Two-dimensional Convolutional Neural Network Research Articles

Deep learning neural networks with input processing for vibration-based bearing fault diagnosis under imbalanced data conditions

Deep learning (DL) networks, such as convolutional neural networks (CNNs) and long short-term memory (LSTM), have gained popularity for bearing fault diagnosis utilizing raw vibration signals. However, their accuracy and stability are compromised when facing imbalanced real-world datasets. This research investigates the impact of imbalanced datasets and explores the potential of signal processing techniques on network inputs compared to the direct use of raw vibration signals. The DL techniques studied include LSTM, one-dimensional CNN, and two-dimensional (2D) CNN, and a novel hybrid 2DCNNLSTM algorithm, incorporating signal processing methods such as Fourier transform and continuous wavelet transform while maintaining nearly equal parameters and the same base architecture. The proposed hybrid 2DCNNLSTM algorithm combines the strengths of LSTM and CNN, allowing for improved bearing diagnosis by capturing both spatial and temporal information in vibration signals. The proposed 2DCNNLSTM algorithm also considers multi-channel input augmenting raw vibration signal, mean, and variance channels to extract meaningful features and enhance classification efficiency. The publicly available Case Western Reserve University benchmark-bearing test rig dataset with ten fault classes, the Paderborn University dataset with three fault classes, and NASA Centre for Intelligent Maintenance Systems bearing datasets with five fault classes are utilized to test the proposed deep learning networks’ accuracy, effectiveness, robustness, and stability. The studies reveal that the hybrid 2DCNNLSTM-based networks outperform both CNN and LSTM networks, even without input processing. Further, utilizing multi-channel input by augmenting the 2D raw signal with mean and variance value channels proves to be more efficient in handling imbalanced and complex datasets while employing a 2DCNNLSTM-based network.

A novel deep learning method based on 2-D CNNs and GRUs for permeability prediction of tight sandstone

The accurate calculation of tight sandstone reservoir permeability is crucial for optimizing production and maximizing natural gas recovery. Traditional physical model-based permeability calculation methods heavily rely on core experimental parameters. However, petrophysical experiments are costly, prompting the development of fast and reliable reservoir permeability prediction methods. In this paper, a novel deep learning model that combines two-dimensional convolutional neural network (2-D CNN) model and gated recurrent neural network is proposed to predict tight sandstone reservoir permeability. The 1-D conventional well logs data is converted into the 2-D geological feature map to fit the model's input dimension. The TimeDistributed layer wrapping technology is used to couple the 2-D CNN and GRU to extract comprehensive features of geological feature maps. By training the model, a depth nonlinear mapping between the spatial and temporal features in the well logs data and permeability is established. The model is divided into two subnetworks, 2-D CNNs and GRUs, which are used to compare and analyze the functions of different modules in the proposed model. Additionally, two traditional machine learning algorithms, SVM and XGBoost, are also introduced as comparison models. Four regression evaluation metrics are used to quantify the predictive performance of different models. The test results in a blind well demonstrate that the proposed method outperforms the comparison models in predicting permeability, with a coefficient of determination (R2) of 0.9305, a root mean square error (RMSE) of 0.0895, a mean squared error (MSE) of 0.0080, and a mean absolute error (MAE) of 0.0725. This model can provide insight into the petrophysical characteristics of reaction reservoirs based on conventional well logs data and can accurately estimate reservoir permeability under limited core data, which has great application potential in improving the exploration accuracy of unconventional oil and gas resources.

Forecasting Gas Well Classification Based on a Two-Dimensional Convolutional Neural Network Deep Learning Model

