The reproducibility across three MRE systems was quantified based on the mean stiffness of each volunteer and phantom.

STATISTICAL TESTS

Intraclass correlation coefficients (ICC), coefficients of variation (CV), and Bland-Altman analyses were used to assess the interobserver reproducibility, the interscan repeatability, and the intersystem reproducibility. Paired t-tests were performed to assess the interobserver and interscan variation. Friedman tests with Dunn's multiple comparison correction were performed to assess the intersystem variation. P values less than 0.05 indicated significant difference.

RESULTS

The reproducibility of stiffness measured by the two observers demonstrated consistency with ICC > 0.92, CV < 4.32%, Mean bias < 2.23%, and P > 0.06. The repeatability of measurements obtained using the electromagnetic system for the liver revealed ICC > 0.96, CV < 3.86%, Mean bias < 0.19%, P > 0.90. When considering the range of reproducibility across the three systems for liver evaluations, results ranged with ICCs from 0.70 to 0.87, CVs from 6.46% to 10.99%, and Mean biases between 1.89% and 6.30%. Phantom studies showed similar results. The values of measured stiffness differed across all three systems significantly.

DATA CONCLUSION

Liver stiffness values measured from different MRE systems can be different, but the measurements across the three MRE systems produced consistent results with excellent reproducibility.

EVIDENCE LEVEL

1 TECHNICAL EFFICACY: Stage 2.

Magnetic soft microfiberbots for robotic embolization

Cerebral aneurysms and brain tumors are leading life-threatening diseases worldwide. By deliberately occluding the target lesion to reduce the blood supply, embolization has been widely used clinically to treat cerebral aneurysms and brain tumors. Conventional embolization is usually performed by threading a catheter through blood vessels to the target lesion, which is often limited by the poor steerability of the catheter in complex neurovascular networks, especially in submillimeter regions. Here, we propose magnetic soft microfiberbots with high steerability, reliable maneuverability, and multimodal shape reconfigurability to perform robotic embolization in submillimeter regions via a remote, untethered, and magnetically controllable manner. Magnetic soft microfiberbots were fabricated by thermal drawing magnetic soft composite into microfibers, followed by magnetizing and molding procedures to endow a helical magnetic polarity. By controlling magnetic fields, magnetic soft microfiberbots exhibit reversible elongated/aggregated shape morphing and helical propulsion in flow conditions, allowing for controllable navigation through complex vasculature and robotic embolization in submillimeter regions. We performed in vitro embolization of aneurysm and tumor in neurovascular phantoms and in vivo embolization of a rabbit femoral artery model under real-time fluoroscopy. These studies demonstrate the potential clinical value of our work, paving the way for a robotic embolization scheme in robotic settings.

Noise-Factorized Disentangled Representation Learning for Generalizable Motor Imagery EEG Classification

Motor Imagery (MI) Electroencephalography (EEG) is one of the most common Brain-Computer Interface (BCI) paradigms that has been widely used in neural rehabilitation and gaming. Although considerable research efforts have been dedicated to developing MI EEG classification algorithms, they are mostly limited in handling scenarios where the training and testing data are not from the same subject or session. Such poor generalization capability significantly limits the realization of BCI in real-world applications. In this paper, we proposed a novel framework to disentangle the representation of raw EEG data into three components, subject/session-specific, MI-task-specific, and random noises, so that the subject/session-specific feature extends the generalization capability of the system. This is realized by a joint discriminative and generative framework, supported by a series of fundamental training losses and training strategies. We evaluated our framework on three public MI EEG datasets, and detailed experimental results show that our method can achieve superior performance by a large margin compared to current state-of-the-art benchmark algorithms.

Chemical Constituents from Berchemia polyphylla var. Leioclada

One previously undescribed naphthoquinone-benzisochromanquinone dimer berpolydiquinone A (1), along with two previously undescribed naphthoquinone-anthraquinone dimers berpolydiquinones B and C (2-3), and one previously undescribed dimeric naphthalene berpolydinaphthalene A (4), were isolated from the stems and leaves of Berchemia polyphylla var. leioclada. The chemical structures of these compounds were determined using high-resolution electrospray ionization mass spectroscopy (HR-ESI-MS), spectroscopic data, the exciton chirality method (ECM), and quantum chemical calculation. Notably, compounds (1-2 and 5) are dimeric quinones that share the same naphthoquinone moiety, specifically identified as 2-methoxystypandron. Compound (4) is a derivative of dimeric naphthalene with a symmetrical structure, which is a new structure type isolated from B. polyphylla var. leioclada for the first time. These findings suggest that B. polyphylla var. leioclada serves as a significant reservoir of structurally diverse phenolic compounds. This study provides a scientific foundation for regarding B. polyphylla var. leioclada as a potential source of "Tiebaojin".

Fiberbots: Robotic fibers for high-precision minimally invasive surgery

Precise manipulation of flexible surgical tools is crucial in minimally invasive surgical procedures, necessitating a miniature and flexible robotic probe that can precisely direct the surgical instruments. In this work, we developed a polymer-based robotic fiber with a thermal actuation mechanism by local heating along the sides of a single fiber. The fiber robot was fabricated by highly scalable fiber drawing technology using common low-cost materials. This low-profile (below 2 millimeters in diameter) robotic fiber exhibits remarkable motion precision (below 50 micrometers) and repeatability. We developed control algorithms coupling the robot with endoscopic instruments, demonstrating high-resolution in situ molecular and morphological tissue mapping. We assess its practicality and safety during in vivo laparoscopic surgery on a porcine model. High-precision motion of the fiber robot delivered endoscopically facilitates the effective use of cellular-level intraoperative tissue identification and ablation technologies, potentially enabling precise removal of cancer in challenging surgical sites.

Beyond Supervised Learning for Pervasive Healthcare

The integration of machine/deep learning and sensing technologies is transforming healthcare and medical practice. However, inherent limitations in healthcare data, namely scarcity, quality, and heterogeneity, hinder the effectiveness of supervised learning techniques which are mainly based on pure statistical fitting between data and labels. In this article, we first identify the challenges present in machine learning for pervasive healthcare and we then review the current trends beyond fully supervised learning that are developed to address these three issues. Rooted in the inherent drawbacks of empirical risk minimization that underpins pure fully supervised learning, this survey summarizes seven key lines of learning strategies, to promote the generalization performance for real-world deployment. In addition, we point out several directions that are emerging and promising in this area, to develop data-efficient, scalable, and trustworthy computational models, and to leverage multi-modality and multi-source sensing informatics, for pervasive healthcare.

