An end-to-end multi-scale airway segmentation framework based on pulmonary CT image.

OBJECTIVE: Automatic and accurate airway segmentation is necessary for lung disease diagnosis. The complex tree-like structures leads to gaps in the different generations of the airway tree, and thus airway segmentation is also considered to be a multi-scale problem. In recent years, convolutional neural networks have facilitated the development of medical image segmentation. In particular, 2D CNNs and 3D CNNs can extract different scale features. Hence, we propose a two-stage and 2D+3D framework for multi-scale airway tree segmentation. APPROACH: In stage 1, we use a 2D Full Airway SegNet(2D FA-SegNet) to segment the complete airway tree. Multi-scale Atros Spatial Pyramid (MASP) and Atros Residual Skip connection (ARSc) modules are inserted to extract different scales feature. We designed a hard sample selection strategy to increase the proportion of intrapulmonary airway samples in stage 2. 3D Airway RefineNet (3D ARNet) as stage 2 takes the results of stage 1 as a priori information. Spatial information extracted by 3D convolutional kernel compensates for the loss of in 2D FA-SegNet. Furthermore, we added False Positive losses and False Negative losses to improve the segmentation performance of airway branches within the lungs. MAIN RESULTS: We performed data enhancement on the publicly available dataset of ISICDM 2020 Challenge 3, and on which evaluated our method. Comprehensive experiments show that the proposed method has the highest DSC of 0.931, and IoU of 0.871 for the whole airway tree and DSC of 0.699, and IoU of 0.543 for the intrapulmonary bronchi tree. In addition, 3D ARNet proposed in this paper cascaded with other State-Of-The-Art methods to increase DLR by up to 46.33% and DBR by up to 42.97%. SIGNIFICANCE: The quantitative and qualitative evaluation results show that our proposed method performs well in segmenting the airway at different scales.

A bimodal feature fusion convolutional neural network for detecting obstructive sleep apnea/hypopnea from nasal airflow and oximetry signals.

The most prevalent sleep-disordered breathing condition is Obstructive Sleep Apnea (OSA), which has been linked to various health consequences, including cardiovascular disease (CVD) and even sudden death. Therefore, early detection of OSA can effectively help patients prevent the diseases induced by it. However, many existing methods have low accuracy in detecting hypopnea events or even ignore them altogether. According to the guidelines provided by the American Academy of Sleep Medicine (AASM), two modal signals, namely nasal pressure airflow and pulse oxygen saturation (SpO2), offer significant advantages in detecting OSA, particularly hypopnea events. Inspired by this notion, we propose a bimodal feature fusion CNN model that primarily comprises of a dual-branch CNN module and a feature fusion module for the classification of 10-second-long segments of nasal pressure airflow and SpO2. Additionally, an Efficient Channel Attention mechanism (ECA) is incorporated into the second module to adaptively weight feature map of each channel for improving classification accuracy. Furthermore, we design an OSA Severity Assessment Framework (OSAF) to aid physicians in effectively diagnosing OSA severity. The performance of both the bimodal feature fusion CNN model and OSAF is demonstrated to be excellent through per-segment and per-patient experimental results, based on the evaluation of our method using two real-world datasets consisting of polysomnography (PSG) recordings from 450 subjects.

Hierarchical classification of early microscopic lung nodule based on cascade network.

Purpose: Early-stage lung cancer is typically characterized clinically by the presence of isolated lung nodules. Thousands of cases are examined each year, and one case usually contains numerous lung CT slices. Detecting and classifying early microscopic lung nodules is demanding due to their diminutive dimensions and restricted characterization capabilities. Therefore, a lung nodule classification model that performs well and is sensitive to microscopic lung nodules is needed to accurately classify lung nodules. Methods: This paper uses the Resnet34 network as a basic classification model. A new cascade lung nodule classification method is proposed to classify lung nodules into 6 classes instead of the traditional 2 or 4 classes. It can effectively classify six different nodule types including ground-glass and solid nodules, benign and malignant nodules, and nodules with predominantly ground-glass or solid components. Results: In this paper, the traditional multi-classification method and the cascade classification method proposed in this paper were tested using real lung nodule data collected in the clinic. The test results demonstrate that the cascade classification method in this study achieves an accuracy of 80.04%, outperforming the conventional multi-classification approach. Conclusions: Different from the existing methods for categorizing the benign and malignant nature of lung nodules, the approach presented in this paper can classify lung nodules into 6 categories more accurately. At the same time, This paper proposes a rapid, precise, and dependable approach for classifying six distinct categories of lung nodules, which increases the accuracy categorization compared with the traditional multivariate categorization method.

