An augmented hybrid multibaseline and referenceless MR thermometry motion compensation algorithm for MRgHIFU hyperthermia

PURPOSE

A hybrid principal component analysis and projection onto dipole fields (PCA-PDF) MR thermometry motion compensation algorithm was optimized with atlas image augmentation and validated.

METHODS

Experiments were conducted on a 3T Philips MRI and Profound V1 Sonalleve high intensity focused ultrasound (high intensity focused ultrasound system. An MR-compatible robot was configured to induce motion on custom gelatin phantoms. Trials with periodic and sporadic motion were introduced on phantoms while hyperthermia was administered. The PCA-PDF algorithm was augmented with a predictive atlas to better compensate for larger sporadic motion.

RESULTS

During periodic motion, the temperature SD in the thermometry was improved from1 . 1 ± 0 . 1 $$ 1.1\pm 0.1 $$ to0 . 5 ± 0 . 1 ∘ $$ 0.5\pm 0.{1}^{\circ } $$ C with both the original and augmented PCA-PDF application. For large sporadic motion, the augmented atlas improved the motion compensation from the original PCA-PDF correction from8 . 8 ± 0 . 5 $$ 8.8\pm 0.5 $$ to0 . 7 ± 0 . 1 ∘ $$ 0.7\pm 0.{1}^{\circ } $$ C.

CONCLUSION

The PCA-PDF algorithm improved temperature accuracy to <1°C during periodic motion, but was not able to adequately address sporadic motion. By augmenting the PCA-PDF algorithm, temperature SD during large sporadic motion was also reduced to <1°C, greatly improving the original PCA-PDF algorithm.

Improved temporal resolution and acceleration on 4D-MR angiography based on superselective pseudo-continuous arterial spin labeling combined with CENTRA-keyhole and view-sharing (4D-S-PACK) using an interpolation algorithm on the temporal axis and compressed sensing-sensitivity encoding (CS-SENSE)

PURPOSE

Two major drawbacks of 4D-MR angiography based on superselective pseudo-continuous arterial spin labeling combined with CENTRA-keyhole and view-sharing (4D-S-PACK) are the low temporal resolution and long scanning time. We investigated the feasibility of increasing the temporal resolution and accelerating the scanning time on 4D-S-PACK by using CS-SENSE and PhyZiodynamics, a novel image-processing program that interpolates images between phases to generate new phases and reduces image noise.

METHODS

Seven healthy volunteers were scanned with a 3.0 T MR scanner to visualize the internal carotid artery (ICA) system. PhyZiodynamics is a novel image-processing that interpolates images between phases to generate new phases and reduces image noise, and by increasing temporal resolution using PhyZiodynamics, inflow dynamic data (reference) were acquired by changing the labeling durations (100-2000 msec, 31 phases) in 4D-S-PACK. From this set of data, we selected seven time intervals to calculate interpolated time points with up to 61 intervals using ×10 for the generation of interpolated phases with PhyZiodynamics. In the denoising process of PhyZiodynamics, we processed the none, low, medium, high noise reduction dataset images. The time intensity curve (TIC), the contrast-to-noise ratio (CNR) were evaluated. In accelerating with CS-SENSE for 4D-S-PACK, 4D-S-PACK were scanned different SENSE or CS-SENSE acceleration factors: SENSE3, CS3-6. Signal intensity (SI), CNR, were evaluated for accelerating the 4D-S-PACK. With regard to arterial vascular visualization, we evaluated the middle cerebral artery (MCA: M1-4 segments).

RESULTS

In increasing temporal resolution, the TIC showed a similar trend between the reference dataset and the interpolated dataset. As the noise reduction weight increased, the CNR of the interpolated dataset were increased compared to that of the reference dataset. In accelerating 4D-S-PACK, the SI values of the SENSE3 dataset and CS dataset with CS3-6 were no significant differences. The image noise increased with the increase of acceleration factor, and the CNR decreased with the increase of acceleration factor. Significant differences in CNR were observed between acceleration factor of SENSE3 and CS6 for the M1-4 (P < 0.05). Visualization of small arteries (M4) became less reliable in CS5 or CS6 images. Significant differences were found for the scores of M2, M3 and M4 segments between SENSE3 and CS6.

CONCLUSION

With PhyZiodynamics and CS-SENSE in 4D-S-PACK, we were able to shorten the scan time while improving the temporal resolution.

A novel predefined-time neurodynamic approach for mixed variational inequality problems and applications

In this paper, we propose a novel neurodynamic approach with predefined-time stability that offers a solution to address mixed variational inequality problems. Our approach introduces an adjustable time parameter, thereby enhancing flexibility and applicability compared to conventional fixed-time stability methods. By satisfying certain conditions, the proposed approach is capable of converging to a unique solution within a predefined-time, which sets it apart from fixed-time stability and finite-time stability approaches. Furthermore, our approach can be extended to address a wide range of mathematical optimization problems, including variational inequalities, nonlinear complementarity problems, sparse signal recovery problems, and nash equilibria seeking problems in noncooperative games. We provide numerical simulations to validate the theoretical derivation and showcase the effectiveness and feasibility of our proposed method.

QSPR for the prediction of critical micelle concentration of different classes of surfactants using machine learning algorithms

The determination of the critical micelle concentration (CMC) is a crucial factor when evaluating surfactants, making it an essential tool in studying the properties of surfactants in various industrial fields. In this present research, we assembled a comprehensive set of 593 different classes of surfactants including, anionic, cationic, nonionic, zwitterionic, and Gemini surfactants to establish a link between their molecular structure and the negative logarithmic value of critical micelle concentration (pCMC) utilizing quantitative structure-property relationship (QSPR) methodologies. Statistical analysis revealed that a set of 14 significant Mordred descriptors (SlogP, GATS6d, nAcid, GATS8dv, GATS4dv, PEOE_VSA11, GATS8d, ATS0p, GATS1d, MATS5p, GATS3d, NdssC, GATS6dv and EState_VSA4), along with temperature, served as appropriate inputs. Different machine learning methods, such as multiple linear regression (MLR), random forest regression (RFR), artificial neural network (ANN), and support vector regression (SVM), were employed in this study to build QSPR models. According to the statistical coefficients of QSPR models, SVR with Dragonfly hyperparameter optimization (SVR-DA) was the most accurate in predicting pCMC values, achieving (R2 = 0.9740, Q2 = 0.9739, r‾m2 = 0.9627, and Δrm2 = 0.0244) for the entire dataset.

Enhanced deep unrolling networks for snapshot compressive hyperspectral imaging

Snapshot compressive hyperspectral imaging necessitates the reconstruction of a complete hyperspectral image from its compressive snapshot measurement, presenting a challenging inverse problem. This paper proposes an enhanced deep unrolling neural network, called EDUNet, to tackle this problem. The EDUNet is constructed via the deep unrolling of a proximal gradient descent algorithm and introduces two innovative modules for gradient-driven update and proximal mapping reflectivity. The gradient-driven update module leverages a memory-assistant descent approach inspired by momentum-based acceleration techniques, for enhancing the unrolled reconstruction process and improving convergence. The proximal mapping is modeled by a sub-network with a cross-stage spectral self-attention, which effectively exploits the inherent self-similarities present in hyperspectral images along the spectral axis. It also enhances feature flow throughout the network, contributing to reconstruction performance gain. Furthermore, we introduce a spectral geometry consistency loss, encouraging EDUNet to prioritize the geometric layouts of spectral curves, leading to a more precise capture of spectral information in hyperspectral images. Experiments are conducted using three benchmark datasets including KAIST, ICVL, and Harvard, along with some real data, comprising a total of 73 samples. The experimental results demonstrate that EDUNet outperforms 15 competing models across four metrics including PSNR, SSIM, SAM, and ERGAS.

Optimizing BP neural network algorithm for Pericarpium Citri Reticulatae (Chenpi) origin traceability based on computer vision and ultra-fast gas-phase electronic nose data fusion

This study utilized computer vision to extract color and texture features of Pericarpium Citri Reticulatae (PCR). The ultra-fast gas-phase electronic nose (UF-GC-E-nose) technique successfully identified 98 volatile components, including olefins, alcohols, and esters, which significantly contribute to the flavor profile of PCR. Multivariate statistical Analysis was applied to the appearance traits of PCR, identifying 57 potential marker-trait factors (VIP > 1 and P < 0.05) from the 118 trait factors that can distinguish PCR from different origins. These factors include color, texture, and odor traits. By integrating multivariate statistical Analysis with the BP neural network algorithm, a novel artificial intelligence algorithm was developed and optimized for traceability of PCR origin. This algorithm achieved a 100% discrimination rate in differentiating PCR samples from various origins. This study offers a valuable reference and data support for developing intelligent algorithms that utilize data fusion from multiple intelligent sensory technologies to achieve rapid traceability of food origins.