In response to the limitations of existing evaluation methods for gas well types in tight sandstone gas reservoirs, characterized by low indicator dimensions and a reliance on traditional methods with low prediction accuracy, therefore, a novel approach based on a two-dimensional convolutional neural network (2D-CNN) is proposed for predicting gas well types. First, gas well features are hierarchically selected using variance filtering, correlation coefficients, and the XGBoost algorithm. Then, gas well types are determined via spectral clustering, with each gas well labeled accordingly. Finally, the selected features are inputted, and classification labels are outputted into the 2D-CNN, where convolutional layers extract features of gas well indicators, and the pooling layer, which, trained by the backpropagation of CNN, performs secondary dimensionality reduction. A 2D-CNN gas well classification prediction model is constructed, and the softmax function is employed to determine well classifications. This methodology is applied to a specific tight gas reservoir. The study findings indicate the following: (1) Via two rounds of feature selection using the new algorithm, the number of gas well indicator dimensions is reduced from 29 to 15, thereby reducing the computational complexity of the model. (2) Gas wells are categorized into high, medium, and low types, addressing a deep learning multi-class prediction problem. (3) The new method achieves an accuracy of 0.99 and a loss value of 0.03, outperforming BP neural networks, XGBoost, LightGBM, long short-term memory networks (LSTMs), and one-dimensional convolutional neural networks (1D-CNNs). Overall, this innovative approach demonstrates superior efficacy in predicting gas well types, which is particularly valuable for tight sandstone gas reservoirs.

Prediction of hydrogen leakage location and intensity in hydrogen refueling stations based on deep learning

The common risk of high-pressure hydrogen leakage seriously affects the safety of hydrogen refueling stations (HRS). In this study, we have developed a hybrid model (WD-KNN-CNN) based on wavelet denoising (WD) and deep learning to predict the hydrogen leakage location and intensity in HRS. The raw data were calculated using validated Computational Fluid Dynamics (CFD), and the WD-processed raw data were utilized as input data for the hybrid model. The K-nearest Neighbors (KNN) is employed to identify the leak location, while a two-dimensional Convolutional Neural Network (CNN) is utilized to extract temporal and spatial distribution features from the concentration sequence for intensity prediction. Bayesian optimization is applied to determine the optimal hyperparameter combination for the model. The results demonstrate that the model achieves a prediction accuracy of 99.14% for leak location and 97.42% for intensity level. Furthermore, through the reduction of sensor quantity, we tested the model's performance and found that utilizing 14 sensors as input resulted in 91.3% accuracy for leak location and 90.8% for intensity level prediction. This hybrid model provides a method for safety monitoring in HRS and offers guidance for personnel response during leak incidents.

Multimodal Convolutional Neural Networks for Sperm Motility and Concentration Predictions

Semen analysis is an important analysis for male infertility primary investigation and manual semen analysis is a conventional method to assess it. Manual semen analysis has been revealed with accuracy and precision limitations due to noncompliance to guidelines and procedures. Sperm motility and concentration are the main indicators for pregnancy and conception rate hence they were selected for parameters prediction. Convolutional neural network (CNN) has benefited computer vision application industry in recent years and has been widely applied in computer vision research tasks. In this paper, three-dimensional CNN (3DCNN) was designed to extract motion and temporal features, which are vital for sperm motility prediction. For sperm concentration, since two-dimensional CNN (2DCNN) is efficient in recognizing and extracting spatial features, well-established Residual Network (ResNet) architecture was adopted and customized for sperm concentration prediction. Multimodal learning approach is a technique to aggregate learnt features from different deep learning architecture that adopted other forms of modalities, which could provide deep learning model with better insights on their tasks. Hence, a multimodal learning deep learning architecture was designed to receive both image-based (frames extracted from video samples) and video-based (stacked frames pre-processed from video samples) input that could provide well-extracted spatial and temporal features for sperm parameters prediction. The results obtained using the proposed methodology have surpassed other similar research works who used deep learning approach. For sperm motility, its best achieved average mean absolute error (MAE) was 8.048, and sperm concentration obtained a competent Pearson’s correlation coefficient (RP) value of 0.853.