Fast-Swimming Soft Robotic Fish Actuated by Bionic Muscle.

Soft underwater swimming robots actuated by smart materials have unique advantages in exploring the ocean, such as low noise, high flexibility, and friendly environment interaction ability. However, most of them typically exhibit limited swimming speed and flexibility due to the inherent characteristics of soft actuation materials. The actuation method and structural design of soft robots are key elements to improve their motion performance. Inspired by the muscle actuation and swimming mechanism of natural fish, a fast-swimming soft robotic fish actuated by a bionic muscle actuator made of dielectric elastomer is presented. The results show that by controlling the two independent actuating units of a biomimetic actuator, the robotic fish can not only achieve continuous C-shaped body motion similar to natural fish but also have a large bending angle (maximum unidirectional angle is about 40°) and thrust force (peak thrust is about 14 mN). In addition, the coupling relationship between the swimming speed and actuating parameters of the robotic fish is established through experiments and theoretical analysis. By optimizing the control strategy, the robotic fish can demonstrate a fast swimming speed of 76 mm/s (0.76 body length/s), which is much faster than most of the reported soft robotic fish driven by nonbiological soft materials that swim in body and/or caudal fin propulsion mode. What's more, by applying programmed voltage excitation to the actuating units of the bionic muscle, the robotic fish can be steered along specific trajectories, such as continuous turning motions and an S-shaped routine. This study is beneficial for promoting the design and development of high-performance soft underwater robots, and the adopted biomimetic mechanisms, as well as actuating methods, can be extended to other various flexible devices and soft robots.

Steering Muscle-Based Bio-Syncretic Robot Through Bionic Optimized Biped Mechanical Design.

Bio-syncretic robots consisting of artificial structures and living muscle cells have attracted much attention owing to their potential advantages, such as high drive efficiency, miniaturization, and compatibility. Motion controllability, as an important factor related to the main performance of bio-syncretic robots, has been explored in numerous studies. However, most of the existing bio-syncretic robots still face challenges related to the further development of steerable kinematic dexterity. In this study, a bionic optimized biped fully soft bio-syncretic robot actuated by two muscle tissues and steered with a direction-controllable electric field generated by external circularly distributed multiple electrodes has been developed. The developed bio-syncretic robot could realize wirelessly steerable motion and effective transportation of microparticle cargo on artificial polystyrene and biological pork tripe surfaces. This study may provide an effective strategy for the development of bio-syncretic robots and other related studies, such as nonliving soft robot design and muscle tissue engineering.

Nanomechanical Analysis of Living Small Extracellular Vesicles to Identify Gastric Cancer Cell Malignancy Based on a Biomimetic Peritoneum.

Gastric cancer is one of the most prevalent digestive malignancies. The lack of effective in vitro peritoneal models has hindered the exploration of the potential mechanisms behind gastric cancer's peritoneal metastasis. An accumulating body of research indicates that small extracellular vesicles (sEVs) play an indispensable role in peritoneal metastasis of gastric cancer cells. In this study, a biomimetic peritoneum was constructed. The biomimetic model is similar to real peritoneum in internal microstructure, composition, and primary function, and it enables the recurrence of peritoneal metastasis process in vitro. Based on this model, the association between the mechanical properties of sEVs and the invasiveness of gastric cancer was identified. By performing nanomechanical analysis on sEVs, we found that the Young's modulus of sEVs can be utilized to differentiate between malignant clinical samples (ascites) and nonmalignant clinical samples (peritoneal lavage). Furthermore, patients' ascites-derived sEVs were verified to stimulate the mesothelial-to-mesenchymal transition, thereby promoting peritoneal metastasis. In summary, nanomechanical analysis of living sEVs could be utilized for the noninvasive diagnosis of malignant degree and peritoneal metastasis of gastric cancer. This finding is expected to contribute future treatments.

Exploring non-invasive precision treatment in non-small cell lung cancer patients through deep learning radiomics across imaging features and molecular phenotypes.