MEF-UNet: An end-to-end ultrasound image segmentation algorithm based on multi-scale feature extraction and fusion

Ultrasound image segmentation is a challenging task due to the complexity of lesion types, fuzzy boundaries, and low-contrast images along with the presence of noises and artifacts. To address these issues, we propose an end-to-end multi-scale feature extraction and fusion network (MEF-UNet) for the automatic segmentation of ultrasound images. Specifically, we first design a selective feature extraction encoder, including detail extraction stage and structure extraction stage, to precisely capture the edge details and overall shape features of the lesions. In order to enhance the representation capacity of contextual information, we develop a context information storage module in the skip-connection section, responsible for integrating information from adjacent two-layer feature maps. In addition, we design a multi-scale feature fusion module in the decoder section to merge feature maps with different scales. Experimental results indicate that our MEF-UNet can significantly improve the segmentation results in both quantitative analysis and visual effects.

A pretrain-finetune approach for improving model generalizability in outcome prediction of acute respiratory distress syndrome patients

BACKGROUND

Early prediction of acute respiratory distress syndrome (ARDS) of critically ill patients in intensive care units (ICUs) has been intensively studied in the past years. Yet a prediction model trained on data from one hospital might not be well generalized to other hospitals. It is therefore essential to develop an accurate and generalizable ARDS prediction model adaptive to different hospital or medical centers.

METHODS

We analyzed electronic medical records of 200,859 and 50,920 hospitalized patients within 24 h after being diagnosed with ARDS from the Philips eICU Institute (eICU-CRD) and the Medical Information Mart for Intensive Care (MIMIC-IV) dataset, respectively. Patients were sorted into three groups, including rapid death, long stay, and recovery, based on their condition or outcome between 24 and 72 h after ARDS diagnosis. To improve prediction performance and generalizability, a "pretrain-finetune" approach was applied, where we pretrained models on the eICU-CRD dataset and performed model finetuning using only a part (35%) of the MIMIC-IV dataset, and then tested the finetuned models on the remaining data from the MIMIC-IV dataset. Well-known machine-learning algorithms, including logistic regression, random forest, extreme gradient boosting, and multilayer perceptron neural networks, were employed to predict ARDS outcomes. Prediction performance was evaluated using the area under the receiver-operating characteristic curve (AUC).

RESULTS

Results show that, in general, multilayer perceptron neural networks outperformed the other models. The use of pretrain-finetune yielded improved performance in predicting ARDS outcomes achieving a micro-AUC of 0.870 for the MIMIC-IV dataset, an improvement of 0.046 over the pretrain model.

CONCLUSIONS

The proposed pretrain-finetune approach can effectively improve model generalizability from one to another dataset in ARDS prediction.

Validation of a novel algorithm with a high specificity in ruling out MDS

INTRODUCTION

A previously published web-based App using Gradient-boosted models (GBMs) of eight laboratory parameters was established by Oster et al. to facilitate diagnosis or exclusion of myelodysplastic syndromes (MDS) in patients.

METHODS

To validate their algorithm, we compared 175 anemic patients with MDS diagnosis from our German MDS Registry with 1378 non-MDS anemic patients who consulted various specialties in the Düsseldorf university hospital.

RESULTS

Based on hemoglobin level, leukocyte and platelet count, mean corpuscular volume, absolute neutrophil count, absolute monocyte count, glucose and creatinine, plus the patients' gender and age, we could not reproduce a high negative predictive value (NPV), but confirmed a useful specificity of 90.9% and a positive predictive value (PPV) of 77.1%. 1192 of 1378 controls were correctly categorized as "probably not MDS (pnMDS)" patients. A total of 65 patients were wrongly classified as "probable MDS (pMDS)," of whom 48 had alternative explanations for their altered laboratory results. In a second analysis, we included 29 patients with chronic myelomonocytic leukemia (CMML) resulting in only one label as possible MDS, suggesting that highly proliferative bone marrow disorders are correctly excluded.

CONCLUSION

The possibility of reliably excluding MDS from differential diagnosis based on peripheral blood lab work appears to be attractive for patients and physicians alike while the confirmation of MDS diagnosis still requires a bone marrow biopsy.

Random, de novo, and conserved proteins: How structure and disorder predictors perform differently

Understanding the emergence and structural characteristics of de novo and random proteins is crucial for unraveling protein evolution and designing novel enzymes. However, experimental determination of their structures remains challenging. Recent advancements in protein structure prediction, particularly with AlphaFold2 (AF2), have expanded our knowledge of protein structures, but their applicability to de novo and random proteins is unclear. In this study, we investigate the structural predictions and confidence scores of AF2 and protein language model-based predictor ESMFold for de novo and conserved proteins from Drosophila and a dataset of comparable random proteins. We find that the structural predictions for de novo and random proteins differ significantly from conserved proteins. Interestingly, a positive correlation between disorder and confidence scores (pLDDT) is observed for de novo and random proteins, in contrast to the negative correlation observed for conserved proteins. Furthermore, the performance of structure predictors for de novo and random proteins is hampered by the lack of sequence identity. We also observe fluctuating median predicted disorder among different sequence length quartiles for random proteins, suggesting an influence of sequence length on disorder predictions. In conclusion, while structure predictors provide initial insights into the structural composition of de novo and random proteins, their accuracy and applicability to such proteins remain limited. Experimental determination of their structures is necessary for a comprehensive understanding. The positive correlation between disorder and pLDDT could imply a potential for conditional folding and transient binding interactions of de novo and random proteins.

Significance of neutrophil extracellular traps-related gene in the diagnosis and classification of atherosclerosis

Atherosclerosis (AS) is a pathological process associated with various cardiovascular diseases. Upon different stimuli, neutrophils release reticular complexes known as neutrophil extracellular traps (NETs). Numerous researches have indicated a strong correlation between NETs and AS. However, its role in cardiovascular disease requires further investigation. By utilizing a machine learning algorithm, we examined the genes associated with NETs that were expressed differently in individuals with AS compared to normal controls. As a result, we identified four distinct genes. A nomogram model was built to forecast the incidence of AS. Additionally, we conducted analysis on immune infiltration, functional enrichment and consensus clustering in AS samples. The findings indicated that individuals with AS could be categorized into two groups, exhibiting notable variations in immune infiltration traits among the groups. Furthermore, to measure the NETs model, the principal component analysis algorithm was developed and cluster B outperformed cluster A in terms of NETs. Additionally, there were variations in the expression of multiple chemokines between the two subtypes. By studying AS NETs, we acquired fresh knowledge about the molecular patterns and immune mechanisms implicated, which could open up new possibilities for AS immunotherapy.

BAGAIL: Multi-modal imitation learning from imbalanced demonstrations

Expert demonstrations in imitation learning often contain different behavioral modes, e.g., driving modes such as driving on the left, keeping the lane, and driving on the right in the driving tasks. Although most existing multi-modal imitation learning methods allow learning from demonstrations of multiple modes, they have strict constraints on the data of each mode, generally requiring a near data ratio of all modes. Otherwise, it tends to fall into a mode collapse or only learn the data distribution of the mode that has the largest data volume. To address the problem, an algorithm that balances real-fake loss and classification loss by modifying the output of the discriminator, referred to as BAlanced Generative Adversarial Imitation Learning (BAGAIL), is proposed. With this modification, the generator is only rewarded for generating real trajectories with correct modes. BAGAIL is therefore able to deal with imbalanced expert demonstrations and carry out efficient learning for each mode. The learning process of BAGAIL is divided into a pre-training stage and an imitation learning stage. During the pre-training stage, BAGAIL initializes the generator parameters by means of conditional Behavioral Cloning, laying the foundation for the direction of parameter optimization. During the imitation learning stage, BAGAIL optimizes the parameters by using the adversary between the generator and the modified discriminator so that the finally obtained policy can successfully learn the distribution of imbalanced expert data. The experiments showed that BAGAIL accurately distinguished different behavioral modes with imbalanced demonstrations. What is more, the learning result of each mode is close to the expert standard and more stable than other multi-modal imitation learning methods.

Set stabilization of logical control networks: A minimum node control approach

In network systems, control using minimum nodes or pinning control can be effectively used for stabilization problems to cut down the cost of control. In this paper, we investigate the set stabilization problem of logical control networks. In particular, we study the set stabilization problem of probabilistic Boolean networks (PBNs) and probabilistic Boolean control networks (PBCNs) via controlling minimal nodes. Firstly, an algorithm is given to search for the minimum index set of pinning nodes. Then, based on the analysis of its high computational complexity, we present optimized algorithms with lower computational complexity to ascertain the network control using minimum node sets. Moreover, some sufficient and necessary conditions are proposed to ensure the feasibility and effectiveness of the proposed algorithms. Furthermore, a theorem is presented for PBCNs to devise all state-feedback controllers corresponding to the set of pinning nodes. Finally, two models of gene regulatory networks are considered to show the efficacy of obtained results.

