Irbm最新文献

1) Introduction: Preterm labor (PL) has globally become the leading cause of death in children under the age of 5 years. One of the most significant keys to preventing preterm labor is its early detection. 2) Objectives: The primary objectives of this study are to address the problem of PL by providing a new approach by analyzing the electrohysterographic (EHG) signals, which are recorded on the mother's abdomen during labor and pregnancy. 3) Methods: The EHG signal reflects the electrical activity that induces the mechanical contraction of the myometrium. Because EHGs are known to be non-stationary signals, and because we anticipate connectivity to alter during contraction (due to electrical diffusion and the mechanotransduction process), we applied the windowing approach on real signals to identify the best windows and the best nodes with the most significant data to be used for classification. The suggested pipeline includes: i) dividing the 16 EHG signals that are recorded from the abdomen of pregnant women in N windows; ii) apply the connectivity matrices on each window; iii) apply the Graph theory-based measures on the connectivity matrices on each window; iv) apply the consensus Matrix on each window in order to retrieve the best windows and the best nodes. Following that, several neural network and machine learning methods are applied to the best windows and best nodes to categorize pregnancy and labor contractions, based on the different input parameters (connectivity method alone, connectivity method plus graph parameters, best nodes, all nodes, best windows, all windows). 4) Results: Results showed that the best nodes are nodes 8, 9, 10, 11, and 12; while the best windows are 2, 4, and 5. The classification results obtained by using only these best nodes are better than when using the whole nodes. The results are always better when using the full burst, whatever the chosen nodes. 5) Conclusion: The windowing approach proved to be an innovative technique that can improve the differentiation between labor and pregnancy EHG signals.

Background

The global prevalence of autism spectrum disorder (ASD) is around 1%. Yet the current diagnosis of ASD mainly depends on clinicians' experience and caregivers' report, which are subjective, time consuming, and labor demanding. An objective and efficient way to diagnose ASD is urgently needed. The objective of this study was to quantify an omnipresent yet least studied behavioral characteristic in children with ASD – interpersonal motor coordination (IMC), and to investigate the feasibility of using IMC related features to identify ASD by implementing machine learning algorithms.

Methods

Twenty children with ASD and twenty-three children with typical development (TD) were filmed in a conversation with an interviewer. Motion energy analysis was implemented to obtain the movement time series, and cross wavelet analysis (CWA) quantified the level of IMC at different movement frequencies. Machine learning algorithms were utilized to examine whether these two groups of children could be accurately classified using features of IMC.

Results

Statistical analysis revealed reduced IMC in the ASD group at relatively high movement frequencies. The establishment of machine learning (ML) models showed that the maximum classification accuracy was 85.37% (specificity = 95.24%, sensitivity = 75.00%) using five original coherence values computed with CWA. In addition, the classification accuracy could be improved to 92.68% (specificity = 95.24%, sensitivity = 90.00%) with three novel features created by taking the sum of statistically significant features.

Conclusions

Children with ASD demonstrated an atypical profile of IMC, and IMC could be used to objectively classify children with ASD and TD. In addition, our analyses showed that creating novel features based on statistically significant features could help improve classification performance. It is proposed that such economic, contactless, and calibration-free approach to data collection might well serve both ASD research and practice, particularly early objective identification. However, this study could be improved with respect to larger sample size with balanced gender ratio and different severity.

Objective: Alzheimer's disease (AD) is an irreversible neurodegenerative disease, while mild cognitive impairment (MCI) is a clinical precursor of AD, thus differentiation of AD, MCI and normal control (NC) from noninvasive magnetic resonance imaging (MRI) has positive clinical implications. Material and method: We utilize a 3D residual network to classify AD, MCI, and NC, and add a multiscale module to the original network to enhance the feature representation capability of the network, as well as a cross-dimensional attentional mechanism to enhance the network's attention to important brain regions. We experimentally verified that the network is more inclined to overestimate the brain age of patients in AD and MCI subgroups, thus proving that there is a high correlation between the brain age prediction task and the AD classification task. Therefore, we adopted a multi-task learning approach, using brain age prediction as a supplementary task for AD classification to reduce the risk of overfitting of the network during the training process. Results: Our method achieved 96.02% accuracy, 93.40% precision, 91.48% recall, and 92.24% F1 value in AD/MCI/NC classification. Conclusions: Ablation experiments confirmed that our proposed cross-dimensional attention and multiscale modules can improve the diagnostic performance of AD and MCI, and that multi-task learning in conjunction with brain age prediction can further improve the performance.

Objective

Unstable plaques can cause acute cardiovascular and cerebrovascular diseases. The stability and instability of plaque are related to the plaque echo status in ultrasound. Carotid videos provide detailed plaque information compared to static images. Ultrasound-based plaque echo classification is challenging due to noise, interference frames, small targets (plaques), and complex shape changes.

Methods

This study proposes a Multi-dimensional Aggregation Network (MA-Net) guided by key features for plaque diagnosis based on carotid ultrasound video, which uses only video-level labels. MA-Net consists of Key-Feature (KF) and Temporal-Channel-Spatial (TCS) modules. The KF module learns the contribution of each frame to the classification at the feature level, adaptively infers the importance score of each frame, thereby reducing the influence of interference frames. The TCS module includes the Temporal-Channel (TC) and Temporal-Spatial (TS) sub-modules. In addition to studying the temporal dimension, it delves into the relationship between the channel and spatial dimensions. TC analyses the temporal dependencies among the channels and filters noise. Moreover, TS extracts features more accurately through the spatio-temporal information contained in the surrounding environment of the plaque.

Results

The performance of MA-Net on the SHU-Ultrasound-Video-2020 dataset is better than that of the state-of-the-art models of video classification, showing at least a 5% increase in accuracy, with an accuracy rate of 87.36%.

