Closed thoracic drainage can be performed using a steel-needle-guided chest tube to treat pleural effusion or pneumothorax in clinics. However, the puncture procedure during surgery is invisible, increasing the risk of surgical failure. Therefore, it is necessary to design a visualization system for closed thoracic drainage. Augmented reality (AR) technology can assist in visualizing the internal anatomical structure and determining the insertion point on the body surface. The structure of the currently used steel-needle-guided chest tube was modified by integrating it with an ultrafine diameter camera to provide real-time visualization of the puncture process. After simulation experiments, the overall registration error of the AR method was measured to be within (3.59±0.53) mm, indicating its potential for clinical application. The ultrafine diameter camera module and improved steel-needle-guided chest tube can timely reflect the position of the needle tip in the human body. A comparative experiment showed that video guidance could improve the safety of the puncture process compared to the traditional method. Finally, a qualitative evaluation of the usability of the system was conducted through a questionnaire. This system facilitates the visualization of closed thoracic drainage puncture procedure and provides an implementation scheme to enhance the accuracy and safety of the operative step, which is conducive to reducing the learning curve and improving the proficiency of the doctors.

An optimizing method for designing the wireless power receiving coil (RC) is proposed in this paper to address issues such as insufficient and fluctuating power supply in the near-infrared capsule robot. An electromagnetic and circuit analysis is conducted to establish the magnetic induction intensity and equivalent circuit models for the wireless power transmission system. Combining these models involves using the number of layers in each dimension as the optimization variable. Constraints are imposed based on the normalized standard deviation of the receiving-end load power and spatial dimensions. At the same time, the optimization objective aims to maximize the average power of the receiving-end load. This process leads to formulating an optimization model for the RC. Finally, three-dimensional RCs with three different sets of parameters are wound, and the receiving-end load power of these coils is experimentally tested under various drive currents. The experimental values of the receiving-end load power exhibit a consistent trend with theoretical values, with experimental values consistently lower than theoretical values. The optimized coil parameters are determined by conducting comparative experiments, with a theoretical value of 4.6% for the normalized standard deviation of the receiving-end load power and an average experimental value of 9.6%. The study addressed the power supply issue of near-infrared capsule robots, which is important for early diagnosing and treating gastrointestinal diseases.

Irregular bone scaffolds fabricated using the Voronoi tessellation method resemble the morphology and properties of human cancellous bones. This has become a prominent topic in bone tissue engineering research in recent years. However, studies on the radial-gradient design of irregular bionic scaffolds are limited. Therefore, this study aims to develop a radial-gradient structure similar to that of natural long bones, enhancing the development of bionic bone scaffolds. A novel gradient method was adopted to maintain constant porosity, control the seed sitespecific distribution within the irregular porous structure, and vary the strut diameter to generate radial gradients. The irregular scaffolds were compared with four conventional scaffolds (cube, pillar BCC, vintiles, and diamond) in terms of permeability, stress concentration characteristics, and mechanical properties. The results indicate that the radial-gradient irregular porous structure boasts the widest permeability range and superior stress distribution compared to conventional scaffolds. With an elastic modulus ranging from 4.20 GPa to 22.96 GPa and a yield strength between 68.37 MPa and 149.40 MPa, it meets bone implant performance requirements and demonstrates significant application potential.

It is not easy to reduce the metal artifacts of computed tomography images. However, the pixel values inside the metal artifact regions vary smoothly, while those on the borders of the metal and the bone regions vary sharply. When the Canny operation by adaptive thresholding is conducted on the raw image, the almost continuous edges can be formed obviously on the borders of the metal and the bone regions, but this kind of information cannot be formed for the metal artifact regions. In this paper, by searching the closed areas formed by the border edges of the bone regions in the Canny image, the metal artifact regions, which are very difficult to discriminate only by intensity thresholding, can be excluded effectively. A novel prior image-based method is thus developed for metal artifact reduction. The experiments demonstrate that the proposed method can be realized easily and reduce the metal artifacts effectively even if multiple large metal objects exist simultaneously in the image. The method is suitable for the clinical application.

