[A deep transfer learning approach for cross-subject recognition of mental tasks based on functional near-infrared spectroscopy].

In the field of brain-computer interfaces (BCIs) based on functional near-infrared spectroscopy (fNIRS), traditional subject-specific decoding methods suffer from the limitations of long calibration time and low cross-subject generalizability, which restricts the promotion and application of BCI systems in daily life and clinic. To address the above dilemma, this study proposes a novel deep transfer learning approach that combines the revised inception-residual network (rIRN) model and the model-based transfer learning (TL) strategy, referred to as TL-rIRN. This study performed cross-subject recognition experiments on mental arithmetic (MA) and mental singing (MS) tasks to validate the effectiveness and superiority of the TL-rIRN approach. The results show that the TL-rIRN significantly shortens the calibration time, reduces the training time of the target model and the consumption of computational resources, and dramatically enhances the cross-subject decoding performance compared to subject-specific decoding methods and other deep transfer learning methods. To sum up, this study provides a basis for the selection of cross-subject, cross-task, and real-time decoding algorithms for fNIRS-BCI systems, which has potential applications in constructing a convenient and universal BCI system.

[Early classification and recognition algorithm for sudden cardiac arrest based on limited electrocardiogram data trained with a two-stages convolutional neural network].

Sudden cardiac arrest (SCA) is a lethal cardiac arrhythmia that poses a serious threat to human life and health. However, clinical records of sudden cardiac death (SCD) electrocardiogram (ECG) data are extremely limited. This paper proposes an early prediction and classification algorithm for SCA based on deep transfer learning. With limited ECG data, it extracts heart rate variability features before the onset of SCA and utilizes a lightweight convolutional neural network model for pre-training and fine-tuning in two stages of deep transfer learning. This achieves early classification, recognition and prediction of high-risk ECG signals for SCA by neural network models. Based on 16 788 30-second heart rate feature segments from 20 SCA patients and 18 sinus rhythm patients in the international publicly available ECG database, the algorithm performance evaluation through ten-fold cross-validation shows that the average accuracy (Acc), sensitivity (Sen), and specificity (Spe) for predicting the onset of SCA in the 30 minutes prior to the event are 91.79%, 87.00%, and 96.63%, respectively. The average estimation accuracy for different patients reaches 96.58%. Compared to traditional machine learning algorithms reported in existing literatures, the method proposed in this paper helps address the requirement of large training datasets for deep learning models and enables early and accurate detection and identification of high-risk ECG signs before the onset of SCA.

Simultaneously predicting SPAD and water content in rice leaves using hyperspectral imaging with deep multi-task regression and transfer component analysis.

BACKGROUND: Water content and chlorophyll content are important indicators for monitoring rice growth status. Simultaneous detection of water content and chlorophyll content is of significance. Different varieties of rice show differences in phenotype, resulting in the difficulties of establishing a universal model. In this study, hyperspectral imaging was used to detect the Soil and Plant Analyzer Development (SPAD) values and water content of fresh rice leaves of three rice varieties (Jiahua 1, Xiushui 121 and Xiushui 134). RESULTS: Both partial least squares regression and convolutional neural networks were used to establish single-task and multi-task models. Transfer component analysis (TCA) was used as transfer learning to learn the common features to achieve an approximate identical distribution between any two varieties. Single-task and multi-task models were also built using the features of the source domain, and these models were applied to the target domain. These results indicated that for models of each rice variety the prediction accuracy of most multi-task models was close to that of single-task models. As for TCA, the results showed that the single-task model achieved good performance for all transfer learning tasks. CONCLUSION: Compared with the original model, good and differentiated results were obtained for the models using features learned by TCA for both the source domain and target domain. The multi-task models could be constructed to predict SPAD values and water content simultaneously and then transferred to another rice variety, which could improve the efficiency of model construction and realize rapid detection of rice growth indicators. © 2024 Society of Chemical Industry.

StochCA: A novel approach for exploiting pretrained models with cross-attention.