Micro/nanosystems for controllable drug delivery to the brain

Drug delivery to the brain is crucial in the treatment for central nervous system disorders. While significant progress has been made in recent years, there are still major challenges in achieving controllable drug delivery to the brain. Unmet clinical needs arise from various factors, including controlled drug transport, handling large drug doses, methods for crossing biological barriers, the use of imaging guidance, and effective models for analyzing drug delivery. Recent advances in micro/nanosystems have shown promise in addressing some of these challenges. These include the utilization of microfluidic platforms to test and validate the drug delivery process in a controlled and biomimetic setting, the development of novel micro/nanocarriers for large drug loads across the blood-brain barrier, and the implementation of micro-intervention systems for delivering drugs through intraparenchymal or peripheral routes. In this article, we present a review of the latest developments in micro/nanosystems for controllable drug delivery to the brain. We also delve into the relevant diseases, biological barriers, and conventional methods. In addition, we discuss future prospects and the development of emerging robotic micro/nanosystems equipped with directed transportation, real-time image guidance, and closed-loop control.

[Prognostic value of the Second Revision of the International Staging System in patients with newly diagnosed transplant-eligible multiple myeloma]

Objective: To verify the predictive value of the Second Revision of the International Staging System (R2-ISS) in newly diagnosed patients with multiple myeloma (MM) who underwent first-line autologous hematopoietic stem cell transplantation (ASCT) in a new drug era in China. Methods: This multicenter retrospective cohort study enrolled patients with newly diagnosed MM from three centers in China (Beijing Chao-Yang Hospital, Capital Medical University; the First Affiliated Hospital, Sun Yat-Sen University, and the Second Affiliated Hospital of Naval Medical University) from June 2008 to June 2018. A total of 401 newly diagnosed patients with MM who were candidates for ASCT were enrolled in this cohort, all received proteasome inhibitor and/or immunomodulator-based induction chemotherapy followed by ASCT. Baseline and follow-up data were collected. The patients were regrouped using R2-ISS. Progression-free survival (PFS) and overall survival (OS) were analyzed. The Kaplan-Meier method was used to analyze the survival curve and two survival curves were compared using the log-rank test. Cox regression analysis were performed to analyze the relationship between risk factors and survival. Results: The median age of the patients was 53 years (range 25-69 years) and 59.5% (240 cases) were men. Newly diagnosed patients with renal impairment accounted for 11.5% (46 cases). According to Revised-International Staging System (R-ISS), 74 patients (18.5 %) were diagnosed with stage Ⅰ, 259 patients (64.6%) with stage Ⅱ, and 68 patients (17.0%) with stage Ⅲ. According to the R2-ISS, the distribution of patients in each group was as follows: 50 patients (12.5%) in stage Ⅰ, 95 patients (23.7%) in stage Ⅱ, 206 patients (51.4%) in stage Ⅲ, and 50 patients (12.5%) in stage Ⅳ. The median follow-up time was 35.9 months (range, 6-119 months). According to the R2-ISS stage, the median PFS in each group was: 75.3 months for stage Ⅰ; 62.0 months for stage Ⅱ, 39.2 months for stage Ⅲ, and 30.3 months for stage Ⅳ; and the median OS was not reached, 86.6 months, 71.6 months, and 38.5 months, respectively. There were statistically significant differences in PFS and OS between different groups (both P<0.001). Multivariate Cox regression analysis showed that stages Ⅲ and Ⅳ of the R2-ISS were independent prognostic factors for PFS (HR=2.37, 95%CI 1.30-4.30; HR=4.50, 95%CI 2.35-9.01) and OS (HR=4.20, 95%CI 1.50-11.80; HR=9.53, 95%CI 3.21-28.29). Conclusions: The R2-ISS has significant predictive value for PFS and OS for transplant-eligible patients with MM in the new drug era. However, the universality of the R2-ISS still needs to be further verified in different populations.

A deep unrolled neural network for real-time MRI-guided brain intervention

Accurate navigation and targeting are critical for neurological interventions including biopsy and deep brain stimulation. Real-time image guidance further improves surgical planning and MRI is ideally suited for both pre- and intra-operative imaging. However, balancing spatial and temporal resolution is a major challenge for real-time interventional MRI (i-MRI). Here, we proposed a deep unrolled neural network, dubbed as LSFP-Net, for real-time i-MRI reconstruction. By integrating LSFP-Net and a custom-designed, MR-compatible interventional device into a 3 T MRI scanner, a real-time MRI-guided brain intervention system is proposed. The performance of the system was evaluated using phantom and cadaver studies. 2D/3D real-time i-MRI was achieved with temporal resolutions of 80/732.8 ms, latencies of 0.4/3.66 s including data communication, processing and reconstruction time, and in-plane spatial resolution of 1 × 1 mm2. The results demonstrated that the proposed method enables real-time monitoring of the remote-controlled brain intervention, and showed the potential to be readily integrated into diagnostic scanners for image-guided neurosurgery.

Multi-site, Multi-domain Airway Tree Modeling

Open international challenges are becoming the de facto standard for assessing computer vision and image analysis algorithms. In recent years, new methods have extended the reach of pulmonary airway segmentation that is closer to the limit of image resolution. Since EXACT'09 pulmonary airway segmentation, limited effort has been directed to the quantitative comparison of newly emerged algorithms driven by the maturity of deep learning based approaches and extensive clinical efforts for resolving finer details of distal airways for early intervention of pulmonary diseases. Thus far, public annotated datasets are extremely limited, hindering the development of data-driven methods and detailed performance evaluation of new algorithms. To provide a benchmark for the medical imaging community, we organized the Multi-site, Multi-domain Airway Tree Modeling (ATM'22), which was held as an official challenge event during the MICCAI 2022 conference. ATM'22 provides large-scale CT scans with detailed pulmonary airway annotation, including 500 CT scans (300 for training, 50 for validation, and 150 for testing). The dataset was collected from different sites and it further included a portion of noisy COVID-19 CTs with ground-glass opacity and consolidation. Twenty-three teams participated in the entire phase of the challenge and the algorithms for the top ten teams are reviewed in this paper. Both quantitative and qualitative results revealed that deep learning models embedded with the topological continuity enhancement achieved superior performance in general. ATM'22 challenge holds as an open-call design, the training data and the gold standard evaluation are available upon successful registration via its homepage (https://atm22.grand-challenge.org/).

Central and peripheral analgesic active components of triterpenoid saponins from Stauntonia chinensis and their action mechanism

Triterpenoid saponins from Stauntonia chinensis have been proven to be a potential candidate for inflammatory pain relief. Our pharmacological studies confirmed that the analgesic role of triterpenoid saponins from S. chinensis occurred via a particular increase in the inhibitory synaptic response in the cortex at resting state and the modulation of the capsaicin receptor. However, its analgesic active components and whether its analgesic mechanism are limited to this are not clear. In order to further determine its active components and analgesic mechanism, we used the patch clamp technique to screen the chemical components that can increase inhibitory synaptic response and antagonize transient receptor potential vanilloid 1, and then used in vivo animal experiments to evaluate the analgesic effect of the selected chemical components. Finally, we used the patch clamp technique and molecular biology technology to study the analgesic mechanism of the selected chemical components. The results showed that triterpenoid saponins from S. chinensis could enhance the inhibitory synaptic effect and antagonize the transient receptor potential vanilloid 1 through different chemical components, and produce central and peripheral analgesic effects. The above results fully reflect that "traditional Chinese medicine has multi-component, multi-target, and multi-channel synergistic regulation".