GGNet: A novel graph structure for power forecasting in renewable power plants considering temporal lead-lag correlations

Power forecast for each renewable power plant (RPP) in the renewable energy clusters is essential. Though existing graph neural networks (GNN)-based models achieve satisfactory prediction performance by capturing dependencies among distinct RPPs, the static graph structure employed in these models ignores crucial lead-lag correlations among RPPs, resulting from the time difference of the air flow at spatially dispersed RPPs. To address this problem, this paper proposes a novel dynamic graph structure using multiple temporal granularity groups (TGGs) to characterize the lead-lag correlations among RPPs. A granular-based GNN called GGNet is designed to generate an optimal adjacency matrix for the proposed graph structure. Specifically, a two-dimensional convolutional neural network (2D-CNN) is used to quantify the uncertain lead-lag correlations among RPPs; secondly, a gate mechanism is used to calculate a dynamic adjacency matrix; Finally, a graph attention network (GAT) is used to aggregate the information on RPPs based on the well-learned adjacency matrix. Case studies conducted using real-world datasets, with wind power plants and photovoltaic power plants, show our method outperforms state-of-the-art (SoTA) ones with better performance. Compared with the SoTA models, the RMSEN and MAEN of wind power plants for 1–4 h forecast steps decreased on average by 22.925% and 13.18%, respectively; the RMSEN and MAEN of photovoltaic power plants for 1–4 h forecast steps decreased on average by 48.95% and 18.75%, respectively. The results show that the proposed framework can generate improved performance with accuracy and robustness.

Diagnostic performance of a deep-learning model using 18F-FDG PET/CT for evaluating recurrence after radiation therapy in patients with lung cancer.

We developed a deep learning model for distinguishing radiation therapy (RT)-related changes and tumour recurrence in patients with lung cancer who underwent RT, and evaluated its performance. We retrospectively recruited 308 patients with lung cancer with RT-related changes observed on 18F-fluorodeoxyglucose positron emission tomography-computed tomography (18F-FDG PET/CT) performed after RT. Patients were labelled as positive or negative for tumour recurrence through histologic diagnosis or clinical follow-up after 18F-FDG PET/CT. A two-dimensional (2D) slice-based convolutional neural network (CNN) model was created with a total of 3329 slices as input, and performance was evaluated with five independent test sets. For the five independent test sets, the area under the curve (AUC) of the receiver operating characteristic curve, sensitivity, and specificity were in the range of 0.98-0.99, 95-98%, and 87-95%, respectively. The region determined by the model was confirmed as an actual recurred tumour through the explainable artificial intelligence (AI) using gradient-weighted class activation mapping (Grad-CAM). The 2D slice-based CNN model using 18F-FDG PET imaging was able to distinguish well between RT-related changes and tumour recurrence in patients with lung cancer.

Improving Protein-Protein Interaction Prediction Using Protein Language Model and Protein Network Features

Interactions between proteins are ubiquitous in a wide variety of biological processes. Accurately identifying the protein-protein interactions (PPI) is of significant importance for understanding the mechanisms of protein functions and facilitating drug discovery. Although the wet-lab technological methods are the best way to identify PPI, their major constraints are their time-consuming nature, high cost, and labor-intensiveness. Hence, lots of efforts have been made towards developing computational methods to improve the performance of PPI prediction. In this study, we propose a novel hybrid computational method (called KSGPPI) that aims at improving the prediction performance of PPI via extracting the discriminative information from protein sequences and interaction networks. The KSGPPI model comprises two feature extraction modules. In the first feature extraction module, a large protein language model, ESM-2, is employed to exploit the global complex patterns concealed within protein sequences. Subsequently, feature representations are further extracted through CKSAAP, and a two-dimensional convolutional neural network (CNN) is utilized to capture local information. In the second feature extraction module, the query protein acquires its similar protein from the STRING database via the sequence alignment tool NW-align and then captures the graph embedding feature for the query protein in the protein interaction network of the similar protein using the algorithm of Node2vec. Finally, the features of these two feature extraction modules are efficiently fused; the fused features are then fed into the fully connected neural networks to predict PPI. The results of five-fold cross-validation on the used benchmarked datasets demonstrate that KSGPPI achieves an average prediction accuracy of 88.96%. Additionally, the average Matthews correlation coefficient value (0.781) of KSGPPI is significantly higher than that of those state-of-the-art PPI prediction methods. The standalone package of KSGPPI is freely downloaded at https://github.com/rickleezhe/KSGPPI.