In the face of the large number of people with motor function disabilities, rehabilitation robots have attracted more and more attention. In order to promote the active participation of the user’s motion intention in the assisted rehabilitation process of the robots, it is crucial to establish the human motion prediction model. In this paper, a hybrid prediction model built on long short-term memory (LSTM) neural network using surface electromyography (sEMG) is applied to predict the elbow motion of the users in advance. This model includes two sub-models: a back-propagation neural network and an LSTM network. The former extracts a preliminary prediction of the elbow motion, and the latter corrects this prediction to increase accuracy. The proposed model takes time series data as input, which includes the sEMG signals measured by electrodes and the continuous angles from inertial measurement units. The offline and online tests were carried out to verify the established hybrid model. Finally, average root mean square errors of 3.52 ◦ and 4.18 ◦ were reached respectively for offline and online tests, and the correlation coefficients for both were above 0.98.

It is significant to process textures with special functions similar to animal surfaces based on bionics and improve the friction stability and contact comfort of contact surfaces for the surface texture design of tactile products. In this paper, a bionic hexagonal micro-convex texture was prepared on an acrylic surface by laser processing. The friction mechanism of a finger touching the bionic hexagonal micro-convex texture under different touch speeds and pressures, and the effect of the height of the texture on tactile perception were investigated by finite element, subjective evaluation, friction, and EEG tests. The results showed that the deformation friction was the main friction component when the finger touched the bionic hexagonal texture, and the slipperiness and friction factor showed a significant negative correlation. As the touch speed decreased or the touch force increased, the hysteresis friction of the fingers as well as the interlocking friction increased, and the slipperiness perception decreased. The bionic hexagonal texture with higher convexity caused a higher friction factor, lower slipperiness perception, and lower P300 peak. Hexagonal textures with lower convexity, lower friction factor, and higher slipperiness perception required greater brain attentional resources and intensity of tactile information processing during tactile perception.

The introduction of path planning and visual navigation in vascular interventional surgery can provide an intuitive reference and guidance for doctors. In this study, based on the preprocessing results of vessel skeleton extraction and stenosis diagnosis in X-ray coronary angiography images, clustering is used to determine the connectivity of the intersection points, and then the improved Dijkstra algorithm is used to automatically plan the surgical path. On this basis, the intermediate point is introduced to piecewise correct the path and improve the accuracy of the system. Finally, the epipolar constrained inverse projection transformation is used to reconstruct the coronary artery 3D model, and the optimal path is marked to achieve a multi-angle 3D visual navigation. Clinical experimental results show that compared with the traditional Dijkstra algorithm, the improved method can reduce the need for intermediate points, which improves computational efficiency, and the average error of manual calibration path is reduced to 4% of that before overall optimization. The results of 3D reconstruction and reprojection further qualitatively and quantitatively verify the effectiveness of the whole scheme.

The dynamic response characteristics of scoliosis and kyphosis to vibration are currently unclear. The finite element method (FEM) was employed to study the vibration response of patients with idiopathic scoliosis and kyphosis. The objective is to analyze the dynamic characteristics of idiopathic scoliosis and kyphosis using FEM. The finite element model of T1—S1 segments was established and verified using the CT scanning images. The established scoliosis and kyphosis models were verified statistically and dynamically. The finite element software Abaqus was utilized to analyze the mode, harmonic response, and transient dynamics of scoliosis and kyphosis. The first four natural frequencies extracted from modal analysis were 1.34, 2.26, 4.49 and 17.69 Hz respectively. Notably, the first three natural frequencies decreased with the increase of upper body mass. In harmonic response analysis, the frequency corresponding to the maximum amplitude in x direction was the first order natural frequency, and the frequency corresponding to the maximum amplitude in y and z directions was the second order natural frequency. At the same resonance frequency, the amplitude of the thoracic spine was larger relative to that of the lumbar spine. The time domain results of transient analysis showed that the displacement dynamic response of each segment presented cyclic response characteristics over time. Under 2.26Hz excitation, the dynamic response of the research object appeared as resonance. The higher the degree of spinal deformity, the greater the fundamental frequency. The first three natural modes of scoliosis and kyphosis contain vibration components in the vertical direction. The second order natural frequency was the most harmful to patients with scoliosis and kyphosis. Under cyclic loading, the deformation of the thoracic cone exceeds that of the lumbar cone.