Development of a Probabilistic Boolean network (PBN) to model intraoperative blood pressure management

BACKGROUND

Blood pressure is a vital sign for organ perfusion that anesthesiologists measure and modulate during surgery. However, current decision-making processes rely heavily on clinicians' experience, which can lead to variability in treatment across surgeries. With the advent of machine learning, we can now create models to predict the outcomes of interventions and guide perioperative decision-making. The first step in this process involves translating the clinical decision-making process into a framework understood by an algorithm. Probabilistic Boolean networks (PBNs) provide an information-rich approach to this problem. A PBN trends toward a steady state, and its decisions are easily understood via its Boolean predictor functions. We hypothesize that a PBN can be developed that corrects hemodynamic instability in patients by selecting clinical interventions to maintain blood pressure within a given range.

METHODS

Data on patients over the age of 65 undergoing surgery with general anesthesia from 2018 to 2020 were drawn from the UF Health PRECEDE data set with IRB approval (IRB201700747). Parameters examined included heart rate, blood pressure, and frequency of medications given 15 min after anesthetic induction and 15 min before awakening. The medication frequency data were truncated into a 66/33 split for the training and validation set used in the PBN. The model was coded using Python 3 and evaluated by comparing the frequency of medications chosen by the program to the values in the testing set via linear regression analysis.

RESULTS

The network developed successfully models a hemodynamically unstable patient and corrects the imbalance by administering medications. This is evidenced by the model achieving a stable, steady state matrix in all iterations. However, the model's ability to emulate clinical drug selection was variable. It was successful with its use of vasodilator selection but struggled with the appropriate selection of vasopressors.

CONCLUSIONS

The PBN has demonstrated the ability to choose appropriate interventions based on a patient's current vitals. Additional work must be done to have the network emulate the frequency at which drugs are selected from in clinical practice. In its current state, the model provides an understanding of how a PBN behaves in the context of correcting hemodynamic instability and can aid in developing more robust models in the future.

The coefficient of conformism of a correlative prediction (CCCP): Building up reliable nano-QSPRs/QSARs for endpoints of nanoparticles in different experimental conditions encoded via quasi-SMILES

Simulation of the physicochemical and biochemical behavior of nanomaterials has its own specifics. However, the main goal of modeling for both traditional substances and nanomaterials is the same. This is an ecologic risk assessment. The universal indicator of toxicity is the n-octanol/water partition coefficient. Mutagenicity indicates the possibility of future undesirable environmental effects, possibly greater than toxicity. Models have been proposed for the octanol/water distribution coefficient of gold nanoparticles and the mutagenicity of silver nanoparticles. Unlike the previous studies, here the models are built using an updated scheme, which includes two improvements. Firstly, the computing involves a new criterion for prediction potential, the so-called coefficient of conformism of a correlative prediction (CCCP); secondly, the Las Vegas algorithm is used to select the potentially most promising models from a group of models obtained by the Monte Carlo algorithm. Apparently, CCCP is a measure of the predictive potential (not only correlation). This can give an advantage in developing a model in comparison to using the classic determination coefficient. Likely, CCCP can be more informative than the classical determination coefficient. The Las Vegas algorithm is able to improve the model obtained by the Monte Carlo method.

ASPTF: A computational tool to predict abiotic stress-responsive transcription factors in plants by employing machine learning algorithms

BACKGROUND

Abiotic stresses pose serious threat to the growth and yield of crop plants. Several studies suggest that in plants, transcription factors (TFs) are important regulators of gene expression, especially when it comes to coping with abiotic stresses. Therefore, it is crucial to identify TFs associated with abiotic stress response for breeding of abiotic stress tolerant crop cultivars.

METHODS

Based on a machine learning framework, a computational model was envisaged to predict TFs associated with abiotic stress response in plants. To numerically encode TF sequences, four distinct sequence derived features were generated. The prediction was performed using ten shallow learning and four deep learning algorithms. For prediction using more pertinent and informative features, feature selection techniques were also employed.

RESULTS

Using the features chosen by the light-gradient boosting machine-variable importance measure (LGBM-VIM), the LGBM achieved the highest cross-validation performance metrics (accuracy: 86.81%, auROC: 92.98%, and auPRC: 94.03%). Further evaluation of the proposed model (LGBM prediction method + LGBM-VIM selected features) was also done using an independent test dataset, where the accuracy, auROC and auPRC were observed 81.98%, 90.65% and 91.30%, respectively.

CONCLUSIONS

To facilitate the adoption of the proposed strategy by users, the approach was implemented as a prediction server called ASPTF, accessible at https://iasri-sg.icar.gov.in/asptf/. The developed approach and the corresponding web application are anticipated to supplement experimental methods in the identification of transcription factors (TFs) responsive to abiotic stress in plants.

Predictive models for sensory score and physicochemical composition of Yuezhou Longjing tea using near-infrared spectroscopy and data fusion

In this study, NIR quantitative prediction model was established for sensory score and physicochemical components of different varieties and quality grades of Yuezhou Longjing tea. Firstly, L, a, b color factors and diffuse reflection spectral data are collected for each sample. Subsequently, the original spectrum is preprocessed. Three techniques for selecting variables, CARS, BOSS, and SPA, were utilized to extract optimal feature bands. Finally, the spectral data extracted from feature bands were fused with L, a and b color factors to build SVR and PLSR prediction models. enabling the rapid non-destructive discrimination of different varieties and grades of Yuezhou Longjing tea. The outcomes demonstrated that BOSS was the best variable selection technique for sensory score and the distinctive caffeine wavelengths, CARS, however, was the best variable selection technique for catechins distinctive wavelengths. Additionally, the middle-level data fusion-based non-linear prediction models greatly outperformed the linear prediction models. For the prediction models of sensory score, catechins, and caffeine, the relative percent deviation (RPD) values were 2.8, 1.6, and 2.6, respectively, suggesting the good predictive ability of the models. In conclusion, evaluating the quality of the five Yuezhou Longjing tea varieties using near-infrared spectroscopy and data fusion have proved as feasible.

Phantom evaluation of electrical conductivity mapping by MRI: Comparison to vector network analyzer measurements and spatial resolution assessment

PURPOSE

To evaluate the performance of various MR electrical properties tomography (MR-EPT) methods at 3 T in terms of absolute quantification and spatial resolution limit for electrical conductivity.

METHODS

Absolute quantification as well as spatial resolution performance were evaluated on homogeneous phantoms and a phantom with holes of different sizes, respectively. Ground-truth conductivities were measured with an open-ended coaxial probe connected to a vector network analyzer (VNA). Four widely used MR-EPT reconstruction methods were investigated: phase-based Helmholtz (PB), phase-based convection-reaction (PB-cr), image-based (IB), and generalized-image-based (GIB). These methods were compared using the same complex images from a 1 mm-isotropic UTE sequence. Alternative transceive phase acquisition sequences were also compared in PB and PB-cr.

RESULTS

In large homogeneous phantoms, all methods showed a strong correlation with ground truth conductivities (r > 0.99); however, GIB was the best in terms of accuracy, spatial uniformity, and robustness to boundary artifacts. In the resolution phantom, the normalized root-mean-squared error of all methods grew rapidly (>0.40) when the hole size was below 10 mm, with simplified methods (PB and IB), or below 5 mm, with generalized methods (PB-cr and GIB).

CONCLUSION

VNA measurements are essential to assess the accuracy of MR-EPT. In this study, all tested MR-EPT methods correlated strongly with the VNA measurements. The UTE sequence is recommended for MR-EPT, with the GIB method providing good accuracy for structures down to 5 mm. Structures below 5 mm may still be detected in the conductivity maps, but with significantly lower accuracy.

Geometry-driven self-supervision for 3D human pose estimation

Although 3D human pose estimation has recently made strides, it is still difficult to precisely recreate a 3D human posture from a single image without the aid of 3D annotation for the following reasons. Firstly, the process of reconstruction inherently suffers from ambiguity, as multiple 3D poses can be projected onto the same 2D pose. Secondly, accurately measuring camera rotation without laborious camera calibration is a difficult task. While some approaches attempt to address these issues using traditional computer vision algorithms, they are not differentiable and cannot be optimized through training. This paper introduces two modules that explicitly leverage geometry to overcome these challenges, without requiring any 3D ground-truth or camera parameters. The first module, known as the relative depth estimation module, effectively mitigates depth ambiguity by narrowing down the possible depths for each joint to only two candidates. The second module, referred to as the differentiable pose alignment module, calculates camera rotation by aligning poses from different views. The use of these geometrically interpretable modules reduces the complexity of training and yields superior performance. By adopting our proposed method, we achieve state-of-the-art results on standard benchmark datasets, surpassing other self-supervised methods and even outperforming several fully-supervised approaches that heavily rely on 3D annotations.