Lithium-ion battery state of health prognostication employing multi-model fusion approach based on image coding of charging voltage and temperature data

The employment of readily obtainable parameters, including voltage and temperature, derived from the battery management system facilitates the real-time evaluation of lithium-ion batteries health status. However, the complex aging mechanisms and noise interference inherent in lithium-ion batteries present substantial challenges to accurate estimation. Hence, this research proposes a multi-model fusion approach utilizing Gramian Angular Field encoding to accurately predict the health status of lithium-ion batteries. Battery voltage and temperature data are acquired through system aging experiments. The Gramian Angular Field-based image encoding transforms these data into images, accentuating the subtle characteristics of the data and thereby enhancing its recognizability by neural networks. Subsequently, two-dimensional Convolutional Neural Networks are employed to extract features from the images, followed by one-dimensional Convolutional Neural Networks to reduce the dimensions of the feature matrix. The battery health status is then predicted utilizing Long Short-term Memory Networks. The culmination is the evaluation of the model performance through the application of quantitative error metrics. The study results show that the method can pinpoint the prediction error to within 2%, significantly improving accuracy over traditional prediction methods. It proves the great potential of direct data, such as voltage and temperature, in predicting battery health.

Unsupervised classification of multi-contrast magnetic resonance histology of peripheral arterial disease lesions using a convolutional variational autoencoder with a Gaussian mixture model in latent space: A technical feasibility study

PurposeTo investigate the feasibility of a deep learning algorithm combining variational autoencoder (VAE) and two-dimensional (2D) convolutional neural networks (CNN) for automatically quantifying hard tissue presence and morphology in multi-contrast magnetic resonance (MR) images of peripheral arterial disease (PAD) occlusive lesions. MethodsMulti-contrast MR images (T2-weighted and ultrashort echo time) were acquired from lesions harvested from six amputated legs with high isotropic spatial resolution (0.078 mm and 0.156 mm, respectively) at 9.4 T. A total of 4014 pseudo-color combined images were generated, with 75% used to train a VAE employing custom 2D CNN layers. A Gaussian mixture model (GMM) was employed to classify the latent space data into four tissue classes: I) concentric calcified (c), II) eccentric calcified (e), III) occluded with hard tissue (h) and IV) occluded with soft tissue (s). Test image probabilities, encoded by the trained VAE were used to evaluate model performance. ResultsGMM component classification probabilities ranged from 0.92 to 0.97 for class (c), 1.00 for class (e), 0.82–0.95 for class (h) and 0.56–0.93 for the remaining class (s). Due to the complexity of soft-tissue lesions reflected in the heterogeneity of the pseudo-color images, more GMM components (n=17) were attributed to class (s), compared to the other three (c, e and h) (n=6). ConclusionCombination of 2D CNN VAE and GMM achieves high classification probabilities for hard tissue-containing lesions. Automatic recognition of these classes may aid therapeutic decision-making and identifying uncrossable lesions prior to endovascular intervention.

Malignancy diagnosis of liver lesion in contrast enhanced ultrasound using an end-to-end method based on deep learning

BackgroundContrast-enhanced ultrasound (CEUS) is considered as an efficient tool for focal liver lesion characterization, given it allows real-time scanning and provides dynamic tissue perfusion information. An accurate diagnosis of liver lesions with CEUS requires a precise interpretation of CEUS images. However,it is a highly experience dependent task which requires amount of training and practice. To help improve the constrains, this study aims to develop an end-to-end method based on deep learning to make malignancy diagnosis of liver lesions using CEUS.MethodsA total of 420 focal liver lesions with 136 benign cases and 284 malignant cases were included. A deep learning model based on a two-dimensional convolution neural network, a long short-term memory (LSTM), and a linear classifier (with sigmoid) was developed to analyze the CEUS loops from different contrast imaging phases. For comparison, a 3D-CNN based method and a machine-learning (ML)-based time-intensity curve (TIC) method were also implemented for performance evaluation.ResultsResults of the 4-fold validation demonstrate that the mean AUC is 0.91, 0.88, and 0.78 for the proposed method, the 3D-CNN based method, and the ML-based TIC method, respectively.ConclusionsThe proposed CNN-LSTM method is promising in making malignancy diagnosis of liver lesions in CEUS without any additional manual features selection.