BACKGROUND: Accurate prediction of tumor molecular alterations is vital for optimizing cancer treatment. Traditional tissue-based approaches encounter limitations due to invasiveness, heterogeneity, and molecular dynamic changes. We aim to develop and validate a deep learning radiomics framework to obtain imaging features that reflect various molecular changes, aiding first-line treatment decisions for cancer patients. METHODS: We conducted a retrospective study involving 508 NSCLC patients from three institutions, incorporating CT images and clinicopathologic data. Two radiomic scores and a deep network feature were constructed on three data sources in the 3D tumor region. Using these features, we developed and validated the 'Deep-RadScore,' a deep learning radiomics model to predict prognostic factors, gene mutations, and immune molecule expression levels. FINDINGS: The Deep-RadScore exhibits strong discrimination for tumor molecular features. In the independent test cohort, it achieved impressive AUCs: 0.889 for lymphovascular invasion, 0.903 for pleural invasion, 0.894 for T staging; 0.884 for EGFR and ALK, 0.896 for KRAS and PIK3CA, 0.889 for TP53, 0.895 for ROS1; and 0.893 for PD-1/PD-L1. Fusing features yielded optimal predictive power, surpassing any single imaging feature. Correlation and interpretability analyses confirmed the effectiveness of customized deep network features in capturing additional imaging phenotypes beyond known radiomic features. INTERPRETATION: This proof-of-concept framework demonstrates that new biomarkers across imaging features and molecular phenotypes can be provided by fusing radiomic features and deep network features from multiple data sources. This holds the potential to offer valuable insights for radiological phenotyping in characterizing diverse tumor molecular alterations, thereby advancing the pursuit of non-invasive personalized treatment for NSCLC patients.

Non-invasive decision support for clinical treatment of non-small cell lung cancer using a multiscale radiomics approach.

BACKGROUND: Selecting therapeutic strategies for cancer patients is typically based on key target-molecule biomarkers that play an important role in cancer onset, progression, and prognosis. Thus, there is a pressing need for novel biomarkers that can be utilized longitudinally to guide treatment selection. METHODS: Using data from 508 non-small cell lung cancer (NSCLC) patients across three institutions, we developed and validated a comprehensive predictive biomarker that distinguishes six genotypes and infiltrative immune phenotypes. These features were analyzed to establish the association between radiological phenotypes and tumor genotypes/immune phenotypes and to create a radiological interpretation of molecular features. In addition, we assessed the sensitivity of the models by evaluating their performance at five different voxel intervals, resulting in improved generalizability of the proposed approach. FINDINGS: The radiomics model we developed, which integrates clinical factors and multi-regional features, outperformed the conventional model that only uses clinical and intratumoral features. Our combined model showed significant performance for EGFR, KRAS, ALK, TP53, PIK3CA, and ROS1 mutation status with AUCs of 0.866, 0.874, 0.902, 0.850, 0.860, and 0.900, respectively. Additionally, the predictive performance for PD-1/PD-L1 was 0.852. Although the performance of all models decreased to different degrees at five different voxel space resolutions, the performance advantage of the combined model did not change. CONCLUSIONS: We validated multiscale radiomic signatures across tumor genotypes and immunophenotypes in a multi-institutional cohort. This imaging-based biomarker offers a non-invasive approach to select patients with NSCLC who are sensitive to targeted therapies or immunotherapy, which is promising for developing personalized treatment strategies during therapy.

Self-supervised neural network-based endoscopic monocular 3D reconstruction method.

Based on deep learning, monocular visual 3D reconstruction methods have been applied in various conventional fields. In the aspect of medical endoscopic imaging, due to the difficulty in obtaining real information, self-supervised deep learning has always been a focus of research. However, current research on endoscopic 3D reconstruction is mainly conducted in laboratory environments, lacking experience in dealing with complex clinical surgical environments. In this work, we use an optical flow-based neural network to address the problem of inconsistent brightness between frames. Additionally, attention modules and inter-layer losses are introduced to tackle the complexity of endoscopic scenes in clinical surgeries. The attention mechanism allows the network to better focus on pixel texture details and depth differences, while the inter-layer losses supervise the network at different scales. We have established a complete monocular endoscopic 3D reconstruction framework and conducted quantitative experiments on a clinical dataset using the cross-correlation coefficient as a metric. Compared with other self-supervised methods, our framework can better simulate the mapping relationship between adjacent frames during endoscope motion. To validate the generalization performance of our framework, we tested the model trained on the clinical dataset on the SCARED dataset and achieved equally excellent results.