Generalization analysis of deep CNNs under maximum correntropy criterion

Convolutional neural networks (CNNs) have gained immense popularity in recent years, finding their utility in diverse fields such as image recognition, natural language processing, and bio-informatics. Despite the remarkable progress made in deep learning theory, most studies on CNNs, especially in regression tasks, tend to heavily rely on the least squares loss function. However, there are situations where such learning algorithms may not suffice, particularly in the presence of heavy-tailed noises or outliers. This predicament emphasizes the necessity of exploring alternative loss functions that can handle such scenarios more effectively, thereby unleashing the true potential of CNNs. In this paper, we investigate the generalization error of deep CNNs with the rectified linear unit (ReLU) activation function for robust regression problems within an information-theoretic learning framework. Our study demonstrates that when the regression function exhibits an additive ridge structure and the noise possesses a finite pth moment, the empirical risk minimization scheme, generated by the maximum correntropy criterion and deep CNNs, achieves fast convergence rates. Notably, these rates align with the mini-max optimal convergence rates attained by fully connected neural network model with the Huber loss function up to a logarithmic factor. Additionally, we further establish the convergence rates of deep CNNs under the maximum correntropy criterion when the regression function resides in a Sobolev space on the sphere.

ARPruning: An automatic channel pruning based on attention map ranking

Structured pruning is a representative model compression technology for convolutional neural networks (CNNs), aiming to prune some less important filters or channels of CNNs. Most recent structured pruning methods have established some criteria to measure the importance of filters, which are mainly based on the magnitude of weights or other parameters in CNNs. However, these judgment criteria lack explainability, and it is insufficient to simply rely on the numerical values of the network parameters to assess the relationship between the channel and the model performance. Moreover, directly utilizing these pruning criteria for global pruning may lead to suboptimal solutions, therefore, it is necessary to complement search algorithms to determine the pruning ratio for each layer. To address these issues, we propose ARPruning (Attention-map-based Ranking Pruning), which reconstructs a new pruning criterion as the importance of the intra-layer channels and further develops a new local neighborhood search algorithm for determining the optimal inter-layer pruning ratio. To measure the relationship between the channel to be pruned and the model performance, we construct an intra-layer channel importance criterion by considering the attention map for each layer. Then, we propose an automatic pruning strategy searching method that can search for the optimal solution effectively and efficiently. By integrating the well-designed pruning criteria and search strategy, our ARPruning can not only maintain a high compression rate but also achieve outstanding accuracy. In our work, it is also experimentally concluded that compared with state-of-the-art pruning methods, our ARPruning method is capable of achieving better compression results. The code can be obtained at https://github.com/dozingLee/ARPruning.

Performance improvement of atherosclerosis risk assessment based on feature interaction

BACKGROUND AND OBJECTIVE

Cardiovascular disease is a leading cause of mortality and premature death. Early intervention in asymptomatic individuals through risk assessment can reduce the incidence of disease. Atherosclerosis is a major cause of cardiovascular disease and early detection can effectively prevent and treat it. In this study, we used real patient data to evaluate the risk of atherosclerosis, assisting doctors in diagnosis and reducing the incidence of cardiovascular disease.

METHODS

We proposed a multi-stage atherosclerosis risk assessment model that includes three main stages: (i) SMOTE and decorrelation weighting algorithm technology were added to the causal stability middle layer to address class imbalance in the dataset and reduce the impact of feature-induced dataset distribution shifts on model differences. (ii) The feature interaction layer considered possible feature interactions and classified features by different categories. By adding more effective feature information, the accuracy and generalizability of the model were improved. (iii) In the integrated model layer, we chose LightGBM as the decision tree integration model for risk assessment because it has higher accuracy and robustness compared to other machine learning algorithms.

RESULTS

The final model used a dataset containing 21 original features and 17 interaction features, achieving excellent performance under a 10-fold cross-validation strategy. The macro accuracy reached 93.86%, macro precision was 94.82%, macro recall was 93.52%, and macro F1 score was as high as 93.37%. These indicators demonstrate the accuracy and robustness of the model in atherosclerosis risk assessment.

CONCLUSION

The model provides strong support for the prevention and diagnosis of cardiovascular disease. Through atherosclerosis risk assessment, the model can help doctors develop personalized prevention and treatment plans, which is of great significance for the prevention and treatment of cardiovascular disease.

Porphyrin fluorescence imaging for real-time monitoring and visualization of the freshness of beef stored at different temperatures

This study presents a novel fluorescence imaging method for the real-time monitoring of beef quality deterioration and freshness. The fluorescence property of porphyrin in the form of heme can be used to characterize quality changes in beef during storage. Therefore, a fluorescence imaging system with an excitation light source of 440 nm and a CCD camera with a specific wavelength filter of 595 nm was constructed, and the porphyrin fluorescence images of beef samples stored at different temperatures were then collected. The quantitative model for predicting the microbial freshness indicator (TVC) of beef was built with the support vector machine regression (SVR) algorithm and produced satisfactory results with Rc2 and Rp2 of 0.858 and 0.812, respectively. The classification model based on the support vector machine (SVM) algorithm classified beef freshness into "fresh" and "spoiled", with calibration and prediction accuracy of 100 % and 90.9 %, respectively.

4 mC site recognition algorithm based on pruned pre-trained DNABert-Pruning model and fused artificial feature encoding

DNA 4 mC plays a crucial role in the genetic expression process of organisms. However, existing deep learning algorithms have shortcomings in the ability to represent DNA sequence features. In this paper, we propose a 4 mC site identification algorithm, DNABert-4mC, based on a fusion of the pruned pre-training DNABert-Pruning model and artificial feature encoding to identify 4 mC sites. The algorithm prunes and compresses the DNABert model, resulting in the pruned pre-training model DNABert-Pruning. This model reduces the number of parameters and removes redundancy from output features, yielding more precise feature representations while upholding accuracy.Simultaneously, the algorithm constructs an artificial feature encoding module to assist the DNABert-Pruning model in feature representation, effectively supplementing the information that is missing from the pre-trained features. The algorithm also introduces the AFF-4mC fusion strategy, which combines artificial feature encoding with the DNABert-Pruning model, to improve the feature representation capability of DNA sequences in multi-semantic spaces and better extract 4 mC sites and the distribution of nucleotide importance within the sequence. In experiments on six independent test sets, the DNABert-4mC algorithm achieved an average AUC value of 93.81%, outperforming seven other advanced algorithms with improvements of 2.05%, 5.02%, 11.32%, 5.90%, 12.02%, 2.42% and 2.34%, respectively.

Model-based iterative reconstruction for direct imaging with point spread function encoded echo planar MRI

BACKGROUND

Echo planar imaging (EPI) is a fast measurement technique commonly used in magnetic resonance imaging (MRI), but is highly sensitive to measurement non-idealities in reconstruction. Point spread function (PSF)-encoded EPI is a multi-shot strategy which alleviates distortion, but acquisition of encodings suitable for direct distortion-free imaging prolongs scan time. In this work, a model-based iterative reconstruction (MBIR) framework is introduced for direct imaging with PSF-EPI to improve image quality and acceleration potential.

METHODS

An MBIR platform was developed for accelerated PSF-EPI. The reconstruction utilizes a subspace representation, is regularized to promote local low-rankedness (LLR), and uses variable splitting for efficient iteration. Comparisons were made against standard reconstructions from prospectively accelerated PSF-EPI data and with retrospective subsampling. Exploring aggressive partial Fourier acceleration of the PSF-encoding dimension, additional comparisons were made against an extension of Homodyne to direct PSF-EPI in numerical experiments. A neuroradiologists' assessment was completed comparing images reconstructed with MBIR from retrospectively truncated data directly against images obtained with standard reconstructions from non-truncated datasets.

RESULTS

Image quality results were consistently superior for MBIR relative to standard and Homodyne reconstructions. As the MBIR signal model and reconstruction allow for arbitrary sampling of the PSF space, random sampling of the PSF-encoding dimension was also demonstrated, with quantitative assessments indicating best performance achieved through nonuniform PSF sampling combined with partial Fourier. With retrospective subsampling, MBIR reconstructs high-quality images from sub-minute scan datasets. MBIR was shown to be superior in a neuroradiologists' assessment with respect to three of five performance criteria, with equivalence for the remaining two.

CONCLUSIONS

A novel image reconstruction framework is introduced for direct imaging with PSF-EPI, enabling arbitrary PSF space sampling and reconstruction of diagnostic-quality images from highly accelerated PSF-encoded EPI data.

An evolutionary theory on virus mutation in COVID-19

With the rapid evolution of SARS-CoV-2, the emergence of new strains is an intriguing question. This paper presents an evolutionary theory to analyze the mutations of the virus and identify the conditions that lead to the generation of new strains. We represent the virus variants using a 4-letter sequence based on amino acid mutations on the spike protein and employ an n-distance algorithm to derive a variant phylogenetic tree. We show that the theoretically-derived tree aligns with experimental data on virus evolution. Additionally, we propose an A-X model, utilizing the set of existing mutation sites (A) and a set of randomly generated sites (X), to calculate the emergence of new strains. Our findings demonstrate that a sufficient number of random iterations can predict the generation of new macro-lineages when the number of sites in X is large enough. These results provide a crucial theoretical basis for understanding the evolution of SARS-CoV-2.