Utilizing large-scale pretrained models is a well-known strategy to enhance performance on various target tasks. It is typically achieved through fine-tuning pretrained models on target tasks. However, naï ve fine-tuning may not fully leverage knowledge embedded in pretrained models. In this study, we introduce a novel fine-tuning method, called stochastic cross-attention (StochCA), specific to Transformer architectures. This method modifies the Transformer's self-attention mechanism to selectively utilize knowledge from pretrained models during fine-tuning. Specifically, in each block, instead of self-attention, cross-attention is performed stochastically according to the predefined probability, where keys and values are extracted from the corresponding block of a pretrained model. By doing so, queries and channel-mixing multi-layer perceptron layers of a target model are fine-tuned to target tasks to learn how to effectively exploit rich representations of pretrained models. To verify the effectiveness of StochCA, extensive experiments are conducted on benchmarks in the areas of transfer learning and domain generalization, where the exploitation of pretrained models is critical. Our experimental results show the superiority of StochCA over state-of-the-art approaches in both areas. Furthermore, we demonstrate that StochCA is complementary to existing approaches, i.e., it can be combined with them to further improve performance. We release the code at https://github.com/daintlab/stochastic_cross_attention.

Partial Transfer Learning Method Based on Inter-Class Feature Transfer for Rolling Bearing Fault Diagnosis.

Rolling bearing fault diagnosis methods based on transfer learning always assume that the sample classes in the target domain are consistent with those in the source domain during the training phase. However, it is difficult to collect all fault classes in the early stage of mechanical application. The more likely situation is that the training data in the target domain only contain a subset of the entire health state, which will lead to the problem of label imbalance compared with the source domain. The outlier classes in the source domain that do not have corresponding target domain samples for feature alignment will interfere with the feature transfer of other classes. To address this specific challenge, this study introduces an innovative inter-class feature transfer fault diagnosis approach. By leveraging label information, the method distinctively computes the distribution discrepancies among shared classes, thereby circumventing the deleterious influence of outlier classes on the transfer procedure. Empirical evaluations on two rolling bearing datasets, encompassing multiple partial transfer tasks, substantiate that the proposed method surpasses other approaches, offering a novel and efficacious solution for the realm of intelligent bearing fault diagnosis.

Fault Diagnosis Method of Special Vehicle Bearing Based on Multi-Scale Feature Fusion and Transfer Adversarial Learning.

To address the issues of inadequate feature extraction for rolling bearings, inaccurate fault diagnosis, and overfitting in complex operating conditions, this paper proposes a rolling bearing diagnosis method based on multi-scale feature fusion and transfer adversarial learning. Firstly, a multi-scale convolutional fusion layer is designed to effectively extract fault features from the original vibration signals at multiple time scales. Through a feature encoding fusion module based on the multi-head attention mechanism, feature fusion extraction is performed, which can model long-distance contextual information and significantly improve diagnostic accuracy and anti-noise capability. Secondly, based on the domain adaptation (DA) cross-domain feature adversarial learning strategy of transfer learning methods, the extraction of optimal domain-invariant features is achieved by reducing the gap in data distribution between the target domain and the source domain, addressing the call for research on fault diagnosis across operating conditions, equipment, and virtual-real migrations. Finally, experiments were conducted to verify and optimize the effectiveness of the feature extraction and fusion network. A public bearing dataset was used as the source domain data, and special vehicle bearing data were selected as the target domain data for comparative experiments on the effect of network transfer learning. The experimental results demonstrate that the proposed method exhibits an exceptional performance in cross-domain and variable load environments. In multiple bearing cross-domain transfer learning tasks, the method achieves an average migration fault diagnosis accuracy rate of up to 98.65%. When compared with existing methods, the proposed method significantly enhances the ability of data feature extraction, thereby achieving a more robust diagnostic performance.

Dense Convolutional Neural Network-Based Deep Learning Pipeline for Pre-Identification of Circular Leaf Spot Disease of Diospyros kaki Leaves Using Optical Coherence Tomography.