The non-contact vital signs measurement technology based on millimeter wave radar has important medical value and unique advantages. However, because of its weak vibration characteristics, wide range of values, and the presence of respiratory harmonics and irrelevant motion interference in the detection signal, it is still difficult to perform a robust extraction in real time. To solve the above problems, the adaptive extraction of heart rates with a wide range of distribution is summarized as a multi-scale detection problem, and the distinction between heartbeat features and other irrelevant body motion features is summarized as a feature attention problem. Then, multi-scale detection module and heart rate feature attention module are designed and combined into a basic network module to build a heart rate extraction neural network. Through experiments based on properly designed datasets, a reasonable parameter design of the module is first explored. Experimental results show that in the signal data with unrelated motion data interference, average absolute error of the proposed method model for heart rate extraction can reach 1.87 beats/min, and average relative accuracy can reach 97.51%.

The structured low-rank model for parallel magnetic resonance (MR) imaging can efficiently reconstruct MR images with limited auto-calibration signals. To improve the reconstruction quality of MR images, we integrate the joint sparsity and sparsifying transform learning (JTL) into the simultaneous auto-calibrating and k-space estimation (SAKE) structured low-rank model, named JTLSAKE. The alternate direction method of multipliers is exploited to solve the resulting optimization problem, and the optimized gradient method is used to improve the convergence speed. In addition, a graphics processing unit is used to accelerate the proposed algorithm. The experimental results on four in vivo human datasets demonstrate that the reconstruction quality of the proposed algorithm is comparable to that of JTL-based low-rank modeling of local k-space neighborhoods with parallel imaging (JTL-PLORAKS), and the proposed algorithm is 46 times faster than the JTL-PLORAKS, requiring only 4 s to reconstruct a 200 × 200 pixels MR image with 8 channels.

Essential proteins are an indispensable part of cells and play an extremely significant role in genetic disease diagnosis and drug development. Therefore, the prediction of essential proteins has received extensive attention from researchers. Many centrality methods and machine learning algorithms have been proposed to predict essential proteins. Nevertheless, the topological characteristics learned by the centrality method are not comprehensive enough, resulting in low accuracy. In addition, machine learning algorithms need sufficient prior knowledge to select features, and the ability to solve imbalanced classification problems needs to be further strengthened. These two factors greatly affect the performance of predicting essential proteins. In this paper, we propose a deep learning framework based on temporal convolutional networks to predict essential proteins by integrating gene expression data and protein-protein interaction (PPI) network. We make use of the method of network embedding to automatically learn more abundant features of proteins in the PPI network. For gene expression data, we treat it as sequence data, and use temporal convolutional networks to extract sequence features. Finally, the two types of features are integrated and put into the multi-layer neural network to complete the final classification task. The performance of our method is evaluated by comparing with seven centrality methods, six machine learning algorithms, and two deep learning models. The results of the experiment show that our method is more effective than the comparison methods for predicting essential proteins.