SecureNet: Proactive intellectual property protection and model security defense for DNNs based on backdoor learning

With the widespread application of deep neural networks (DNNs), the risk of privacy breaches against DNN models is constantly on the rise, resulting in an increasing need for intellectual property (IP) protection for such models. Although neural network watermarking techniques are widely used to safeguard the IP of DNNs, they can only achieve passive protection and cannot actively prevent unauthorized users from illicit use or embezzlement of the trained DNN models. Therefore, the development of proactive protection techniques to prevent IP infringement is imperative. To this end, we propose SecureNet, a key-based access license framework for DNN models. The proposed approach involves injecting license keys into the model through backdoor learning, enabling correct model functionality only when the appropriate license key is included in the input. To ensure the reusability of DNN models, we also propose a license key replacement algorithm. In addition, based on SecureNet, we designed defense mechanisms against adversarial attacks and backdoor attacks, respectively. Furthermore, we introduce a fine-grained authorization method that enables flexible granting of model permissions to different users. We have designed four license-key schemes with different privileges, tailored to various scenarios. We evaluated SecureNet on five benchmark datasets including MNIST, Cifar10, Cifar100, FaceScrub, and CelebA, and assessed its performance on six classic DNN models: LeNet-5, VGG16, ResNet18, ResNet101, NFNet-F5, and MobileNetV3. The results demonstrate that our approach outperforms the state-of-the-art model parameter encryption methods by at least 95% in terms of computational efficiency. Additionally, it provides effective defense against adversarial attacks and backdoor attacks without compromising the model's overall performance.

Communication-efficient distributed cubic Newton with compressed lazy Hessian

Recently, second-order distributed optimization algorithms have been becoming a research hot in distributed learning, due to their faster convergence rate than the first-order algorithms. However, second-order algorithms always suffer from serious communication bottleneck. To conquer such challenge, we propose communication-efficient second-order distributed optimization algorithms in the parameter-server framework, by incorporating cubic Newton methods with compressed lazy Hessian. Specifically, our algorithms require each worker communicate compressed Hessians with the server only at some particular iterations, which can save both communication bits and communication rounds. For non-convex problems, we theoretically prove that our algorithms can reduce the communication cost comparing to the state-of-the-art second-order algorithms, while maintaining the same iteration complexity order O(ϵ-3/2) as the centralized cubic Newton methods. By further using gradient regularization technique, our algorithms can achieve global convergence for convex problems. Moreover, for strongly convex problems, our algorithms can achieve local superlinear convergence rate without any requirement on initial conditions. Finally, numerical experiments are conducted to show the high efficiency of the proposed algorithms.

DWSSA: Alleviating over-smoothness for deep Graph Neural Networks

Graph Neural Networks (GNNs) have demonstrated great potential in achieving outstanding performance in various graph-related tasks, e.g., graph classification and link prediction. However, most of them suffer from the following issue: shallow networks capture very limited knowledge. Prior works design deep GNNs with more layers to solve the issue, which however introduces a new challenge, i.e., the infamous over-smoothness. Graph representation over emphasizes node features but only considers the static graph structure with a uniform weight are the key reasons for the over-smoothness issue. To alleviate the issue, this paper proposes a Dynamic Weighting Strategy (DWS) for addressing over-smoothness. We first employ Fuzzy C-Means (FCM) to cluster all nodes into several groups and get each node's fuzzy assignment, based on which a novel metric function is devised for dynamically adjusting the aggregation weights. This dynamic weighting strategy not only enables the intra-cluster interactions, but also inter-cluster aggregations, which well addresses undifferentiated aggregation caused by uniform weights. Based on DWS, we further design a Structure Augmentation (SA) step for addressing the issue of underutilizing the graph structure, where some potentially meaningful connections (i.e., edges) are added to the original graph structure via a parallelable KNN algorithm. In general, the optimized Dynamic Weighting Strategy with Structure Augmentation (DWSSA) alleviates over-smoothness by reducing noisy aggregations and utilizing topological knowledge. Extensive experiments on eleven homophilous or heterophilous graph benchmarks demonstrate the effectiveness of our proposed method DWSSA in alleviating over-smoothness and enhancing deep GNNs performance.

Warfarin-A natural anticoagulant: A review of research trends for precision medication

BACKGROUND

Warfarin is a widely prescribed anticoagulant in the clinic. It has a more considerable individual variability, and many factors affect its variability. Mathematical models can quantify the quantitative impact of these factors on individual variability.

PURPOSE

The aim is to comprehensively analyze the advanced warfarin dosing algorithm based on pharmacometrics and machine learning models of personalized warfarin dosage.

METHODS

A bibliometric analysis of the literature retrieved from PubMed and Scopus was performed using VOSviewer. The relevant literature that reported the precise dosage of warfarin calculation was retrieved from the database. The multiple linear regression (MLR) algorithm was excluded because a recent systematic review that mainly reviewed this algorithm has been reported. The following terms of quantitative systems pharmacology, mechanistic model, physiologically based pharmacokinetic model, artificial intelligence, machine learning, pharmacokinetic, pharmacodynamic, pharmacokinetics, pharmacodynamics, and warfarin were added as MeSH Terms or appearing in Title/Abstract into query box of PubMed, then humans and English as filter were added to retrieve the literature.

RESULTS

Bibliometric analysis revealed important co-occuring MeShH and index keywords. Further, the United States, China, and the United Kingdom were among the top countries contributing in this domain. Some studies have established personalized warfarin dosage models using pharmacometrics and machine learning-based algorithms. There were 54 related studies, including 14 pharmacometric models, 31 artificial intelligence models, and 9 model evaluations. Each model has its advantages and disadvantages. The pharmacometric model contains biological or pharmacological mechanisms in structure. The process of pharmacometric model development is very time- and labor-intensive. Machine learning is a purely data-driven approach; its parameters are more mathematical and have less biological interpretation. However, it is faster, more efficient, and less time-consuming. Most published models of machine learning algorithms were established based on cross-sectional data sourced from the database.

CONCLUSION

Future research on personalized warfarin medication should focus on combining the advantages of machine learning and pharmacometrics algorithms to establish a more robust warfarin dosage algorithm. Randomized controlled trials should be performed to evaluate the established algorithm of warfarin dosage. Moreover, a more user-friendly and accessible warfarin precision medicine platform should be developed.

Multi-scale characterization and analysis of cellular viscoelastic mechanical phenotypes by atomic force microscopy

The viscoelasticity of cells serves as a biomarker that reveals changes induced by malignant transformation, which aids the cytological examinations. However, differences in the measurement methods and parameters have prevented the consistent and effective characterization of the viscoelastic phenotype of cells. To address this issue, nanomechanical indentation experiments were conducted using an atomic force microscope (AFM). Multiple indentation methods were applied, and the indentation parameters were gradually varied to measure the viscoelasticity of normal liver cells and cancerous liver cells to create a database. This database was employed to train machine-learning algorithms in order to analyze the differences in the viscoelasticity of different types of cells and as well as to identify the optimal measurement methods and parameters. These findings indicated that the measurement speed significantly influenced viscoelasticity and that the classification difference between the two cell types was most evident at 5 μm/s. In addition, the precision and the area under the receiver operating characteristic curve were comparatively analyzed for various widely employed machine-learning algorithms. Unlike previous studies, this research validated the effectiveness of measurement parameters and methods with the assistance of machine-learning algorithms. Furthermore, the results confirmed that the viscoelasticity obtained from the multiparameter indentation measurement could be effectively used for cell classification. RESEARCH HIGHLIGHTS: This study aimed to analyze the viscoelasticity of liver cancer cells and liver cells. Different nano-indentation methods and parameters were used to measure the viscoelasticity of the two kinds of cells. The neural network algorithm was used to reverse analyze the dataset, and the methods and parameters for accurate classification and identification of cells are successfully found.