A new segment method for pulmonary artery and vein.

Accurate differentiation between pulmonary arteries and veins (A/V) holds pivotal importance in the realm of diagnosing and treating pulmonary ailments. This study presents a new approach that leverages grayscale differences between A/V. Distinctions are measured using median and mean grayscale values within the vessel area. Initially, adherent regions are removed based on vessel structure. The trunk regions are segmented using gray level information near the heart region of the lung boundary. Incorrectly segmented vessels are corrected based on connectivity. For distal lung vessels, a similar distance field is established using a graph-cut method. Experimental results show the algorithm's superior segmentation accuracy, achieving 97.26% compared to the CNN-based average accuracy of 91.67%. Error branches are more concentrated, aiding subsequent manual and automatic correction. This demonstrates the algorithm's effective segmentation of pulmonary A/V.

Uncover rock-climbing fish's secret of balancing tight adhesion and fast sliding for bioinspired robots.

The underlying principle of the unique dynamic adaptive adhesion capability of a rock-climbing fish (Beaufortia kweichowensis) that can resist a pull-off force of 1000 times its weight while achieving simultaneous fast sliding (7.83 body lengths per second (BL/S)) remains a mystery in the literature. This adhesion-sliding ability has long been sought for underwater robots. However, strong surface adhesion and fast sliding appear to contradict each other due to the need for high surface contact stress. The skillfully balanced mechanism of the tight surface adhesion and fast sliding of the rock-climbing fish is disclosed in this work. The Stefan force (0.1 mN/mm2) generated by micro-setae on pectoral fins and ventral fins leads to a 70 N/m2 adhesion force by conforming the overall body of the fish to a surface to form a sealing chamber. The pull-off force is neutralized simultaneously due to the negative pressure caused by the volumetric change of the chamber. The rock-climbing fish's micro-setae hydrodynamic interaction and sealing suction cup work cohesively to contribute to low friction and high pull-off-force resistance and can therefore slide rapidly while clinging to the surface. Inspired by this unique mechanism, an underwater robot is developed with incorporated structures that mimic the functionality of the rock-climbing fish via a micro-setae array attached to a soft self-adaptive chamber, a setup which demonstrates superiority over conventional structures in terms of balancing tight underwater adhesion and fast sliding.

Radiomics under 2D regions, 3D regions, and peritumoral regions reveal tumor heterogeneity in non-small cell lung cancer: a multicenter study.

PURPOSE: Lung cancer has significant genetic and phenotypic heterogeneity, leading to poor prognosis. Radiomic features have emerged as promising predictors of the tumor phenotype. However, the role of underlying information surrounding the cancer remains unclear. MATERIALS AND METHODS: We conducted a retrospective study of 508 patients with NSCLC from three institutions. Radiomics models were built using features from six tumor regions and seven classifiers to predict three prognostically significant tumor phenotypes. The models were evaluated and interpreted by the mean area under the receiver operating characteristic curve (AUC) under nested cross-validation and Shapley values. The best-performing predictive models corresponding to six tumor regions and three tumor phenotypes were identified for further comparative analysis. In addition, we designed five experiments with different voxel spacing to assess the sensitivity of the experimental results to the spatial resolution of the voxels. RESULTS: Our results demonstrated that models based on 2D, 3D, and peritumoral region features yielded mean AUCs and 95% confidence intervals of 0.759 and [0.747-0.771] for lymphovascular invasion, 0.889 and [0.882-0.896] for pleural invasion, and 0.839 and [0.829-0.849] for T-staging in the testing cohort, which was significantly higher than all other models. Similar results were obtained for the model combining the three regional features at five voxel spacings. CONCLUSION: Our study revealed the predictive role of the developed methods with multi-regional features for the preoperative assessment of prognostic factors in NSCLC. The analysis of different voxel spacing and model interpretability strengthens the experimental findings and contributes to understanding the biological significance of the radiological phenotype.

A Review of the Development and Challenges of Cell Mechanical Models.