MR Elastography With Optimization-Based Phase Unwrapping and Traveling Wave Expansion-Based Neural Network (TWENN)

Magnetic Resonance Elastography (MRE) can characterize biomechanical properties of soft tissue for disease diagnosis and treatment planning. However, complicated wavefields acquired from MRE coupled with noise pose challenges for accurate displacement extraction and modulus estimation. Using optimization-based displacement extraction and Traveling Wave Expansion-based Neural Network (TWENN) modulus estimation, we propose a new pipeline for processing MRE images. An objective function with Dual Data Consistency (Dual-DC) has been used to ensure accurate phase unwrapping and displacement extraction. For the estimation of complex wavenumbers, a complex-valued neural network with displacement covariance as an input has been developed. A model of traveling wave expansion is used to generate training datasets for the network with varying levels of noise. The complex shear modulus map is obtained through fusion of multifrequency and multidirectional data. Validation using brain and liver simulation images demonstrates the practical value of the proposed pipeline, which can estimate the biomechanical properties with minimal root-mean-square errors when compared to state-of-the-art methods. Applications of the proposed method for processing MRE images of phantom, brain, and liver reveal clear anatomical features, robustness to noise, and good generalizability of the pipeline.

Garbractin A, a Polycyclic Polyprenylated Acylphloroglucinol with a 4,11-dioxatricyclo[4.4.2.01,5]Dodecane Skeleton from Garcinia bracteata Fruits

Garbractin A (1), a structurally complicated polycyclic polyprenylated acylphloroglucinol (PPAP) with an unprecedented 4,11-dioxatricyclo[4.4.2.01,5] dodecane skeleton, was isolated from the fruits of Garcinia bracteata, along with five new biosynthetic analogues named garcibracteatones A-E (2-6). Their structures containing absolute configurations were revealed using spectroscopic data, the residual dipolar coupling-enhanced NMR approach, and quantum chemical calculations. The antihyperglycemic effect of these PPAPs (1-6) was evaluated using insulin-resistant HepG2 cells (IR-HepG2 cells) induced through palmitic acid (PA). Compounds 1, 3, and 4 were found to significantly promote glucose consumption in the IR-HepG2 cells and, therefore, may hold potential as candidates for treating hyperglycemia.

Contrastive Adversarial Learning for Endomicroscopy Imaging Super-Resolution

Endomicroscopy is an emerging imaging modality for real-time optical biopsy. One limitation of existing endomicroscopy based on coherent fibre bundles is that the image resolution is intrinsically limited by the number of fibres that can be practically integrated within the small imaging probe. To improve the image resolution, Super-Resolution (SR) techniques combined with image priors can enhance the clinical utility of endomicroscopy whereas existing SR algorithms suffer from the lack of explicit guidance from ground truth high-resolution (HR) images. In this article, we propose an unsupervised SR pipeline to allow stable offline and kernel-generic learning. Our method takes advantage of both internal statistics and external cross-modality priors. To improve the joint learning process, we present a Sharpness-aware Contrastive Generative Adversarial Network (SCGAN) with two dedicated modules, a sharpness-aware generator and a contrastive-learning discriminator. In the generator, an auxiliary task of sharpness discrimination is formulated to facilitate internal learning by considering the rankings of training instances in various sharpness levels. In the discriminator, we design a contrastive-learning module to mitigate the ill-posed nature of SR tasks via constraints from both positive and negative images. Experiments on multiple datasets demonstrate that SCGAN reduces the performance gap between previous unsupervised approaches and the upper bounds defined in supervised settings by more than 50%, delivering a new state-of-the-art performance score for endomicroscopy super-resolution. Further application on a realistic Voronoi-based pCLE downsampling kernel proves that SCGAN attains PSNR of 35.851 dB, improving 5.23 dB compared with the traditional Delaunay interpolation.

Navigation with minimal occupation volume for teleoperated snake-like surgical robots: MOVE

Master-Slave control is a common mode of operation for surgical robots as it ensures that surgeons are always in control and responsible for the procedure. Most teleoperated surgical systems use low degree-of-freedom (DOF) instruments, thus facilitating direct mapping of manipulator position to the instrument pose and tip location (tip-to-tip mapping). However, with the introduction of continuum and snake-like robots with much higher DOF supported by their inherent redundant architecture for navigating through curved anatomical pathways, there is a need for developing effective kinematic methods that can actuate all the joints in a controlled fashion. This paper introduces the concept of navigation with Minimal Occupation VolumE (MOVE), a teleoperation method that extends the concept of follow-the-leader navigation. It defines the path taken by the head while using all the available space surrounding the robot constrained by individual joint limits. The method was developed for the i ² Snake robot and validated with detailed simulation and control experiments. The results validate key performance indices such as path following, body weights, path weights, fault tolerance and conservative motion. The MOVE solver can run in real-time on a standard computer at frequencies greater than 1 kHz.

Microfluidic Approaches for Microactuators: From Fabrication, Actuation, to Functionalization

Microactuators can autonomously convert external energy into specific mechanical motions. With the feature sizes varying from the micrometer to millimeter scale, microactuators offer many operation and control possibilities for miniaturized devices. In recent years, advanced microfluidic techniques have revolutionized the fabrication, actuation, and functionalization of microactuators. Microfluidics can not only facilitate fabrication with continuously changing materials but also deliver various signals to stimulate the microactuators as desired, and consequently improve microfluidic chips with multiple functions. Herein, this cross-field that systematically correlates microactuator properties and microfluidic functions is comprehensively reviewed. The fabrication strategies are classified into two types according to the flow state of the microfluids: stop-flow and continuous-flow prototyping. The working mechanism of microactuators in microfluidic chips is discussed in detail. Finally, the applications of microactuator-enriched functional chips, which include tunable imaging devices, micromanipulation tools, micromotors, and microsensors, are summarized. The existing challenges and future perspectives are also discussed. It is believed that with the rapid progress of this cutting-edge field, intelligent microsystems may realize high-throughput manipulation, characterization, and analysis of tiny objects and find broad applications in various fields, such as tissue engineering, micro/nanorobotics, and analytical devices.

Deep anatomy learning for lung airway and artery-vein modeling with contrast-enhanced CT synthesis

PURPOSE

Endobronchial intervention requires detailed modeling of pulmonary anatomical substructure, such as lung airway and artery-vein maps, which are commonly extracted from non-contrast computed tomography (NCCT) independently using automatic segmentation approaches. We aim to make the first attempt to jointly train a CNN-based model for airway and artery-vein segmentation along with synthetic contrast-enhanced CT (CECT) generation.