Spectrometer as a quantitative sensor for predicting the weld depth in laser welding

Spectrometers have been used for quality assessment or as auxiliary sensors in multisensor systems to record emissions at different wavelengths. This study developed a single-spectrometer sensor to quantitatively predict the laser penetration depth in Cu/Al overlap and dual-phase (DP) steel bead-on-plate welding joints. The spectrometer signals were sampled at 100 Hz, and an optical coherence tomography (OCT) sensor was used to estimate the in situ penetration depth. A one-dimensional (1D) convolution neural network (CNN) model was developed using the current step of the spectrometer signal. Furthermore, a two-dimensional (2D) CNN model was developed using 10 recent steps of the spectrometer signals. Consequently, 1732 and 3000 data points for DP steel and Al/Cu overlap welding were used for the penetration-depth modeling. Through model training, mean absolute errors of 22.67 and 52.96 µm were achieved for the Al/Cu overlap and DP steel welding, respectively, corresponding to approximately 1.5 % of the substrate thickness. The 2D-CNN models outperformed the 1D-CNN models for both material combinations. However, a slight delay was observed in the response of the models when confronted with abrupt output changes. This study confirmed that spectroscopy can serve as the sole quantitative sensing method for laser welding monitoring and control.

Multi-modal data cross-domain fusion network for gearbox fault diagnosis under variable operating conditions

Gearbox fault diagnosis is a critical aspect of machinery maintenance and reliability, as it ensures the safe and efficient operation of various industrial systems. The cross-domain fault diagnosis method based on transfer learning has been extensively researched to enhance the engineering applications of data-driven methods. Currently, the state-of-the-art gearbox cross-domain fault diagnosis primarily relies on single-modal data, which may not capture the full information needed for robust fault diagnosis under varying conditions. To address this issue, we propose a novel multi-modal data cross-domain fusion network that utilizes vibration signals and thermal images to capture comprehensive information about the gearbox’s health conditions. First, one-dimensional and two-dimensional convolutional neural networks are constructed for feature extraction and fusion of multi-modal data. Then, the Maximum Mean Discrepancy loss is introduced to achieve cross-domain feature alignments within the modal. Finally, the cross-modal consistency learning strategy is constructed to enhance the cross-domain diagnosis performance of the model. To validate the effectiveness of the proposed method, we conducted experiments on a real-world gearbox test rig. Experimental results demonstrate that the proposed method is superior to single-modal methods and existing fusion methods in terms of diagnosis performance, proving that the proposed method offers a promising solution for gearbox fault diagnosis in industrial settings.

Battery health prediction using two-dimensional multi-channel ensemble models

The accurate assessment of battery health is crucial for maximizing the performance and lifespan of lithium-ion batteries. In this paper, an ensemble model based on a two-dimensional multi-channel convolutional neural network is proposed to predict the maximum usable capacity of lithium-ion batteries. First, based on the charge–discharge process, the characteristic-derived lines of the capacity–voltage (Q–V) curve are extracted. Next, the Gramian angular summation field (GASF) is employed to convert one-dimensional time series data into image matrices, with the aim of preserving temporal characteristics. Finally, the paper discusses the impact of 2D matrix size on prediction results and introduces the use of a 2D multi-channel ensemble model for the unsupervised recognition and extraction of battery decline information from images, marking the first time this approach has been utilized. The experimental results demonstrate that the proposed method effectively captures battery degradation patterns from multiple perspectives. The average mean absolute percentage error of 1.95% and root mean square error of 1.41% on the verified batteries indicate the robustness and generalizability of the approach.

Source printer identification using convolutional neural network and transfer learning approach

<span>In recent years, Source printer identification has become increasingly important for detecting forged documents. A printer's distinguishing feature is its fingerprints. Each printer has a unique collection of fingerprints on every printed page. A model for identifying the source printer and classifying the questioned document into one of the printer classes is provided by source printer identification. A paper proposes a new approach that trains three different approaches on the dataset to choose the more accurate model for determining the printer's source. In the first, some pre-trained models are used as feature extractors, and support vector machine (SVM) is used to classify the generated features. In the second, we construct a two-dimensional convolutional neural network (2D-CNN) to address the source printer identification (SPI) problem. Instead of SoftMax, 2D-CNN is employed for feature extractors and SVM as a classifier. This approach obtains 93.75% 98.5% accuracy for 2D-CNN-SVM in the experiments. The SVM classifier enhanced the 2D-CNN accuracy by roughly 5% over the initial configuration. Finally, we adjusted 13 already-pre-trained CNN architectures using the dataset. Among the 13 pre-trained CNN models, DarkNet-19 has the greatest accuracy of 99.2 %. On the same dataset, the suggested approaches achieve well in terms of classification accuracy than the other recently released algorithms. </span>