Circular leaf spot (CLS) disease poses a significant threat to persimmon cultivation, leading to substantial harvest reductions. Existing visual and destructive inspection methods suffer from subjectivity, limited accuracy, and considerable time consumption. This study presents an automated pre-identification method of the disease through a deep learning (DL) based pipeline integrated with optical coherence tomography (OCT), thereby addressing the highlighted issues with the existing methods. The investigation yielded promising outcomes by employing transfer learning with pre-trained DL models, specifically DenseNet-121 and VGG-16. The DenseNet-121 model excels in differentiating among three stages of CLS disease (healthy (H), apparently healthy (or healthy-infected (HI)), and infected (I)). The model achieved precision values of 0.7823 for class-H, 0.9005 for class-HI, and 0.7027 for class-I, supported by recall values of 0.8953 for class-HI and 0.8387 for class-I. Moreover, the performance of CLS detection was enhanced by a supplemental quality inspection model utilizing VGG-16, which attained an accuracy of 98.99% in discriminating between low-detail and high-detail images. Moreover, this study employed a combination of LAMP and A-scan for the dataset labeling process, significantly enhancing the accuracy of the models. Overall, this study underscores the potential of DL techniques integrated with OCT to enhance disease identification processes in agricultural settings, particularly in persimmon cultivation, by offering efficient and objective pre-identification of CLS and enabling early intervention and management strategies.

Time-series domain adaptation via sparse associative structure alignment: Learning invariance and variance.

Domain adaptation on time-series data, which is often encountered in the field of industry, like anomaly detection and sensor data forecasting, but received limited attention in academia, is an important but challenging task in real-world scenarios. Most of the existing methods for time-series data use the covariate shift assumption for non-time-series data to extract the domain-invariant representation, but this assumption is hard to meet in practice due to the complex dependence among variables and a small change of the time lags may lead to a huge change of future values. To address this challenge, we leverage the stableness of causal structures among different domains. To further avoid the strong assumptions in causal discovery like linear non-Gaussian assumption, we relax it to mine the stable sparse associative structures instead of discovering the causal structures directly. Besides the domain-invariant structures, we also find that some domain-specific information like the strengths of the structures is important for prediction. Based on the aforementioned intuition, we extend the sparse associative structure alignment model in the conference version to the Sparse Associative Structure Alignment model with domain-specific information enhancement (SASA2 in short), which aligns the invariant unweighted spare associative structures and considers the variant information for time-series unsupervised domain adaptation. Specifically, we first generate the segment set to exclude the obstacle of offsets. Second, we extract the unweighted sparse associative structures via sparse attention mechanisms. Third, we extract the domain-specific information via an autoregressive module. Finally, we employ a unidirectional alignment restriction to guide the transformation from the source to the target. Moreover, we further provide a generalization analysis to show the theoretical superiority of our method. Compared with existing methods, our method yields state-of-the-art performance, with a 5% relative improvement in three real-world datasets, covering different applications: air quality, in-hospital healthcare, and anomaly detection. Furthermore, visualization results of sparse associative structures illustrate what knowledge can be transferred, boosting the transparency and interpretability of our method.

Deep convolutional neural network-based 3D fluorescence sensor array for sugar identification in serum based on the oxidase-mimicking property of CuO nanoparticles.

Developing sensor arrays capturing comprehensive fluorescence (FL) spectra from a single probe is crucial for understanding sugar structures with very high similarity in biofluids. Therefore, the analysis of highly similar sugar' structures in biofluids based on the entire FL of a single nanozyme probe needs more concern, which makes the development of novel alternative approaches highly wanted for biomedical and other applications. Herein, a well-designed deep learning model with intrinsic information of 3D FL of CuO nanoparticles (NPs)' oxidase-like activity was developed to classify and predict the concentration of a group of sugars with very similar chemical structures in different media. The findings presented that the overall accuracy of the developed model in classifying the nine selected sugars was (99-100 %), which prompted us to transfer the developed model to predict the concentration of the selected sugars at a concentration range of (1-100 µM). The transferred model also gave excellent results (R2 = 97-100 %). Therefore, the model was extended to other more complex applications, namely the identification of mixtures of sugars in serum and the detection of polysaccharides in different media such as serum and lake water. Notably, LOD for fructose was determined at 4.23 nM, marking a 120-fold decrease compared to previous studies. Our developed model was also compared with other deep learning-based models, and the results have demonstrated remarkable progress. Moreover, the identification of other possible coexisting interference substances in lake water samples was considered. This work marks a significant advancement, opening avenues for the widespread application of sensor arrays integrating nanozymes and deep learning techniques in biomedical and other diverse fields.