METHODS

A multi-task framework is proposed to simultaneously generate three segmentation maps and synthesize CECTs. We first design a collaborative learning model with tissue knowledge interaction for lung airway and artery-vein segmentation. Meanwhile, a conditional adversarial training strategy is applied to generate CECTs from NCCTs guided by artery maps. Additionally, CECT identity and reconstruction help to regularize the model for plausible NCCT to CECT translation.

RESULTS

Extensive experiments were conducted to evaluate the performance of the proposed framework based on three datasets (90 NCCTs for the airway task, 55 NCCTs for the artery-vein task and 100 CECTs for the artery task). The results demonstrate the effective improvement of our proposed method compared to other methods and configurations that can achieve more accurate segmentation maps (Dice score coefficients for these three tasks are: 93.6%, 80.7% and 82.4%, respectively) and realistic CECTs simultaneously. The ablation study further verifies the effectiveness of the components of the designed model.

CONCLUSION

This study demonstrates the model potential in multi-task learning that integrates anatomically relevant segmentation and performs NCCT to CECT translation. Such an interaction approach promotes mutually for both promising segmentation results and plausible synthesis.

Assessment of vibration modulated regional cerebral blood flow with MRI

Human brain experiences vibration of certain magnitude and frequency during various physical activities such as vehicle transportation and machine operation, which may cause traumatic brain injury or other brain diseases. However, the mechanisms of brain pathogenesis due to vibration are not fully elucidated due to the lack of techniques to study brain functions while applying vibration to the brain at a specific magnitude and frequency. Here, this study reported a custom-built head-worn electromagnetic actuator that applied vibration to the brain in vivo at an accurate frequency inside a magnetic resonance imaging scanner while cerebral blood flow (CBF) was acquired. Using this technique, CBF values from 45 healthy volunteers were quantitatively measured immediately following vibration at 20, 30, 40 Hz, respectively. Results showed increasingly reduced CBF with increasing frequency at multiple regions of the brain, while the size of the regions expanded. Importantly, the vibration-induced CBF reduction regions largely fell inside the brain's default mode network (DMN), with about 58 or 46% overlap at 30 or 40 Hz, respectively. These findings demonstrate that vibration as a mechanical stimulus can change strain conditions, which may induce CBF reduction in the brain with regional differences in a frequency-dependent manner. Furthermore, the overlap between vibration-induced CBF reduction regions and DMN suggested a potential relationship between external mechanical stimuli and cognitive functions.

Trustworthy learning with (un)sure annotation for lung nodule diagnosis with CT

Recent evolution in deep learning has proven its value for CT-based lung nodule classification. Most current techniques are intrinsically black-box systems, suffering from two generalizability issues in clinical practice. First, benign-malignant discrimination is often assessed by human observers without pathologic diagnoses at the nodule level. We termed these data as "unsure-annotation data". Second, a classifier does not necessarily acquire reliable nodule features for stable learning and robust prediction with patch-level labels during learning. In this study, we construct a sure-annotation dataset with pathologically-confirmed labels and propose a collaborative learning framework to facilitate sure nodule classification by integrating unsure-annotation data knowledge through nodule segmentation and malignancy score regression. A loss function is designed to learn reliable features by introducing interpretability constraints regulated with nodule segmentation maps. Furthermore, based on model inference results that reflect the understanding from both machine and experts, we explore a new nodule analysis method for similar historical nodule retrieval and interpretable diagnosis. Detailed experimental results demonstrate that our approach is beneficial for achieving improved performance coupled with trustworthy model reasoning for lung cancer prediction with limited data. Extensive cross-evaluation results further illustrate the effect of unsure-annotation data for deep-learning based methods in lung nodule classification.

TNN: Tree Neural Network for Airway Anatomical Labeling

Detailed anatomical labeling of bronchial trees extracted from CT images can be used as fine-grained maps for intra-operative navigation. To cater to the sparse distribution of airway voxels and large class imbalance in 3D image space, a graph-neural-network-based method is proposed to map branches to nodes in a graph space and assign anatomical labels down to subsegmental level. To address the inherent problem of overlapping distribution of positional and morphological features, especially for subsegmental categories, the proposed method focuses on the relative position between sibling subsegments which is fixed in most cases. The hierarchical nomenclature is represented by multi-level labeling and each category is associated with one or two subtrees in the graph. Hyperedges are used to extract the representation of subtrees while a hypergraph neural network is developed to encode their intrinsic relationship through hyperedge interaction. A filter module is further designed to guide feature aggregation between nodes and hyperedges. With the proposed method, the final accuracies for segmental and subsegmental node classification can achieve 93.6% and 82.0% respectively. The corresponding code is publicly available at https://github.com/haozheng-sjtu/airway-labeling.

Vision-Kinematics Interaction for Robotic-Assisted Bronchoscopy Navigation

Endobronchial intervention is increasingly used as a minimally invasive means for the treatment of pulmonary diseases. In order to acquire the position of bronchoscopy, vision-based localization approaches are clinically preferable but are sensitive to visual variations. The static nature of pre-operative planning makes mapping of intraoperative anatomical features challenging for learning-based methods using visual features alone. In this work, we propose a robust navigation framework based on Vision Kinematic Interaction (VKI) for monocular bronchoscopic videos. To address visual-imbalance between the virtual and real views of bronchoscopy images, a Visual Similarity Network (VSN) is proposed to extract domain-invariant features to represent the lumen structure from endoscopic views, as well as domain-specific features to characterize the surface texture and visual artefacts. To improve the robustness of online estimation of camera pose, we also introduce a Kinematic Refinement Network (KRN) that allows progressive refinement of camera pose estimation based on network prediction and robot control signals. The accuracy of camera localization is validated on phantom and porcine lung datasets from a robotically controlled endobronchial intervention system, with both quantitative and qualitative results demonstrating the performance of the techniques. Results show that the features extracted by the proposed method can preserve the structural information of small airways in the presence of large visual variations along with the much-improved camera localization accuracy. The absolute trajectory errors (ATE) on phantom data and porcine data are 8.01 mm and 8.62 mm respectively.

Toward Robust Histology-Prior Embedding for Endomicroscopy Image Classification

Representation learning is the critical task for medical image analysis in computer-aided diagnosis. However, it is challenging to learn discriminative features due to the limited size of the dataset and the lack of labels. In this paper, we propose a stochastic routing normalization and neighborhood embedding framework with application to breast tissue classification by learning discriminative features of probe-based confocal laser endomicroscopy. In order to align the low-level and mid-level of pCLE and histology domain, we firstly build the domain-specific normalization module with stochastic activation strategy considering both depth-wise and feature-wise criterion. For high-level features, the latent centers are learned from the histology domain as the template for feature matching. The proposed method is evaluated on a clinical database with 700 pCLE mosaics. The accuracy of image classification with limited training samples demonstrates that the proposed method can outperform previous works on domain alignment.