Multi-level feature fusion and joint refinement for simultaneous object pose estimation and camera localization

Object pose estimation and camera localization are critical in various applications. However, achieving algorithm universality, which refers to category-level pose estimation and scene-independent camera localization, presents challenges for both techniques. Although the two tasks keep close relationships due to spatial geometry constraints, different tasks require distinct feature extractions. This paper pays attention to a unified RGB-D based framework that simultaneously performs category-level object pose estimation and scene-independent camera localization. The framework consists of a pose estimation branch called SLO-ObjNet, a localization branch called SLO-LocNet, a pose confidence calculation process and object-level optimization. At the start, we obtain the initial camera and object results from SLO-LocNet and SLO-ObjNet. In these two networks, we design there-level feature fusion modules as well as the loss function to achieve feature sharing between two tasks. Then the proposed approach involves a confidence calculation process to determine the accuracy of object poses obtained. Additionally, an object-level Bundle Adjustment (BA) optimization algorithm is further used to improve the precision of these techniques. The BA algorithm establishes relationships among feature points, objects, and cameras with the usage of camera-point, camera-object, and object-point metrics. To evaluate the performance of this approach, experiments are conducted on localization and pose estimation datasets including REAL275, CAMERA25, LineMOD, YCB-Video, 7 Scenes, ScanNet and TUM RGB-D. The results show that this approach outperforms existing methods in terms of both estimation and localization accuracy. Additionally, SLO-LocNet and SLO-ObjNet are trained on ScanNet data and tested on 7 Scenes and TUM RGB-D datasets to demonstrate its universality performance. Finally, we also highlight the positive effects of fusion modules, loss function, confidence process and BA for improving overall performance.

Cosine convolutional neural network and its application for seizure detection

Traditional convolutional neural networks (CNNs) often suffer from high memory consumption and redundancy in their kernel representations, leading to overfitting problems and limiting their application in real-time, low-power scenarios such as seizure detection systems. In this work, a novel cosine convolutional neural network (CosCNN), which replaces traditional kernels with the robust cosine kernel modulated by only two learnable factors, is presented, and its effectiveness is validated on the tasks of seizure detection. Meanwhile, based on the cosine lookup table and KL-divergence, an effective post-training quantization algorithm is proposed for CosCNN hardware implementation. With quantization, CosCNN can achieve a nearly 75% reduction in the memory cost with almost no accuracy loss. Moreover, we design a configurable cosine convolution accelerator on Field Programmable Gate Array (FPGA) and deploy the quantized CosCNN on Zedboard, proving the proposed seizure detection system can operate in real-time and low-power scenarios. Extensive experiments and comparisons were conducted using two publicly available epileptic EEG databases, the Bonn database and the CHB-MIT database. The results highlight the performance superiority of the CosCNN over traditional CNNs as well as other seizure detection methods.

Deformable registration of preoperative MR and intraoperative long-length tomosynthesis images for guidance of spine surgery via image synthesis

PURPOSE

Improved integration and use of preoperative imaging during surgery hold significant potential for enhancing treatment planning and instrument guidance through surgical navigation. Despite its prevalent use in diagnostic settings, MR imaging is rarely used for navigation in spine surgery. This study aims to leverage MR imaging for intraoperative visualization of spine anatomy, particularly in cases where CT imaging is unavailable or when minimizing radiation exposure is essential, such as in pediatric surgery.

METHODS

This work presents a method for deformable 3D-2D registration of preoperative MR images with a novel intraoperative long-length tomosynthesis imaging modality (viz., Long-Film [LF]). A conditional generative adversarial network is used to translate MR images to an intermediate bone image suitable for registration, followed by a model-based 3D-2D registration algorithm to deformably map the synthesized images to LF images. The algorithm's performance was evaluated on cadaveric specimens with implanted markers and controlled deformation, and in clinical images of patients undergoing spine surgery as part of a large-scale clinical study on LF imaging.

RESULTS

The proposed method yielded a median 2D projection distance error of 2.0 mm (interquartile range [IQR]: 1.1-3.3 mm) and a 3D target registration error of 1.5 mm (IQR: 0.8-2.1 mm) in cadaver studies. Notably, the multi-scale approach exhibited significantly higher accuracy compared to rigid solutions and effectively managed the challenges posed by piecewise rigid spine deformation. The robustness and consistency of the method were evaluated on clinical images, yielding no outliers on vertebrae without surgical instrumentation and 3% outliers on vertebrae with instrumentation.

CONCLUSIONS

This work constitutes the first reported approach for deformable MR to LF registration based on deep image synthesis. The proposed framework provides access to the preoperative annotations and planning information during surgery and enables surgical navigation within the context of MR images and/or dual-plane LF images.

Referenceless Nyquist ghost correction outperforms standard navigator-based method and improves efficiency of in vivo diffusion tensor cardiovascular magnetic resonance

PURPOSE

The study aims to assess the potential of referenceless methods of EPI ghost correction to accelerate the acquisition of in vivo diffusion tensor cardiovascular magnetic resonance (DT-CMR) data using both computational simulations and data from in vivo experiments.

METHODS

Three referenceless EPI ghost correction methods were evaluated on mid-ventricular short axis DT-CMR spin echo and STEAM datasets from 20 healthy subjects at 3T. The reduced field of view excitation technique was used to automatically quantify the Nyquist ghosts, and DT-CMR images were fit to a linear ghost model for correction.

RESULTS

Numerical simulation showed the singular value decomposition (SVD) method is the least sensitive to noise, followed by Ghost/Object method and entropy-based method. In vivo experiments showed significant ghost reduction for all correction methods, with referenceless methods outperforming navigator methods for both spin echo and STEAM sequences at b = 32, 150, 450, and 600  smm - 2 $$ {\mathrm{smm}}^{-2} $$ . It is worth noting that as the strength of the diffusion encoding increases, the performance gap between the referenceless method and the navigator-based method diminishes.

CONCLUSION

Referenceless ghost correction effectively reduces Nyquist ghost in DT-CMR data, showing promise for enhancing the accuracy and efficiency of measurements in clinical practice without the need for any additional reference scans.

Registration of preoperative temporal bone CT-scan to otoendoscopic video for augmented-reality based on convolutional neural networks

PURPOSE

Patient-to-image registration is a preliminary step required in surgical navigation based on preoperative images. Human intervention and fiducial markers hamper this task as they are time-consuming and introduce potential errors. We aimed to develop a fully automatic 2D registration system for augmented reality in ear surgery.

METHODS

CT-scans and corresponding oto-endoscopic videos were collected from 41 patients (58 ears) undergoing ear examination (vestibular schwannoma before surgery, profound hearing loss requiring cochlear implant, suspicion of perilymphatic fistula, contralateral ears in cases of unilateral chronic otitis media). Two to four images were selected from each case. For the training phase, data from patients (75% of the dataset) and 11 cadaveric specimens were used. Tympanic membranes and malleus handles were contoured on both video images and CT-scans by expert surgeons. The algorithm used a U-Net network for detecting the contours of the tympanic membrane and the malleus on both preoperative CT-scans and endoscopic video frames. Then, contours were processed and registered through an iterative closest point algorithm. Validation was performed on 4 cases and testing on 6 cases. Registration error was measured by overlaying both images and measuring the average and Hausdorff distances.

RESULTS

The proposed registration method yielded a precision compatible with ear surgery with a 2D mean overlay error of 0.65 ± 0.60 mm for the incus and 0.48 ± 0.32 mm for the round window. The average Hausdorff distance for these 2 targets was 0.98 ± 0.60 mm and 0.78 ± 0.34 mm respectively. An outlier case with higher errors (2.3 mm and 1.5 mm average Hausdorff distance for incus and round window respectively) was observed in relation to a high discrepancy between the projection angle of the reconstructed CT-scan and the video image. The maximum duration for the overall process was 18 s.

CONCLUSIONS

A fully automatic 2D registration method based on a convolutional neural network and applied to ear surgery was developed. The method did not rely on any external fiducial markers nor human intervention for landmark recognition. The method was fast and its precision was compatible with ear surgery.

PDE-LEARN: Using deep learning to discover partial differential equations from noisy, limited data

In this paper, we introduce PDE-LEARN, a novel deep learning algorithm that can identify governing partial differential equations (PDEs) directly from noisy, limited measurements of a physical system of interest. PDE-LEARN uses a Rational Neural Network, U, to approximate the system response function and a sparse, trainable vector, ξ, to characterize the hidden PDE that the system response function satisfies. Our approach couples the training of U and ξ using a loss function that (1) makes U approximate the system response function, (2) encapsulates the fact that U satisfies a hidden PDE that ξ characterizes, and (3) promotes sparsity in ξ using ideas from iteratively reweighted least-squares. Further, PDE-LEARN can simultaneously learn from several data sets, allowing it to incorporate results from multiple experiments. This approach yields a robust algorithm to discover PDEs directly from realistic scientific data. We demonstrate the efficacy of PDE-LEARN by identifying several PDEs from noisy and limited measurements.