Cell models can express a variety of cell information, including mechanical properties, electrical properties, and chemical properties. Through the analysis of these properties, we can fully understand the physiological state of cells. As such, cell modeling has gradually become a topic of great interest, and a number of cell models have been established over the last few decades. In this paper, the development of various cell mechanical models has been systematically reviewed. First, continuum theoretical models, which were established by ignoring cell structures, are summarized, including the cortical membrane droplet model, solid model, power series structure damping model, multiphase model, and finite element model. Next, microstructural models based on the structure and function of cells are summarized, including the tension integration model, porous solid model, hinged cable net model, porous elastic model, energy dissipation model, and muscle model. What's more, from multiple viewpoints, the strengths and weaknesses of each cell mechanical model have been analyzed in detail. Finally, the potential challenges and applications in the development of cell mechanical models are discussed. This paper contributes to the development of different fields, such as biological cytology, drug therapy, and bio-syncretic robots.

miR-144-3p represses hepatocellular carcinoma progression by affecting cell aerobic glycolysis via FOXK1.

Aerobic glycolysis is a unique mark of cancer cells, which enables therapeutic intervention in cancer. Forkhead box K1 (FOXK1) is a transcription factor that facilitates the progression of multiple cancers including hepatocellular carcinoma (HCC). Nevertheless, it is unclear whether or not FOXK1 can affect HCC cell glycolysis. This study attempted to study the effect of FOXK1 on HCC cell glycolysis. Expression of mature miRNAs and mRNAs, as well as clinical data, was downloaded from The Cancer Genome Atlas-Liver hepatocellular carcinoma (TCGA-LIHC) dataset. FOXK1 and miR-144-3p levels were assessed through quantitative real-time polymerase chain reaction (qRT-PCR) and western blot. Targeting of the relationship between miR-144-3p and FOXK1 was verified via a dual-luciferase assay. Pathway enrichment analysis of FOXK1 was performed by Gene Set Enrichment Analysis (GSEA). Cell function assays revealed the glycolytic ability, cell viability, migration, invasion, cell cycle, and apoptosis of HCC cells in each treatment group. Bioinformatics analysis suggested that FOXK1 was upregulated in tissues of HCC patients, while the upstream miR-144-3p was downregulated in tumour tissues. Dual-luciferase assay implied a targeting relationship between miR-144-3p and FOXK1. Cellular experiments implied that silencing FOXK1 repressed HCC cell glycolysis, which in turn inhibited the HCC malignant progression. Rescue assay confirmed that miR-144-3p repressed glycolysis in HCC cells by targeting FOXK1, and then repressed HCC malignant progression. miR-144-3p/FOXK1 axis repressed malignant progression of HCC via affecting the aerobic glycolytic process of HCC cells. miR-144-3p and FOXK1 have the potential to become new therapeutic targets for HCC, which provide new insights for HCC treatment.

COVID-19 lung infection segmentation from chest CT images based on CAPA-ResUNet.

Coronavirus disease 2019 (COVID-19) epidemic has devastating effects on personal health around the world. It is significant to achieve accurate segmentation of pulmonary infection regions, which is an early indicator of disease. To solve this problem, a deep learning model, namely, the content-aware pre-activated residual UNet (CAPA-ResUNet), was proposed for segmenting COVID-19 lesions from CT slices. In this network, the pre-activated residual block was used for down-sampling to solve the problems of complex foreground and large fluctuations of distribution in datasets during training and to avoid gradient disappearance. The area loss function based on the false segmentation regions was proposed to solve the problem of fuzzy boundary of the lesion area. This model was evaluated by the public dataset (COVID-19 Lung CT Lesion Segmentation Challenge-2020) and compared its performance with those of classical models. Our method gains an advantage over other models in multiple metrics. Such as the Dice coefficient, specificity (Spe), and intersection over union (IoU), our CAPA-ResUNet obtained 0.775 points, 0.972 points, and 0.646 points, respectively. The Dice coefficient of our model was 2.51% higher than Content-aware residual UNet (CARes-UNet). The code is available at https://github.com/malu108/LungInfectionSeg.

Electroactive Polymer-Based Soft Actuator with Integrated Functions of Multi-Degree-of-Freedom Motion and Perception.