Fabric computing: Concepts, opportunities, and challenges

With the advent of the Internet of Everything, people can easily interact with their environments immersively. The idea of pervasive computing is becoming a reality, but due to the inconvenience of carrying silicon-based entities and a lack of fine-grained sensing capabilities for human-computer interaction, it is difficult to ensure comfort, esthetics, and privacy in smart spaces. Motivated by the rapid developments in intelligent fabric technology in the post-Moore era, we propose a novel computing approach that creates a paradigm shift driven by fabric computing and advocate a new concept of non-chip sensing in living spaces. We discuss the core notion and benefits of fabric computing, including its implementation, challenges, and future research opportunities.

•

Fabric computing constructs a non-chip sensing with non-disturbance and ultra-dense structure

•

Multifunctional fibers obtain first-view sensory data; The value of sensory data will be distilled by intelligent fabric agents

•

Potential cognitive applications can be enabled by integrating fabric computing with AI

[The prognostic value of dynamic minimal residual disease after autologous hematopoietic stem cell transplantation in patients with multiple myeloma in novel-agent era]

Objective: To investigate the clinical prognostic value of dynamic minimal residual disease (MRD) after autologous hematopoietic stem cell transplantation (AHSCT) in patients with multiple myeloma (MM). Methods: Patients with MM who underwent AHSCT in Beijing Chao-Yang Hospital from February 2016 to December 2019 were enrolled in this study. All the patients in the study had complete baseline data at the diagnosis. AHSCT was performed after induction chemotherapy. Response evaluation was performed after induction therapy. All the patients were assessed at approximately 100 days after AHSCT. Bone marrow MRD by NGF was performed every three months and dynamically monitored for at least 12 months. All the patients were divided into different groups according to cytogenetics and MRD status. Survivals in different groups were analyzed by IBM SPSS 22.0 statistical software. Results: A total of 150 patients with MM were enrolled in this study at last, including 66 patients in the cytogenetic standard risk group and 84 patients in the cytogenetic high-risk group. The median age was 54 years (range 30-68 years) and 87 male patients (58.0%) was in the study. The median follow-up was 36 months (range 16-72 months). Patients in the standard-risk group had better clinical prognosis than those in the high-risk group [median PFS in the standard-risk group was not achieved, and median PFS in the high-risk group was 45 months (P<0.001); median OS of both groups was not reached, and the estimated 3-year OS rate of the standard-risk group and the high-risk group was 95.2% and 78.9%, respectively (P=0.001)]. According to MRD status of patients, patients in each group were divided into three subgroups: persistent positive (Ppos), transient negative (Tneg) and persistent negative (Pneg). The median OS and median PFS of all subgroups in the standard-risk group was not reached (P=0.324 and P=0.086). In high-risk group, the median OS of MRD Pneg subgroup was not reached, and the estimated 3-year OS rate was 100%; The median OS of MRD Ppos subgroup was 52 months, and MRD Tneg subgroup only 31 months (P=0.002); the median PFS of MRD Pneg group was not reached, and the estimated 3-year PFS rate was 85.4%; median PFS of MRD Ppos subgroup was 40 months, and MRD Tneg subgroup only 17 months (P=0.001). Conclusions: MRD Pneg might overcome the adverse prognosis of MM patients with high-risk cytogenetics. However, MRD Tneg might be a poor prognostic factor for the patients with cytogenetic high-risk MM.

Towards Occlusion-Aware Pose Estimation of Surgical Suturing Threads

OBJECTIVE

The technique of robust suture detection is vital in many applications including the skill evaluation for trainee, and suture augmentation in robotic-assisted surgery. Due to the complicated environment in surgery, the pose estimation of suture threads is challenged by the foreground and the background occlusion.

METHODS

To address this problem, we proposed an Occlusion-Aware Spatial Propagation model in this work. The challenging cases caused by self-intersection of threads are resolved by modeling the region connectivity. By taking the advantage of context-information, the background-occlusion is handled with the guided spatial propagation mechanism.

RESULTS

Experiments on phantom and ex-vivo datasets demonstrate the proposed method achieves superior accuracy on pose estimation compared to baseline methods, indicating the effectiveness of the occlusion-aware connectivity and spatial propagation.

CONCLUSION

Our proposed method provides a general framework for fully end-to-end pose estimation of suturing thread that achieves promising quality without the external simulation.

SIGNIFICANCE

Our fully automated algorithm addresses the occlusion problem including foreground and background occlusion which are common in surgery, and we anticipate that it will substantially provide the prior for future autonomy of robotic surgery.

Detection and assessment of Parkinson's disease based on gait analysis: A survey

Neurological disorders represent one of the leading causes of disability and mortality in the world. Parkinson's Disease (PD), for example, affecting millions of people worldwide is often manifested as impaired posture and gait. These impairments have been used as a clinical sign for the early detection of PD, as well as an objective index for pervasive monitoring of the PD patients in daily life. This review presents the evidence that demonstrates the relationship between human gait and PD, and illustrates the role of different gait analysis systems based on vision or wearable sensors. It also provides a comprehensive overview of the available automatic recognition systems for the detection and management of PD. The intervening measures for improving gait performance are summarized, in which the smart devices for gait intervention are emphasized. Finally, this review highlights some of the new opportunities in detecting, monitoring, and treating of PD based on gait, which could facilitate the development of objective gait-based biomarkers for personalized support and treatment of PD.

[Meta analysis of hearing loss caused by the combined effect of noise and heat in the working population]

Objective: To explore the effect of the combined effect of noise and heat on occupational hearing loss of workers by using Meta-analysis method. Methods: In August 2020, the Chinese and English literature on the relationship between exposure to noise and heat and occupational hearing loss published from January 2005 to August 2020 by CNKI, China Biomedical Literature Service System, Wanfang Data Knowledge Service Platform, VIP Official Database, Medline and PubMed Databases were searched, using noise, heat or hyperthermia, hearing as keywords. The selected data were analyzed by Stata 12.0 software, and the combined OR (95% CI) value included in the literature was calculated. Sensitivity analysis was used to explore the source of heterogeneity and analyze publication bias. Results: A total of 14 literatures (14 in Chinese, 0 in English) were included in the analysis, and 38654 subjects were included, including 6411 workers in the noise and heat combined effect group and 32243 workers in the noise alone group. The probability of hearing loss in the noise and heat combined effect group was 1.39 times higher than that in the noise alone group (95%CI: 1.14-1.69). The effect size OR was stable after sensitivity analysis, and there was no publication bias in the included literatures tested by Egger's and Begg's Method (z=0.38, P=0.702, t=-0.74, P=0.476) . Conclusion: Simultaneous exposure to noise and heat may increase the risk of hearing loss for workers in noisy workplaces.