Colorectal cancer is the most common cancer with a second mortality rate. Polyp lesion is a precursor symptom of colorectal cancer. Detection and removal of polyps can effectively reduce the mortality of patients in the early period. However, mass images will be generated during an endoscopy, which will greatly increase the workload of doctors, and long-term mechanical screening of endoscopy images will also lead to a high misdiagnosis rate. Aiming at the problem that computer-aided diagnosis models deeply depend on the computational power in the polyp detection task, we propose a lightweight model, coordinate attention-YOLOv5-Lite-Prune, based on the YOLOv5 algorithm, which is different from state-of-the-art methods proposed by the existing research that applied object detection models or their variants directly to prediction task without any lightweight processing, such as faster region-based convolutional neural networks, YOLOv3, YOLOv4, and single shot multibox detector. The innovations of our model are as follows: First, the lightweight EfficientNetLite network is introduced as the new feature extraction network. Second, the depthwise separable convolution and its improved modules with different attention mechanisms are used to replace the standard convolution in the detection head structure. Then, the α-intersection over union loss function is applied to improve the precision and convergence speed of the model. Finally, the model size is compressed with a pruning algorithm. Our model effectively reduces parameter amount and computational complexity without significant accuracy loss. Therefore, the model can be successfully deployed on the embedded deep learning platform, and detect polyps with a speed above 30 frames per second, which means the model gets rid of the limitation that deep learning models must rely on high-performance servers.

There is still a dearth of systematic study on picture stitching techniques for the natural tubular structures of intestines, and traditional stitching techniques have a poor application to endoscopic images with deep scenes. In order to recreate the intestinal wall in two dimensions, a method is developed. The normalized Laplacian algorithm is used to enhance the image and transform it into polar coordinates according to the characteristics that intestinal images are not obvious and usually arranged in a circle, in order to extract the new image segments of the current image relative to the previous image. The improved weighted fusion algorithm is then used to sequentially splice the segment images. The experimental results demonstrate that the suggested approach can improve image clarity and minimize noise while maintaining the information content of intestinal images. In addition, the method’s seamless transition between the final portions of a panoramic image also demonstrates that the stitching trace has been removed.

The pandemic SARS-CoV-2 has become an undying virus to spread a sustainable disease named COVID-19 for upcoming few years. Mortality rates are rising rapidly as approved drugs are not yet available. Isolation from the infected person or community is the preferred choice to protect our health. Since humans are the only carriers, it might be possible to control the positive rate if the infected population or host carriers are isolated from each other. Isolation alone may not be a proper solution. These are the resolutions of previous research work carried out on COVID-19 throughout the world. The present scenario of the world and public health is knocking hard with a big question of critical uncertainty of COVID-19 because of its imprecise database as per daily positive cases recorded all over the world and in India as well. In this research work, we have presented an optimal control model for COVID-19 using granular differentiability based on fuzzy dynamical systems. In the first step, we created a fuzzy Susceptible-Exposed-Infected-Asymptomatic-Hospitalized-Recovered-Death (SEIAHRD) model for COVID-19, analyzed it using granular differentiability, and reported disease dynamics for time-independent disease control parameters. In the second step, we upgraded the fuzzy dynamical system and granular differentiability model related to time-dependent disease control parameters as an optimal control problem invader. Theoretical studies have been validated with some practical data from the epidemic COVID-19 related to the Indian perspective during first wave and early second wave.

Sensitivity encoding (SENSE) is a parallel magnetic resonance imaging (MRI) reconstruction model by utilizing the sensitivity information of receiver coils to achieve image reconstruction. The existing SENSE-based reconstruction algorithms usually used nonadaptive sparsifying transforms, resulting in a limited reconstruction accuracy. Therefore, we proposed a new model for accurate parallel MRI reconstruction by combining the L0 norm regularization term based on the efficient sum of outer products dictionary learning (SOUPDIL) with the SENSE model, called SOUPDIL-SENSE. The SOUPDIL-SENSE model is mainly solved by utilizing the variable splitting and alternating direction method of multipliers techniques. The experimental results on four human datasets show that the proposed algorithm effectively promotes the image sparsity, eliminates the noise and artifacts of the reconstructed images, and improves the reconstruction accuracy.

Various and intricate varieties of lung disease have made it challenging for computer aided diagnosis to appropriately segment lung lesions utilizing computed tomography (CT) images. This study integrates transfer learning with the attention mechanism to construct a deep learning model that can automatically detect new coronary pneumonia on lung CT images. In this study, using VGG16 pre-trained by ImageNet as the encoder, the decoder was established utilizing the U-Net structure. The attention module is incorporated during each concatenate procedure, permitting the model to concentrate on the critical information and identify the crucial components efficiently. The public COVID-19-CT-Seg-Benchmark dataset was utilized for experiments, and the highest scores for Dice, F1, and Accuracy were 0.907 1, 0.907 6, and 0.996 5, respectively. The generalization performance was assessed concurrently, with performance metrics including Dice, F1, and Accuracy over 0.8. The experimental findings indicate the feasibility of the segmentation network proposed in this study.