Distribution-free Bayesian regularized learning framework for semi-supervised learning

In machine learning it is often necessary to assume or know the distribution of the data, however it is difficult to do so in practical applications. Aiming to this problem, this work, we propose a novel distribution-free Bayesian regularized learning framework for semi-supervised learning, which is called Hessian regularized twin minimax probability extreme learning machine (HRTMPELM). In this framework, we attempt to construct two non-parallel hyperplanes by introducing the high separation probability assumption, such that each hyperplane separates samples from one class with maximum probability while moving away from samples from the other class. Subsidiently, the framework can be utilized to construct reasonable semi-supervised classifiers by using the information of the inherent geometric distribution of the samples through the Hessian regularization term. Additionally, the proposed framework controls the misclassification error of samples by minimizing the upper limit of the worst-case misclassification probability, and improves the generalization performance of the model by introducing the idea of regularization to avoid the occurrence of ill-posedness and overfitting problems. More importantly, the framework has no hyperparameters, making the learning process very simplified and efficient. Finally, a simple and reliable algorithm with globally optimal solutions via multivariate Chebyshev inequalities is designed for solving the proposed learning framework. Experiments on multiple datasets demonstrate the reliability and effectiveness of the proposed learning framework compared to other methods. Especially, we applied the framework to Ningxia wolfberry quality detection, which greatly enriches and facilitates the application of machine learning algorithms in the agricultural field.

Optimized self-attention based cycle-consistent generative adversarial network adopted melanoma classification from dermoscopic images

Skin is the exposed part of the human body that constantly protected from UV rays, heat, light, dust, and other hazardous radiation. One of the most dangerous illnesses that affect people is skin cancer. A type of skin cancer called melanoma starts in the melanocytes, which regulate the colour in human skin. Reducing the fatality rate from skin cancer requires early detection and diagnosis of conditions like melanoma. In this article, a Self-attention based cycle-consistent generative adversarial network optimized with Archerfish Hunting Optimization Algorithm adopted Melanoma Classification (SACCGAN-AHOA-MC-DI) from dermoscopic images is proposed. Primarily, the input Skin dermoscopic images are gathered via the dataset of ISIC 2019. Then, the input Skin dermoscopic images is pre-processed using adjusted quick shift phase preserving dynamic range compression (AQSP-DRC) for removing noise and increase the quality of Skin dermoscopic images. These pre-processed images are fed to the piecewise fuzzy C-means clustering (PF-CMC) for ROI region segmentation. The segmented ROI region is supplied to the Hexadecimal Local Adaptive Binary Pattern (HLABP) to extract the Radiomic features, like Grayscale statistic features (standard deviation, mean, kurtosis, and skewness) together with Haralick Texture features (contrast, energy, entropy, homogeneity, and inverse different moments). The extracted features are fed to self-attention based cycle-consistent generative adversarial network (SACCGAN) which classifies the skin cancers as Melanocytic nevus, Basal cell carcinoma, Actinic Keratosis, Benign keratosis, Dermatofibroma, Vascular lesion, Squamous cell carcinoma and melanoma. In general, SACCGAN not adapt any optimization modes to define the ideal parameters to assure accurate classification of skin cancer. Hence, Archerfish Hunting Optimization Algorithm (AHOA) is considered to maximize the SACCGAN classifier, which categorizes the skin cancer accurately. The proposed method attains 23.01%, 14.96%, and 45.31% higher accuracy and 32.16%, 11.32%, and 24.56% lesser computational time evaluated to the existing methods, like melanoma prediction method for unbalanced data utilizing optimized Squeeze Net through bald eagle search optimization (CNN-BES-MC-DI), hyper-parameter optimized CNN depending on Grey wolf optimization algorithm (CNN-GWOA-MC-DI), DEANN incited skin cancer finding depending on fuzzy c-means clustering (DEANN-MC-DI). RESEARCH HIGHLIGHTS: This manuscript, self-attention based cycle-consistent. SACCGAN-AHOA-MC-DI method is implemented in Python. (SACCGAN-AHOA-MC-DI) from dermoscopic images is proposed. Adjusted quick shift phase preserving dynamic range compression (AQSP-DRC). Removing noise and increase the quality of Skin dermoscopic images.

Bayesian optimization of gradient trajectory for parallel-transmit pulse design

PURPOSE

Spoke pulses improve excitation homogeneity in parallel-transmit MRI. We propose an efficient global optimization algorithm, Bayesian optimization of gradient trajectory (BOGAT), for single-slice and simultaneous multislice imaging.

THEORY AND METHODS

BOGAT adds an outer loop to optimize kT-space positions. For each position, the RF coefficients are optimized (e.g., with magnitude least squares) and the cost function evaluated. Bayesian optimization progressively estimates the cost function. It automatically chooses the kT-space positions to sample, to achieve fast convergence, often coming close to the globally optimal spoke positions. We investigated the typical features of spokes cost functions by a grid search with field maps comprising 85 slabs from 14 volunteers. We tested BOGAT in this database, and prospectively in a phantom and in vivo. We compared the vendor-provided Fourier transform approach with the same magnitude least squares RF optimizer.

RESULTS

The cost function is nonconvex and seen empirically to be piecewise smooth with discontinuities where the underlying RF optimum changes sharply. BOGAT converged to within 10% of the global minimum cost within 30 iterations in 93% of slices in our database. BOGAT achieved up to 56% lower flip angle RMS error (RMSE) or 55% lower pulse energy in phantoms versus the Fourier transform approach, and up to 30% lower RMSE and 29% lower energy in vivo with 7.8 s extra computation.

CONCLUSION

BOGAT efficiently estimated near-global optimum spoke positions for the two-spoke tests, reducing flip-angle RMSE and/or pulse energy in a computation time (˜10 s), which is suitable for online optimization.

TCMSSD: A comprehensive database focused on syndrome standardization

BACKGROUD

Quantitative and standardized research on syndrome differentiation has always been at the forefront of modernizing Traditional Chinese Medicine (TCM) theory. However, the majority of existing databases primarily concentrate on the network pharmacology of herbal prescriptions, and there are limited databases specifically dedicated to TCM syndrome differentiation.

PURPOSE

In response to this gap, we have developed the Traditional Chinese Medical Syndrome Standardization Database (TCMSSD, http://tcmssd.ratcm.cn).

METHODS

TCMSSD is a comprehensive database that gathers data from various sources, including TCM literature such as TCM Syndrome Studies (Zhong Yi Zheng Hou Xue) and TCM Internal Medicine (Zhong Yi Nei Ke Xue) and various public databases such as TCMID and ETCM. In our study, we employ a deep learning approach to construct the knowledge graph and utilize the BM25 algorithm for syndrome prediction.

RESULTS

The TCMSSD integrates the essence of TCM with the modern medical system, providing a comprehensive collection of information related to TCM. It includes 624 syndromes, 133,518 prescriptions, 8,073 diseases (including 1,843 TCM-specific diseases), 8,259 Chinese herbal medicines, 43,413 ingredients, 17,602 targets, and 8,182 drugs. By analyzing input data and comparing it with the patterns and characteristics recorded in the database, the syndrome prediction tool generates predictions based on established correlations and patterns.

CONCLUSION

The TCMSSD fills the gap in existing databases by providing a comprehensive resource for quantitative and standardized research on TCM syndrome differentiation and laid the foundation for research on the biological basis of syndromes.

AIEgen-deep: Deep learning of single AIEgen-imaging pattern for cancer cell discrimination and preclinical diagnosis

This study introduces AIEgen-Deep, an innovative classification program combining AIEgen fluorescent dyes, deep learning algorithms, and the Segment Anything Model (SAM) for accurate cancer cell identification. Our approach significantly reduces manual annotation efforts by 80%-90%. AIEgen-Deep demonstrates remarkable accuracy in recognizing cancer cell morphology, achieving a 75.9% accuracy rate across 26,693 images of eight different cell types. In binary classifications of healthy versus cancerous cells, it shows enhanced performance with an accuracy of 88.3% and a recall rate of 79.9%. The model effectively distinguishes between healthy cells (fibroblast and WBC) and various cancer cells (breast, bladder, and mesothelial), with accuracies of 89.0%, 88.6%, and 83.1%, respectively. Our method's broad applicability across different cancer types is anticipated to significantly contribute to early cancer detection and improve patient survival rates.

A highly integrated digital PCR system with on-chip heating for accurate DNA quantitative analysis

Digital polymerase chain reaction (dPCR) is extensively used for highly sensitive disease diagnosis due to its single-molecule detection ability. However, current dPCR systems require intricate DNA sample distribution, rely on cumbersome external heaters, and exhibit sluggish thermal cycling, hampering efficiency and speed of the dPCR process. Herein, we presented the development of a microwell array based dPCR system featuring an integrated self-heating dPCR chip. By utilizing hydrodynamic and electrothermal simulations, the chip's structure is optimized, resulting in improved partitioning within microwells and uniform thermal distribution. Through strategic hydrophilic/hydrophobic modifications on the chip's surface, we effectively secured the compartmentalization of sample within the microwells by employing an overlaying oil phase, which renders homogeneity and independence of samples in the microwells. To achieve precise, stable, uniform, and rapid self-heating of the chip, the ITO heating layer and the temperature control algorithm are deliberately designed. With a capacity of 22,500 microwells that can be easily expanded, the system successfully quantified EGFR plasmid solutions, exhibiting a dynamic linear range of 105 and a detection limit of 10 copies per reaction. To further validate its performance, we employed the dPCR platform for quantitative detection of BCR-ABL1 mutation gene fragments, where its performance was compared against the QuantStudio 3D, and the self-heating dPCR system demonstrated similar analytical accuracy to the commercial dPCR system. Notably, the individual chip is produced on a semiconductor manufacturing line, benefiting from mass production capabilities, so the chips are cost-effective and conducive to widespread adoption and accessibility.