Smart implanted access port catheter for therapy intervention with pH and lactate biosensors

Totally implanted access ports (TIAP) are widely used with oncology patients requiring long term central venous access for the delivery of chemotherapeutic agents, infusions, transfusions, blood sample collection and parenteral nutrition. Such devices offer a significant improvement to the quality of life for patients and reduced complication rates, particularly infection, in contrast to the classical central venous catheters. Nevertheless, infections do occur, with biofilm formation bringing difficulties to the treatment of infection-related complications that can ultimately lead to the explantation of the device. A smart TIAP device that is sensor-enabled to detect infection prior to extensive biofilm formation would reduce the cases for potential device explantation, whereas biomarkers detection within body fluids such as pH or lactate would provide vital information regarding metabolic processes occurring inside the body. In this paper, we propose a novel batteryless and wireless device suitable for the interrogation of such markers in an embodiment model of an TIAP, with miniature biochemical sensing needles. Device readings can be carried out by a smartphone equipped with Near Field Communication (NFC) interface at relative short distances off-body, while providing radiofrequency energy harvesting capability to the TIAP, useful for assessing patient's health and potential port infection on demand.

Robots at the Beijing 2022 Winter Olympics

The 2022 Winter Olympics has shown that robots can serve, protect, and help people excel in many aspects of their lives.

[The prognostic relationship between CD56 expression and newly diagnosed multiple myeloma]

Objective: To evaluate the prognostic value of CD56 expression in newly diagnosed MM (NDMM). Methods: A total of 332 NDMM patients were enrolled in Beijing Chaoyang Hospital, Capital Medical University from January 1, 2011 to January 1, 2021, with a median age of 60 years and a male to female ratio of 1.2∶1. CD56 expression on myeloma cells was detected by flow cytometry before induction therapy. Overall survival (OS) and progression-free survival (PFS) data were collected. In order to reduce the confounding factors, the propensity score matching technique was used to match CD56 positive versus negative patients at a ratio of 1∶1. Results: Among 332 patients, CD56 positivity rate was 65.1% (216/332). Patients with CD56 expression had significantly longer median OS (58.4 vs. 43.1 months, P=0.024) and PFS (28.7 vs. 24.1 months, P=0.013) than those with negative CD56. Univariate Cox proportional hazards regression analyses showed that CD56 expression was positively correlated with OS (HR=0.644, 95%CI 0.438-0.947, P=0.025) and a favorable prognostic factor for PFS (HR=0.646, 95%CI 0.457-0.913,P=0.013). The favorable effect of CD56 expression on PFS was confirmed in multivariate analysis (HR=0.705, 95%CI 0.497-0.998, P=0.049), but OS was not affected (P>0.05).In the propensity score matching analysis, 194 patients with 97 in each group were identified. CD56 positivity consistently predicted longer PFS (34.2 vs.25.1 months, P=0.047), but not OS (63.4 vs.43.1 months, P=0.056). Conclusion: These results demonstrate that CD56 expression is a favorable prognostic factor for PFS of newly diagnosed MM patients.

[Clinical trials on intrathecal pemetrexed treated leptomeningeal metastases from solid tumors]

Objective: To investigate the feasibility, safety and efficacy of intrathecal pemetrexed (IP) treated for patients with leptomeningeal metastases (LM) from solid tumors. Methods: Forty-seven patients receiving pemetrexed intrathecal chemotherapy in the First Hospital of Jilin University from 2017 to 2018 were selected. The study of pemetrexed intrathecal chemotherapy adopted the classical dose-climbing model and included 13 patients with meningeal metastasis of non-small cell lung cancer who had relapsed and refractory after multiple previous treatments including intrathecal chemotherapy. Based on the dose climbing study, 34 patients with meningeal metastasis of solid tumor who did not receive intrathecal chemotherapy were enrolled in a clinical study using pemetrexed as the first-line intrathecal chemotherapy combined with radiotherapy. Kaplan-Meier method and Log rank test were used for survival analysis, and Cox regression model was used for influencing factor analysis. Results: The dose climbing study showed that the maximum tolerated dose of pemetrexed intrathecal chemotherapy was 10 mg per single dose, and the recommended dosing regimen was 10 mg once or twice a week. The incidence of adverse reactions was 10 cases, including hematological adverse reactions (7 cases), transaminase elevation (2 cases), nerve root reactions (5 cases), fatigue and weight loss (1 case). The incidence of serious adverse reactions was 4, including grade 4-5 poor hematology (2 cases), grade 4 nerve root irritation (2 cases), and grade 4 elevated aminotransferase (1 case). In the dose climbing study, 4 patients were effectively treated and 7 were disease controlled. The survival time was ranged from 0.3 to 14.0 months and a median survival time was 3.8 months. The clinical study of pemetrexed intrathecal chemotherapy combined with radiotherapy showed that the treatment mode of 10 mg pemetrexed intrathecal chemotherapy once a week combined with synchronous involved area radiotherapy 40 Gy/4 weeks had a high safety and reactivity. The incidence of major adverse reactions was 52.9% (18/34), including hematologic adverse reactions (13 cases), transaminase elevation (10 cases), and nerve root reactions (4 cases). In study 2, the response rate was 67.6% (23/34), the disease control rate was 73.5% (25/34), the overall survival time was ranged from 0.3 to 16.6 months, the median survival time was 5.5 months, and the 1-year survival rate was 21.6%. Clinical response, improvement of neurological dysfunction, completion of concurrent therapy and subsequent systemic therapy were associated with the overall survival (all P<0.05). Conclusions: Pemetrexed is suitable for the intrathecal chemotherapy with a high safety and efficacy. The recommended administration regimen was IP at 10 mg on the schedule of once or twice per week. Hematological toxicity is the main factor affecting the implementation of IP. Vitamin supplement can effectively control the occurrence of hematological toxicity.

In-vivo Validation of a Novel Robotic Platform for Endovascular Intervention

OBJECTIVE

In-vivo validation on animal setting of a pneumatically propelled robot for endovascular intervention, to determine safety and clinical advantage of robotic cannulations compared to manual operation.

METHODS

Robotic assistance and image-guided intervention are increasingly used for improving endovascular procedures with enhanced navigation dexterity and accuracy. However, most platforms developed in the past decade still present inherent limitations in terms of altered clinical workflow, counterintuitive human-robot interaction, and a lack of versatility. We have created a versatile, highly integrated platform for robot-assisted endovascular intervention aimed at addressing such limitations, and here we demonstrate its clinical usability through in-vivo animal trials. A detailed in-vivo study on four porcine models conducted with our robotic platform is reported, involving cannulation and balloon angioplasty of five target arteries.

RESULTS

The trials showed a 100% success rate, and post-mortem histopathological assessment demonstrated a reduction in the incidence and severity of vessel trauma with robotic navigation versus manual manipulation.

CONCLUSION

In-vivo experiments demonstrated that the applicability of our robotic system within the context of this study was well tolerated, with good feasibility, and low risk profile. Comparable results were observed with robotics and manual cannulation, with clinical outcome potentially in favor of robotics.

SIGNIFICANCE

This study showed that the proposed robotic platform can potentially improve the execution of endovascular procedures, paving the way for clinical translation.