Endotracheal intubation has broad application prospects in the biomedical field. At present, visual intubation tools are mainly used to judge the catheter position. However, when patients suffer from pains in the neck, throat, and trachea and other diseases or other conditions, if the exposure of the glottic area is not ideal, there are difficult airways. For difficult airways, this visual intubation tool has great limitations. Studying the new guidance method of endotracheal intubation and providing a reference or solution for difficult airway intubation is a crucial problem in the biomedical clinical field. In this paper, an endotracheal intubation method is proposed based on end-tidal carbon dioxide (ETCO2) perception. The simulation model verifies the feasibility of this method for endotracheal intubation guidance. Then, four micro-cavity tubes are used as a gas collection tube, and a set of endotracheal tube guidance systems based on ETCO2 perception is designed and developed to collect and process the CO2 concentration information in the pharyngeal cavity. The experimental results show that this guidance system can be used for intubation guidance in the simulated pharyngeal cavity without vision. Keywords: end-tidal carbon dioxide, no vision endotracheal intubation, micro-cavity tube

In order to improve the reconstruction accuracy of magnetic resonance imaging (MRI), an accurate natural image compressed sensing (CS) reconstruction network is proposed, which combines the advantages of model-based and deep learning-based CS-MRI methods. In theory, enhancing geometric texture details in linear reconstruction is possible. First, the optimization problem is decomposed into two problems: linear approximation and geometric compensation. Aimed at the problem of image linear approximation, the data consistency module is used to deal with it. Since the processing process will lose texture details, a neural network layer that explicitly combines image and frequency feature representation is proposed, which is named butterfly dilated geometric distillation network. The network introduces the idea of butterfly operation, skillfully integrates the features of image domain and frequency domain, and avoids the loss of texture details when extracting features in a single domain. Finally, a channel feature fusion module is designed by combining channel attention mechanism and dilated convolution. The attention of the channel makes the final output feature map focus on the more important part, thus improving the feature representation ability. The dilated convolution enlarges the receptive field, thereby obtaining more dense image feature data. The experimental results show that the peak signal-tonoise ratio of the network is 5.43 dB, 5.24 dB and 3.89 dB higher than that of ISTA-Net+, FISTA and DGDN networks on the brain data set with a Cartesian sampling mask CS ratio of 10%.

The proposed mass model of vocal fold vibration holds a significant importance in the auxiliary diagnosis and treatment of human vocal fold disorders. Mathematical models are proposed in aerodynamics and acoustics to simulate vocal fold vibration during phonation. This has always been a hot topic in pathological linguistics research. Over the past few decades, researchers have designed various types of mass models of vocal fold vibration based on experiments. These models differ in principles, computational complexity, and degrees of freedom. Therefore, we classify and describe the mass models according to modeling methods. We summarize the research status and characteristics of different models, and based on this, we look forward to future research directions for vocal fold mass models.

With the rapid development of machine learning, the prediction of the performance of acoustic metamaterials using neural networks is replacing the traditional experiment-based testing methods. In this paper, a Gini impurity-based artificial neural network structural optimizer (GIASO) is proposed to optimize the neural network structure, and the effects of five different initialization algorithms on the model performance and structure optimization are investigated. Two physically guided models with additional resonant frequencies and sound transmission loss formula are achieved to further improve the prediction accuracy of the model. The results show that GIASO utilizing the gray wolf optimizer as the initialization method can significantly improve the prediction performance of the model. Simultaneously, the physical guidance model with additional resonant frequencies has the best performance and can better predict the edge data points. Eventually, the effect of each input parameter on the sound transmission loss is explained by combining sensitivity analysis and theoretical formulation.