Development and classification of autonomous vehicle's ambiguous driving scenario

Human drivers are gradually being replaced by highly automated driving systems, and this trend is expected to persist. The response of autonomous vehicles to Ambiguous Driving Scenarios (ADS) is crucial for legal and safety reasons. Our research focuses on establishing a robust framework for developing ADS in autonomous vehicles and classifying them based on AV user perceptions. To achieve this, we conducted extensive literature reviews, in-depth interviews with industry experts, a comprehensive questionnaire survey, and factor analysis. We created 28 diverse ambiguous driving scenarios and examined 548 AV users' perspectives on moral, ethical, legal, utility, and safety aspects. Based on the results, we grouped ADS, with all of them having the highest user perception of safety. We classified these scenarios where autonomous vehicles yield to others as moral, bottleneck scenarios as ethical, cross-over scenarios as legal, and scenarios where vehicles come to a halt as utility-related. Additionally, this study is expected to make a valuable contribution to the field of self-driving cars by presenting new perspectives on policy and algorithm development, aiming to improve the safety and convenience of autonomous driving.

Accurate actual flip angle imaging (AFI) in the presence of fat

PURPOSE

To investigate the effect of incomplete fat spoiling on the accuracy of B1 mapping with actual flip angle imaging (AFI) and to propose a method to minimize the errors using the chemical shift properties of fat.

THEORY AND METHODS

Diffusion-based dephasing is the main spoiling mechanism exploited in AFI. However, a very low diffusion in fat may make the spoiling insufficient, leading to ghosts in the B1 maps. As the errors retain the chemical-shift signature of fat, their impact can be minimized using chemical-shift-based fat signal removal from AFI acquisition modified to include multi-echo readout. The source of the errors and the proposed correction were studied in simulations and phantom and in-vivo imaging experiments.

RESULTS

Our results support that AFI artifacts are caused by the incomplete fat spoiling present in clinically attractive short TR acquisition regimes. The correction eliminated the ghosting and significantly improved the B1 mapping accuracy as well as the accuracy of R1 mapping performed with AFI-derived B1 maps.

CONCLUSIONS

The incomplete fat signal spoiling may be a source of AFI B1 mapping errors, especially in subjects with high fat content. Achieving complete fat spoiling requires longer TR, which is undesirable in clinical applications. The proposed approach based on fat signal removal can reduce errors without significant prolongation of the AFI pulse sequence. We propose that, when attaining complete fat spoiling is not feasible, AFI mapping should be performed in a multi-echo regime with appropriate fat separation or suppression to minimize these errors.

PSE-Net: Channel pruning for Convolutional Neural Networks with parallel-subnets estimator

Channel Pruning is one of the most widespread techniques used to compress deep neural networks while maintaining their performances. Currently, a typical pruning algorithm leverages neural architecture search to directly find networks with a configurable width, the key step of which is to identify representative subnet for various pruning ratios by training a supernet. However, current methods mainly follow a serial training strategy to optimize supernet, which is very time-consuming. In this work, we introduce PSE-Net, a novel parallel-subnets estimator for efficient channel pruning. Specifically, we propose a parallel-subnets training algorithm that simulate the forward-backward pass of multiple subnets by droping extraneous features on batch dimension, thus various subnets could be trained in one round. Our proposed algorithm facilitates the efficiency of supernet training and equips the network with the ability to interpolate the accuracy of unsampled subnets, enabling PSE-Net to effectively evaluate and rank the subnets. Over the trained supernet, we develop a prior-distributed-based sampling algorithm to boost the performance of classical evolutionary search. Such algorithm utilizes the prior information of supernet training phase to assist in the search of optimal subnets while tackling the challenge of discovering samples that satisfy resource constraints due to the long-tail distribution of network configuration. Extensive experiments demonstrate PSE-Net outperforms previous state-of-the-art channel pruning methods on the ImageNet dataset while retaining superior supernet training efficiency. For example, under 300M FLOPs constraint, our pruned MobileNetV2 achieves 75.2% Top-1 accuracy on ImageNet dataset, exceeding the original MobileNetV2 by 2.6 units while only cost 30%/16% times than BCNet/AutoAlim.

Robust T2 estimation with balanced steady state free precession

PURPOSE

To develop a novel signal representation for balanced steady state free precession (bSSFP) displaying its T2 independence on B1 and on magnetization transfer (MT) effects.

METHODS

A signal model for bSSFP is developed that shows only an explicit dependence (up to a scaling factor) on E2 (and, therefore, T2) and a novel parameter c (with implicit dependence on the flip angle and E1). Moreover, it is shown that MT effects, entering the bSSFP signal via a binary spin bath model, can be captured by a redefinition of T1 and, therefore, leading to modification of E1, resulting in the same signal model. Various sets of phase-cycled bSSFP brain scans (different flip angles, different TR, different RF pulse durations, and different number of phase cycles) were recorded at 3 T. The parameters T2 (E2) and c were estimated using a variable projection (VARPRO) method and Monte-Carlo simulations were performed to assess T2 estimation precision.

RESULTS

Initial experiments confirmed the expected independence of T2 on various protocol settings, such as TR, the flip angle, B1 field inhomogeneity, and the RF pulse duration. Any variation (within the explored range) appears to directly affect the estimation of the parameter c only-in agreement with theory.

CONCLUSION

BSSFP theory predicts an extraordinary feature that all MT and B1-related variational aspects do not enter T2 estimation, making it a potentially robust methodology for T2 quantification, pending validation against existing standards.

Joint MAPLE: Accelerated joint T1 and T 2 * $$ {{\mathrm{T}}_2}^{\ast } $$ mapping with scan-specific self-supervised networks

PURPOSE

Quantitative MRI finds important applications in clinical and research studies. However, it is encoding intensive and may suffer from prohibitively long scan times. Accelerated MR parameter mapping techniques have been developed to help address these challenges. Here, an accelerated joint T1,T 2 * $$ {{\mathrm{T}}_2}^{\ast } $$ , frequency and proton density mapping technique with scan-specific self-supervised network reconstruction is proposed to synergistically combine parallel imaging, model-based, and deep learning approaches to speed up parameter mapping.

METHODS

Proposed framework, Joint MAPLE, includes parallel imaging, signal modeling, and data consistency blocks which are optimized jointly in a combined loss function. A scan-specific self-supervised reconstruction is embedded into the framework, which takes advantage of multi-contrast data from a multi-echo, multi-flip angle, gradient echo acquisition.

RESULTS

In comparison with parallel reconstruction techniques powered by low-rank methods, emerging scan specific networks, and model-basedT 2 * $$ {{\mathrm{T}}_2}^{\ast } $$ estimation approaches, the proposed framework reduces the reconstruction error in parameter maps by approximately two-fold on average at acceleration rates as high as R = 16 with uniform sampling. It can outperform evaluated parallel reconstruction techniques up to four-fold on average in the presence of challenging sub-sampling masks. It is observed that Joint MAPLE performs well at extreme acceleration rates of R = 25 and R = 36 with error values less than 20%.

CONCLUSION

Joint MAPLE enables higher fidelity parameter estimation at high acceleration rates by synergistically combining parallel imaging and model-based parameter mapping and exploiting multi-echo, multi-flip angle datasets. Utilizing a scan-specific self-supervised reconstruction obviates the need for large data sets for training while improving the parameter estimation ability.

Temporal validation of a Generalized Additive2 Model (GA2M) to assess the risk of Chronic Kidney Disease (CKD)

OBJECTIVE

To assess the temporal validity of a model predicting the risk of Chronic Kidney Disease (CKD) using Generalized Additive2 Models (GA2M).

MATERIALS

We adopted the Italian Health Search Database (HSD) with which the original algorithm was developed and validated by comparing different machine learnings models.

METHODS

We selected all patients aged >=15 being active in HSD in 2019. They were followed up until December 2022 so being updated with three years of data collection. Those with prior diagnosis of CKD were excluded. A GA2M-based algorithm for CKD prediction was applied to this cohort in order to compare observed and predicted risk. Area Under Curve (AUC) and Average Precision (AP) were calculated.

RESULTS

We obtained an AUC and AP equal to 88% and 30%, respectively.

DISCUSSION

The prediction accuracy of the algorithm was largely consistent with that obtained in our prior work which was based on a different time-window for data collection. We therefore underlined and demonstrated the relevance of temporal validation for this prediction tool.

CONCLUSION

The GA2M confirmed its high accuracy in prediction of CKD. As such, the respective patient- and population-based informatic tools might be implemented in primary care.