An explainable AI-based blood cell classification using optimized convolutional neural network.

White blood cells (WBCs) are a vital component of the immune system. The efficient and precise classification of WBCs is crucial for medical professionals to diagnose diseases accurately. This study presents an enhanced convolutional neural network (CNN) for detecting blood cells with the help of various image pre-processing techniques. Various image pre-processing techniques, such as padding, thresholding, erosion, dilation, and masking, are utilized to minimize noise and improve feature enhancement. Additionally, performance is further enhanced by experimenting with various architectural structures and hyperparameters to optimize the proposed model. A comparative evaluation is conducted to compare the performance of the proposed model with three transfer learning models, including Inception V3, MobileNetV2, and DenseNet201.The results indicate that the proposed model outperforms existing models, achieving a testing accuracy of 99.12%, precision of 99%, and F1-score of 99%. In addition, We utilized SHAP (Shapley Additive explanations) and LIME (Local Interpretable Model-agnostic Explanations) techniques in our study to improve the interpretability of the proposed model, providing valuable insights into how the model makes decisions. Furthermore, the proposed model has been further explained using the Grad-CAM and Grad-CAM++ techniques, which is a class-discriminative localization approach, to improve trust and transparency. Grad-CAM++ performed slightly better than Grad-CAM in identifying the predicted area's location. Finally, the most efficient model has been integrated into an end-to-end (E2E) system, accessible through both web and Android platforms for medical professionals to classify blood cell.

FLAT: Fusing layer representations for more efficient transfer learning in NLP.

Parameter efficient transfer learning (PETL) methods provide an efficient alternative for fine-tuning. However, typical PETL methods inject the same structures to all Pre-trained Language Model (PLM) layers and only use the final hidden states for downstream tasks, regardless of the knowledge diversity across PLM layers. Additionally, the backpropagation path of existing PETL methods still passes through the frozen PLM during training, which is computational and memory inefficient. In this paper, we propose FLAT, a generic PETL method that explicitly and individually combines knowledge across all PLM layers based on the tokens to perform a better transferring. FLAT considers the backbone PLM as a feature extractor and combines the features in a side-network, hence the backpropagation does not involve the PLM, which results in much less memory requirement than previous methods. The results on the GLUE benchmark show that FLAT outperforms other tuning techniques in the low-resource scenarios and achieves on-par performance in the high-resource scenarios with only 0.53% trainable parameters per task and 3.2× less GPU memory usagewith BERTbase. Besides, further ablation study is conducted to reveal that the proposed fusion layer effectively combines knowledge from PLM and helps the classifier to exploit the PLM knowledge to downstream tasks. We will release our code for better reproducibility.

Novel molecular inhibitor design for Plasmodium falciparum Lactate dehydrogenase enzyme using machine learning generated library of diverse compounds.

Generative machine learning models offer a novel strategy for chemogenomics and de novo drug design, allowing researchers to streamline their exploration of the chemical space and concentrate on specific regions of interest. In cases with limited inhibitor data available for the target of interest, de novo drug design plays a crucial role. In this study, we utilized a package called 'mollib,' trained on ChEMBL data containing approximately 365,000 bioactive molecules. By leveraging transfer learning techniques with this package, we generated a series of compounds, starting from five initial compounds, which are potential Plasmodium falciparum (Pf) Lactate dehydrogenase inhibitors. The resulting compounds exhibit structural diversity and hold promise as potential novel Pf Lactate dehydrogenase inhibitors.

Construction and Validation of a General Medical Image Dataset for Pretraining.

In the field of deep learning for medical image analysis, training models from scratch are often used and sometimes, transfer learning from pretrained parameters on ImageNet models is also adopted. However, there is no universally accepted medical image dataset specifically designed for pretraining models currently. The purpose of this study is to construct such a general dataset and validate its effectiveness on downstream medical imaging tasks, including classification and segmentation. In this work, we first build a medical image dataset by collecting several public medical image datasets (CPMID). And then, some pretrained models used for transfer learning are obtained based on CPMID. Various-complexity Resnet and the Vision Transformer network are used as the backbone architectures. In the tasks of classification and segmentation on three other datasets, we compared the experimental results of training from scratch, from the pretrained parameters on ImageNet, and from the pretrained parameters on CPMID. Accuracy, the area under the receiver operating characteristic curve, and class activation map are used as metrics for classification performance. Intersection over Union as the metric is for segmentation evaluation. Utilizing the pretrained parameters on the constructed dataset CPMID, we achieved the best classification accuracy, weighted accuracy, and ROC-AUC values on three validation datasets. Notably, the average classification accuracy outperformed ImageNet-based results by 4.30%, 8.86%, and 3.85% respectively. Furthermore, we achieved the optimal balanced outcome of performance and efficiency in both classification and segmentation tasks. The pretrained parameters on the proposed dataset CPMID are very effective for common tasks in medical image analysis such as classification and segmentation.