Soft actuators have received extensive attention in the fields of soft robotics, biomedicine, and intelligence systems owing to their advantages of pliancy, silence, and essential safety. However, most existing soft actuators have only single actuation elements and lack sensing. Therefore, it is difficult for them to perform complex motions with multiple degrees of freedom (multi-DOFs) and high precision. This article reports a miniature columnar dielectric elastomer actuator (DEA) with multi-DOF actuation and sensing, which was fabricated with an electroactive polymer acrylic film (Very High Bond [VHB] acrylic film by 3M Company) and carbon black grease electrodes. The arrangement of the simulation electrodes on the VHB was optimized to realize multi-DOF actuation, and the sensing electrodes were configured on the outer part of the DEA to realize real-time sensing. The results showed that the soft actuator can achieve all-round actuation through the selective power of the stimulation electrodes with a controllable voltage. The maximum bending angle and axial strain of the actuator reached 50° and 13%, respectively. Moreover, the deformation modes, direction, and quantity could be precisely measured using the integrative sensing function. In addition, to demonstrate the advantages of the proposed actuator, a manipulator with multiple actuators was designed and controlled to realize different actions of screwing and grasping with sensing. This research is useful not only for the design of multifunctional soft actuators but also for the development of soft robots with flexible, complex, and precisely controllable motions.

Macrophage migration inhibitory factor MtMIF3 prevents the premature aging of Medicago truncatula nodules.

Macrophage migration inhibitory factor (MIF) is a proinflammatory cytokine involved in immune response in animals. However, the role of MIFs in plants such as Medicago truncatula, particularly in symbiotic nitrogen fixation, remains unclear. An investigation of M. truncatula-Sinorhizobium meliloti symbiosis revealed that MtMIF3 was mainly expressed in the nitrogen-fixing zone of the nodules. Silencing MtMIF3 using RNA interference (Ri) technology resulted in increased nodule numbers but higher levels of bacteroid degradation in the infected cells of the nitrogen-fixing zone, suggesting that premature aging was induced in MtMIF3-Ri nodules. In agreement with this conclusion, the activities of nitrogenase, superoxide dismutase and catalase were lower than those in controls, but cysteine proteinase activity was increased in nodulated roots at 28 days postinoculation. In contrast, the overexpression of MtMIF3 inhibited nodule senescence. MtMIF3 is localized in the plasma membrane, nucleus, and cytoplasm, where it interacts with methionine sulfoxide reductase B (MsrB), which is also localized in the chloroplasts of tobacco leaf cells. Taken together, these results suggest that MtMIF3 prevents premature nodule aging and protects against oxidation by interacting with MtMsrB.

An interactive game for rehabilitation based on real-time hand gesture recognition.

Currently, cardiovascular and cerebrovascular diseases have become serious global health problems related to their high incidence and fatality rate. Some patients with cardiovascular cerebro-cardiovascular diseases even may face motor or cognitive dysfunction after surgery. In recent years, human-computer interactive systems with artificial intelligence have become an important part of human well-being because they enable novel forms of rehabilitation therapies. We propose an interactive game utilizing real-time skeleton-based hand gesture recognition, which aims to assist rehabilitation exercises by improving the hand-eye coordination of the patients during a game-like experience. For this purpose, we propose a lightweight residual graph convolutional architecture for hand gesture recognition. Furthermore, we designed the whole system using the proposed gesture recognition module and some third-party modules. Finally, some participants were invited to test our system and most of them showed an improvement in their passing rate of the game during the test process.

Prospective clinical research of radiomics and deep learning in oncology: A translational review.

Radiomics and deep learning (DL) hold transformative promise and substantial and significant advances in oncology; however, most methods have been tested in retrospective or simulated settings. There is considerable interest in the biomarker validation, clinical utility, and methodological robustness of these studies and their deployment in real-world settings. This review summarizes the characteristics of studies, the level of prospective validation, and the overview of research on different clinical endpoints. The discussion of methodological robustness shows the potential for independent external replication of prospectively reported results. These in-depth analyses further describe the barriers limiting the translation of radiomics and DL into primary care options and provide specific recommendations regarding clinical deployment. Finally, we propose solutions for integrating novel approaches into the treatment environment to unravel the critical process of translating AI models into the clinical routine and explore strategies to improve personalized medicine.