Using Convolutional Neural Networks for Blocking Prediction in Elastic Optical Networks

This paper presents a study on connection-blocking prediction in Elastic Optical Networks (EONs) using Convolutional Neural Networks (CNNs). In EONs, connections are established and torn down dynamically to fulfill the instantaneous requirements of the users. The dynamic allocation of the connections may cause spectrum fragmentation and lead to network performance degradation as connection blocking increases. Predicting potential blocking situations can be helpful during EON operations. For example, this prediction could be used in real networks to trigger proper spectrum defragmentation mechanisms at suitable moments, thereby enhancing network performance. Extensive simulations over the well-known NSFNET (National Science Foundation Network) backbone network topology were run by generating realistic traffic patterns. The obtained results are later used to train the developed machine learning models, which allow the prediction of connection-blocking events. Resource use was continuously monitored and recorded during the process. Two different Convolutional Neural Network models, a 1D CNN (One-Dimensional Convolutional Neural Network) and 2D CNN (Two-Dimensional Convolutional Neural Network), are proposed as the predicting methods, and their behavior is compared to other conventional models based on an SVM (Support Vector Machine) and KNN (K Nearest Neighbors). The results obtained show that the proposed 2D CNN predicts blocking with the best accuracy (92.17%), followed by the SVM, the proposed 1D CNN, and KNN. Results suggest that 2D CNN can be helpful in blocking prediction and might contribute to increasing the efficiency of future EON networks.

Indirect identification of bridge damage based on coupled vehicle–bridge vibration and 2D-CNN

This study puts forth a methodology to discern structural damage in bridges that employs two-dimensional convolutional neural network (2D-CNN), which is rooted in the principles of continuous wavelet transform (CWT) theory. The method combines the vehicle–bridge coupled vibration response with deep learning models to extend the application of indirect bridge damage identification methods. To test the proposed method, a spatial vehicle and bridge computational model is established for a three-span continuous beam bridge, and bridge damage is simulated by reducing the stiffness of the unit under different damage conditions. Considering the stochastic nature of road roughness, a self-developed vehicle–bridge coupled vibration analysis program is utilized to acquire the vehicle acceleration response signal and construct the dataset. The 2D-CNN model, with its high sensitivity to 2D data features, is used to extract features from the vehicle vertical acceleration vibration signal. The signal undergoes transformation via CWT, resulting in a 2D grayscale time-frequency image. This image is subsequently utilised as input to construct the 2D-CNN model. Results demonstrate that this method performs well in the identification of bridge structural damage, exhibiting high accuracy in identifying the location and severity of such damage. Thus, a novel avenue is provided for the identification and assessment of bridge structural damage.

One-shot screening: Utilization of a two-dimensional convolutional neural network for automatic detection of left ventricular hypertrophy using electrocardiograms