Five years of Science Robotics

[Figure: see text].

A decade retrospective of medical robotics research from 2010 to 2020

[Figure: see text].

A Single RGB Camera Based Gait Analysis With A Mobile Tele-Robot For Healthcare

With the increasing awareness of high-quality life, there is a growing need for health monitoring devices running robust algorithms in home environment. Health monitoring technologies enable real-time analysis of users' health status, offering long-term healthcare support and reducing hospitalization time. The propose of this work is twofold, the software focuses on the analysis of gait, which is widely adopted for joint correction and assessing any lower limb, or spinal problem. On the hardware side, a novel marker-less gait analysis device using a low-cost RGB camera mounted on a mobile tele-robot is designed. As gait analysis with a single camera is much more challenging compared to previous works utilizing multi-cameras, a RGB-D camera or wearable sensors, we propose using vision-based human pose estimation approaches. More specifically, based on the out-put of state-of-the-art human pose estimation models, we devise measurements for four bespoke gait parameters: inversion/eversion, dorsiflexion/plantarflexion, ankle and foot progression angles. We thereby classify walking patterns into normal, supination, pronation and limp. We also illustrate how to run the proposed machine learning models in low-resource environments such as a single entry-level CPU. Experiments show that our single RGB camera method achieves competitive performance compared to multi-camera motion capture systems, at smaller hardware costs.

An MR-Safe Endovascular Robotic Platform: Design, Control, and Ex-Vivo Evaluation

OBJECTIVE

Cardiovascular diseases are the most common cause of global death. Endovascular interventions, in combination with advanced imaging technologies, are promising approaches for minimally invasive diagnosis and therapy. More recently, teleoperated robotic platforms target improved manipulation accuracy, stabilisation of instruments in the vasculature, and reduction of patient recovery times. However, benefits of recent platforms are undermined by a lack of haptics and residual patient exposure to ionising radiation. The purpose of this research was to design, implement, and evaluate a novel endovascular robotic platform, which accommodates emerging non-ionising magnetic resonance imaging (MRI).

METHODS

We proposed a pneumatically actuated MR-safe teleoperation platform to manipulate endovascular instrumentation remotely and to provide operators with haptic feedback for endovascular tasks. The platform task performance was evaluated in an ex vivo cannulation study with clinical experts ( N = 7) under fluoroscopic guidance and haptic assistance on abdominal and thoracic phantoms.

RESULTS

The study demonstrated that the robotic dexterity involving pneumatic actuation concepts enabled successful remote cannulation of different vascular anatomies with success rates of 90%-100%. Compared to manual cannulation, slightly lower interaction forces between instrumentation and phantoms were measured for specific tasks. The maximum robotic interaction forces did not exceed 3N.

CONCLUSION

This research demonstrates a promising versatile robotic technology for remote manipulation of endovascular instrumentation in MR environments.

SIGNIFICANCE

The results pave the way for clinical translation with device deployment to endovascular interventions using non-ionising real-time 3D MR guidance.

Alleviating Class-Wise Gradient Imbalance for Pulmonary Airway Segmentation

Automated airway segmentation is a prerequisite for pre-operative diagnosis and intra-operative navigation for pulmonary intervention. Due to the small size and scattered spatial distribution of peripheral bronchi, this is hampered by a severe class imbalance between foreground and background regions, which makes it challenging for CNN-based methods to parse distal small airways. In this paper, we demonstrate that this problem is arisen by gradient erosion and dilation of the neighborhood voxels. During back-propagation, if the ratio of the foreground gradient to background gradient is small while the class imbalance is local, the foreground gradients can be eroded by their neighborhoods. This process cumulatively increases the noise information included in the gradient flow from top layers to the bottom ones, limiting the learning of small structures in CNNs. To alleviate this problem, we use group supervision and the corresponding WingsNet to provide complementary gradient flows to enhance the training of shallow layers. To further address the intra-class imbalance between large and small airways, we design a General Union loss function that obviates the impact of airway size by distance-based weights and adaptively tunes the gradient ratio based on the learning process. Extensive experiments on public datasets demonstrate that the proposed method can predict the airway structures with higher accuracy and better morphological completeness than the baselines.

Cross-Domain Self-Supervised Complete Geometric Representation Learning for Real-Scanned Point Cloud Based Pathological Gait Analysis

Accurate lower-limb pose estimation is a prerequisite of skeleton based pathological gait analysis. To achieve this goal in free-living environments for long-term monitoring, single depth sensor has been proposed in research. However, the depth map acquired from a single viewpoint encodes only partial geometric information of the lower limbs and exhibits large variations across different viewpoints. Existing off-the-shelf three-dimensional (3D) pose tracking algorithms and public datasets for depth based human pose estimation are mainly targeted at activity recognition applications. They are relatively insensitive to skeleton estimation accuracy, especially at the foot segments. Furthermore, acquiring ground truth skeleton data for detailed biomechanics analysis also requires considerable effort. To address these issues, we propose a novel cross-domain self-supervised complete geometric representation learning framework, with knowledge transfer from the unlabelled synthetic point clouds of full lower-limb surfaces. The proposed method can significantly reduce the number of ground truth skeletons (with only 1\%) in the training phase, meanwhile ensuring accurate and precise pose estimation and capturing discriminative features across different pathological gait patterns compared to other methods.

Urinary Bladder Volume Monitoring Using Magnetic Induction Tomography: A Rotational Simulation Model for Anatomical Slices within the Pelvic Region

Urinary bladder volume monitoring can benefit from contactless measurements, as alternative to the traditional medical methods of transurethral catheterization or ultrasound examination. The emerging modality of Magnetic Induction Tomography (MIT) offers the possibility for estimation of the intravesical volume in the physiological and pathological states using conductivity map reconstructions of the tissues present in the pelvic region. Within MIT, eddy currents originating from the conductive urine can produce their own magnetic field in response to an external magnetic source that is susceptible of being detected outside the body by means of a static ring of sensing coils. However, the ill-conditioned and ill-posed nature of the MIT Inverse Problem make the numerical implementation and conductivity estimation highly laborious. In this paper, we present a rotational frame model based on the MIT principles with application in urodynamic studies, which allows to extend the number of contactless measurements without increasing the overall dimension of the simulation domain, at the expense of solving multiple MIT Forward Problems. On the inversion process, the single-step Gauss-Newton method with Laplacian regularizer is recruited to estimate the bladder volume non-invasively and remotely (estimation error of 19%), paving the way for this technique to surpass the current limitations found in intravesical volume monitoring in quasi-real time.