Conclusion

The outstanding diagnostic capability of the proposed model will help provide a more robust and reproducible diagnostic process with a lower labour cost for clinical carotid plaque diagnosis.

Objective

This study addresses the challenge of user-specific bias in Brain-Computer Interfaces (BCIs) by proposing a novel methodology. The primary objective is to employ a hybrid deep learning model, combining 2D Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) layers, to analyze EEG signals and classify imagined tasks. The overarching goal is to create a generalized model that is applicable to a broader population and mitigates user-specific biases.

Materials and Methods

EEG signals from imagined motor tasks in the public dataset Physionet form the basis of the study. This is due to the need to use other databases in addition to the BCI competition. A model of arrays emulating the electrode arrangement in the head is proposed to capture spatial information using CNN, and LSTM algorithms are used to capture temporal information, followed by signal classification.

Results

The hybrid model is implemented to achieve a high classification rate, reaching up to 90% for specific users and averaging 74.54%. Error detection thresholds are set to eliminate subjects with low task affinity, resulting in a significant improvement in classification accuracy of up to 21.34%.

Conclusion

The proposed methodology makes a significant contribution to the BCI field by providing a generalized system trained on diverse user data that effectively captures spatial and temporal EEG signal features. This study emphasizes the value of the hybrid model in advancing BCIs, highlighting its potential for improved reliability and accuracy in human-computer interaction. It also suggests the exploration of additional advanced layers, such as transformers, to further enhance the proposed methodology.

Objectives

This paper is proposed to identify both promising technologies and potential products in the domain of biosensor using patent-based Technology-Driven Technology Roadmaps (TRM).

Materials and methods

The technology-driven TRM with timelines in this study is developed in three layers: technology, function and product. Patent applications are collected and identified to interpret technologies and functions for biosensors in healthcare, and product manuals or releases can be used as product introductions.

Results

Most biosensors in healthcare patents are concentrated in biochemical (T2) and electroencephalography (T5). Glycated hemoglobin (F1), measuring of glucose (F3), and biological process and molecular systems (F6) have a relatively larger patent count. Biochemical (T2) can combine with biological process and molecular systems (F6), and then brain's real-time electrical activity monitoring can be handled. Biochemical (T2) can also devote to glycated hemoglobin (F1), and glucose monitoring (F3), and thus create QCM sensor, CGM and GlucoWatch etc. applications.

Conclusion

Biochemical (T2) has a wide application among different functions for wearable biosensors in healthcare. This paper identifies and explores new developments biochemical (T2), and electroencephalography (T5) in wearable biosensors are expected to play a significant role over the coming decade in improving the current healthcare infrastructure, and enhancing the democratization of information and allocation of medical resources.

Objectives

This study was performed to prove the concept that transfer learning techniques, assisted with a generative model, could be used to alleviate the ‘big data’ requirement for training high-fidelity deep learning (DL) models in prediction of stiffness tensor of trabecular bone cubes.

Material and methods

Transfer learning approaches of domain adaptation were used, in which a source domain included 1,641 digital trabecular bone cubes synthesized from a generative model, and a target domain included 868 real trabecular bone cubes from human cadaver femurs. Simulated quantitative computed tomography (QCT) images of both the synthesized and real bone cubes were used as input, whereas the stiffness tensor of these cubes determined using finite element simulations were used as output. Three transfer learning algorithms, including instance-based (TrAdaBoostR2 and WANN) and parameter-based (RNN) methods, were used. Two case studies, one with varying sizes of training dataset and the other with a gender-biased training dataset, were performed to evaluate these deep transfer learning models in comparison with a base deep learning (DL) model trained using the dataset from the target domain.

Results

The results indicated that these deep transfer learning models were robust both to sample size and to the gender-biased training dataset, whereas the base DL model was very sensitive to such changes. Among the three transfer learning algorithms, the prediction accuracy of the RNN-based deep transfer learning model was the best (0.92-0.96%) and comparable to that of the base DL model trained using the dataset from the target domain.

Conclusion

This study proved the proposed concept and confirmed that high fidelity QCT-based deep learning models could be obtained for prediction of stiffness tensor of trabecular bone cubes.

Background and objective

Steady-state visual evoked potential (SSVEP)-based brain-computer interfaces (BCIs) aim to detect target frequencies corresponding to specific commands in electroencephalographic (EEG) signals by classification algorithms to achieve the desired control. However, SSVEP signals suffer from low signal-to-noise ratio and large differences in brain activity. Moreover, the existing CNN models have small receptive fields, which make it difficult to receive large range of feature information and limit the effectiveness of classification algorithms.

Methods

To this end, we proposed a high-order temporal convolutional neural network (HOT-CNN) model for enhancing the performance of SSVEP target recognition. Specifically, the SSVEP-EEG signals was divided into equal-length time segments and a time-slice attention module was designed to capture the correlation between time slices. The module improves the local characterization of signals and reduces biological noise interference by automatically assigning high weights to locally relevant temporal sampling cues and lower weights to other temporal cues. Moreover, for global features, a temporal convolutional network module was designed to increases the receptive field of the network and to extract more comprehensive time domain features by using dilated causal convolution. Finally, the fusion and analysis of local and global features are achieved by designing a feature fusion and classification module to accomplish accurate classification of SSVEP signals.

Results

Our method was evaluated on large publicly available datasets containing 35 subjects and 40 categories. Experimental results indicated that HOT-CNN achieved encouraging performance compared with other advanced methods: the highest information transfer rate of 241.01bits/min was obtained using 0.5s stimuli, and the highest average accuracy of 96.39% was obtained using 1.0s stimuli.

Conclusions

The method effectively reinforced the global and local time-domain information and improved the classification performance of SSVEP, which has wide application prospects.