Background and ObjectiveLeft ventricular hypertrophy (LVH) can impair ejection function and elevate the risk of heart failure. Therefore, early detection through screening is crucial. This study aimed to propose a novel method to enhance LVH detection using 12-lead electrocardiogram (ECG) waveforms with a two-dimensional (2D) convolutional neural network (CNN). MethodsUtilizing 42,127 pairs of ECG-transthoracic echocardiogram data, we pre-processed raw data into single-shot images derived from each ECG lead and conducted lead selection to optimize LVH diagnosis. Our proposed one-shot screening method, implemented during pre-processing, enables the superimposition of waveform source data of any length onto a single-frame image, thereby addressing the limitations of the one-dimensional (1D) approach. We developed a deep learning model with a 2D-CNN structure and machine learning models for LVH detection. To assess our method, we also compared our results with conventional ECG criteria and those of a prior study that used a 1D-CNN approach, utilizing the same dataset from the University of Tokyo Hospital for LVH diagnosis. ResultsFor LVH detection, the average area under the receiver operating characteristic curve (AUROC) was 0.916 for the 2D-CNN model, which was significantly higher than that obtained using logistic regression and random forest methods, as well as the two conventional ECG criteria (AUROC of 0.766, 0.790, 0.599, and 0.622, respectively). Incorporating additional metadata, such as ECG measurement data, further improved the average AUROC to 0.921. The model's performance remained stable across two different annotation criteria and demonstrated significant superiority over the performance of the 1D-CNN model used in a previous study (AUROC of 0.807). ConclusionsThis study introduces a robust and computationally efficient method that outperforms 1D-CNN models utilized in previous studies for LVH detection. Our method can transform waveforms of any length into fixed-size images and leverage the selected lead of the ECG, ensuring adaptability in environments with limited computational resources. The proposed method holds promise for integration into clinical practice as a tool for early diagnosis, potentially enhancing patient outcomes by facilitating earlier treatment and management.

Deep-VEGF: deep stacked ensemble model for prediction of vascular endothelial growth factor by concatenating gated recurrent unit with two-dimensional convolutional neural network

Vascular endothelial growth factor (VEGF) is involved in the development and progression of various diseases, including cancer, diabetic retinopathy, macular degeneration and arthritis. Understanding the role of VEGF in various disorders has led to the development of effective treatments, including anti-VEGF drugs, which have significantly improved therapeutic methods. Accurate VEGF identification is critical, yet experimental identification is expensive and time-consuming. This study presents Deep-VEGF, a novel computational model for VEGF prediction based on deep-stacked ensemble learning. We formulated two datasets using primary sequences. A novel feature descriptor named K-Space Tri Slicing-Bigram position-specific scoring metrix (KSTS-BPSSM) is constructed to extract numerical features from primary sequences. The model training is performed by deep learning techniques, including gated recurrent unit (GRU), generative adversarial network (GAN) and convolutional neural network (CNN). The GRU and CNN are ensembled using stacking learning approach. KSTS-BPSSM-based ensemble model secured the most accurate predictive outcomes, surpassing other competitive predictors across both training and testing datasets. This demonstrates the potential of leveraging deep learning for accurate VEGF prediction as a powerful tool to accelerate research, streamline drug discovery and uncover novel therapeutic targets. This insightful approach holds promise for expanding our knowledge of VEGF's role in health and disease. Communicated by Ramaswamy H. Sarma

NIR spectroscopy-CNN-enabled chemometrics for multianalyte monitoring in microbial fermentation.

As the biopharmaceutical industry looks to implement Industry 4.0, the need for rapid and robust analytical characterization of analytes has become a pressing priority. Spectroscopic tools, like near-infrared (NIR) spectroscopy, are finding increasing use for real-time quantitative analysis. Yet detection of multiple low-concentration analytes in microbial and mammalian cell cultures remains an ongoing challenge, requiring the selection of carefully calibrated, resilient chemometrics for each analyte. The convolutional neural network (CNN) is a puissant tool for processing complex data and making it a potential approach for automatic multivariate spectral processing. This work proposes an inception module-based two-dimensional (2D) CNN approach (I-CNN) for calibrating multiple analytes using NIR spectral data. The I-CNN model, coupled with orthogonal partial least squares (PLS) preprocessing, converts the NIR spectral data into a 2D data matrix, after which the critical features are extracted, leading to model development for multiple analytes. Escherichia coli fermentation broth was taken as a case study, where calibration models were developed for 23 analytes, including 20 amino acids, glucose, lactose, and acetate. The I-CNN model result statistics depicted an average R2 values of prediction 0.90, external validation data set 0.86 and significantly lower root mean square error of prediction values ∼0.52 compared to conventional regression models like PLS. Preprocessing steps were applied to I-CNN models to evaluate any augmentation in prediction performance. Finally, the model reliability was assessed via real-time process monitoring and comparison with offline analytics. The proposed I-CNN method is systematic and novel in extracting distinctive spectral features from a multianalyte bioprocess data set and could be adapted to other complex cell culture systems requiring rapid quantification using spectroscopy.