[Autologous stem cell transplantation improve the survival of newly diagnosed multiple myeloma patients]

目的

评估自体造血干细胞移植（auto-HSCT）对初治多发性骨髓瘤（MM）疗效及生存的影响。

方法

回顾性分析2008年10月1日至2019年10月1日243例65岁以下接受auto-HSCT的初治MM患者，同时以同期176例≤65岁适合移植但未进行auto-HSCT的初治MM患者作为对照，评估auto-HSCT对患者疗效及生存的影响。为平衡auto-HSCT和非auto-HSCT患者之间各因素的分布，利用倾向性评分匹配技术按照1∶1比例匹配以减少组间的偏差。

结果

通过倾向性评分匹配分析，共筛选出128例患者（每组64例）。64例患者诱导治疗后接受auto-HSCT，24例（37.5％）获得严格意义的完全缓解（sCR），16例（25.0％）获得完全缓解（CR），15例（23.4％）获得非常好的部分缓解（VGPR），9例（14.1％）获得部分缓解（PR），auto-HSCT组疗效明显优于非auto-HSCT组（P＝0.032）。与非auto-HSCT组相比，auto-HSCT组总生存（OS）和无进展生存（PFS）期明显延长［OS：87.6（95％ CI 57.3～117.9）个月对53.9（95％ CI 36.1～71.7）个月，P＝0.011；PFS：42.2（95％ CI 29.9～54.5）个月对22.4（95％ CI 17.1～27.7）个月，P＝0.007］。多因素分析显示auto-HSCT是OS（HR＝0.448，95％CI 0.260～0.771，P＝0.004）和PFS（HR＝0.446，95％CI 0.280～0.778，P＝0.003）的独立保护因素。

结论

auto-HSCT可改善适合移植初治MM患者的OS和PFS。

Progress in robotics for combating infectious diseases

The world was unprepared for the COVID-19 pandemic, and recovery is likely to be a long process. Robots have long been heralded to take on dangerous, dull, and dirty jobs, often in environments that are unsuitable for humans. Could robots be used to fight future pandemics? We review the fundamental requirements for robotics for infectious disease management and outline how robotic technologies can be used in different scenarios, including disease prevention and monitoring, clinical care, laboratory automation, logistics, and maintenance of socioeconomic activities. We also address some of the open challenges for developing advanced robots that are application oriented, reliable, safe, and rapidly deployable when needed. Last, we look at the ethical use of robots and call for globally sustained efforts in order for robots to be ready for future outbreaks.

How could robotics help establish a new norm after COVID-19?

In a time of upheaval, robotics has an opportunity to offer long-term solutions and radical change.

Learning Tubule-Sensitive CNNs for Pulmonary Airway and Artery-Vein Segmentation in CT

Training convolutional neural networks (CNNs) for segmentation of pulmonary airway, artery, and vein is challenging due to sparse supervisory signals caused by the severe class imbalance between tubular targets and background. We present a CNNs-based method for accurate airway and artery-vein segmentation in non-contrast computed tomography. It enjoys superior sensitivity to tenuous peripheral bronchioles, arterioles, and venules. The method first uses a feature recalibration module to make the best use of features learned from the neural networks. Spatial information of features is properly integrated to retain relative priority of activated regions, which benefits the subsequent channel-wise recalibration. Then, attention distillation module is introduced to reinforce representation learning of tubular objects. Fine-grained details in high-resolution attention maps are passing down from one layer to its previous layer recursively to enrich context. Anatomy prior of lung context map and distance transform map is designed and incorporated for better artery-vein differentiation capacity. Extensive experiments demonstrated considerable performance gains brought by these components. Compared with state-of-the-art methods, our method extracted much more branches while maintaining competitive overall segmentation performance. Codes and models are available at http://www.pami.sjtu.edu.cn/News/56.

Systems pharmacology-based dissection of mechanisms of Tibetan medicinal compound Ruteng as an effective treatment for collagen-induced arthritis rats

ETHNOPHARMACOLOGICAL RELEVANCE

Compound Ruteng (CRT) is a prescribed formulation based on the theory of Tibetan medicine for the treatment of yellow-water-disease. It is consisted with 7 medicinal material include Boswellia carterii Birdw (named "Ruxiang" in Chinese); Tinospora sinensis (Lour.) Merr. (named "Kuan-Jin-Teng" in Chinese), Cassia obtusifolia L (named "Jue-Ming-Zi" in Chinese); Abelmoschus manihot (L.) Medic (named "Huang-Kui-Zi" in Chinese); Terminalia chebula Retz. (named "He-Zi" in Chinese); Lamiophlomis rotata (Benth.) Kudo (named "Du-Yi-Wei" in Chinese) and Pyrethrum tatsienense (Bur. et Franch.) Ling (named "Da-Jian-Ju" in Chinese). They are widely distributed in Tibet area of China and have been used to treat rheumatism, jaundice, and skin diseases for centuries.

AIM OF THE STUDY

The present study was conducted to investigate the anti-arthritis effect of CRT and to disclose the systems pharmacology-based dissection of mechanisms.

MATERIALS AND METHODS

The chemical constituents in CRT were identified using HPLC method, and CRT candidate targets against RA were screened by network pharmacology-based analysis and further experimentally validated based on collagen-induced arthritis (CIA) rat model. Furthermore, therapeutic mechanisms and pathways of CRT were investigated.

RESULTS

391 potential targets (protein) were predicted against 92 active ingredients of 7 medicinal materials in CRT. Enrichment analysis and molecular docking studies also enforced the practiced results. X-ray based physiological imaging showed the attenuated effect of CRT on paw swelling, synovial joints and cartilage with improved inflammation in CIA rats. Moreover, the expression of biomarkers associated with RA such as MMP1, MMP3 and MMP13 and TNF-a, COX2 and iNOS are down-regulated in ankle joints, serum, or liver.

CONCLUSION

In conclusion, CRT compound could attenuate RA symptoms and active ingredients of this compound could be considered for drug designing to treat RA.

MCDCD: Multi-Source Unsupervised Domain Adaptation for Abnormal Human Gait Detection

For gait analysis, especially for the detection of subtle gait abnormalities, the collected datasets involve high variability across subjects due to inherent biometric traits and movement behaviors, leading to limited detection accuracy and poor generalizability. To address this, we propose a novel deep multi-source Unsupervised Domain Adaptation (UDA) approach, namely Maximum Cross-Domain Classifier Discrepancy (MCDCD), which aims to improve the classification performance on the test subject (target domain) by leveraging the information from multiple labelled training subjects (source domains). Specifically, the proposed model consists of a feature extractor and a domain-specific category classifier per source domain. The former feature extractor learns to generate discriminative gait features. For the latter classifiers, we minimize the cross-entropy loss to accurately classify source samples, and simultaneously maximize a novel cross-domain discrepancy loss between any two category classifiers to minimize domain shift between multiple sources and the target domain. To validate the proposed MCDCD for detecting gait abnormalities on novel subjects, we collected both high-quality Motion capture (Mocap) and noisy Electromyography (EMG) data from eighteen subjects with both normal and imitated abnormal gaits. Experiment results using both data modalities demonstrate that the proposed approach can achieve superior performance in abnormal gait classification compared to baseline deep models and state-of-the-art UDA methods.