Harnessing Deep Learning for Accurate Pathological Assessment of Brain Tumor Cell Types.

Primary diffuse central nervous system large B-cell lymphoma (CNS-pDLBCL) and high-grade glioma (HGG) often present similarly, clinically and on imaging, making differentiation challenging. This similarity can complicate pathologists' diagnostic efforts, yet accurately distinguishing between these conditions is crucial for guiding treatment decisions. This study leverages a deep learning model to classify brain tumor pathology images, addressing the common issue of limited medical imaging data. Instead of training a convolutional neural network (CNN) from scratch, we employ a pre-trained network for extracting deep features, which are then used by a support vector machine (SVM) for classification. Our evaluation shows that the Resnet50 (TL + SVM) model achieves a 97.4% accuracy, based on tenfold cross-validation on the test set. These results highlight the synergy between deep learning and traditional diagnostics, potentially setting a new standard for accuracy and efficiency in the pathological diagnosis of brain tumors.

TF-EPI: an interpretable enhancer-promoter interaction detection method based on Transformer.

The detection of enhancer-promoter interactions (EPIs) is crucial for understanding gene expression regulation, disease mechanisms, and more. In this study, we developed TF-EPI, a deep learning model based on Transformer designed to detect these interactions solely from DNA sequences. The performance of TF-EPI surpassed that of other state-of-the-art methods on multiple benchmark datasets. Importantly, by utilizing the attention mechanism of the Transformer, we identified distinct cell type-specific motifs and sequences in enhancers and promoters, which were validated against databases such as JASPAR and UniBind, highlighting the potential of our method in discovering new biological insights. Moreover, our analysis of the transcription factors (TFs) corresponding to these motifs and short sequence pairs revealed the heterogeneity and commonality of gene regulatory mechanisms and demonstrated the ability to identify TFs relevant to the source information of the cell line. Finally, the introduction of transfer learning can mitigate the challenges posed by cell type-specific gene regulation, yielding enhanced accuracy in cross-cell line EPI detection. Overall, our work unveils important sequence information for the investigation of enhancer-promoter pairs based on the attention mechanism of the Transformer, providing an important milestone in the investigation of cis-regulatory grammar.

DenseNet201Plus: Cost-effective transfer-learning architecture for rapid leaf disease identification with attention mechanisms.

Plant leaf diseases are a significant concern in agriculture due to their detrimental impact on crop productivity and food security. Effective disease management depends on the early and accurate detection and diagnosis of these conditions, facilitating timely intervention and mitigation strategies. In this study, we address the pressing need for accurate and efficient methods for detecting leaf diseases by introducing a new architecture called DenseNet201Plus. DenseNet201 was modified by including superior data augmentation and pre-processing techniques, an attention-based transition mechanism, multiple attention modules, and dense blocks. These modifications enhance the robustness and accuracy of the proposed DenseNet201Plus model in diagnosing diseases related to plant leaves. The proposed architecture was trained using two distinct datasets: Banana Leaf Disease and Black Gram Leaf Disease. Through extensive experimentation, we evaluated the performance of DenseNet201Plus in terms of various classification metrics and achieved values of 0.9012, 0.9012, 0.9012, and 0.9716 for accuracy, precision, recall, and AUC for the banana leaf disease dataset, respectively. Similarly, the black gram leaf disease dataset model provides values of 0.9950, 0.9950, 0.9950, and 1.0 for accuracy, precision, recall, and AUC. Compared to other well-known pre-trained convolutional neural network (CNN) architectures, our proposed model demonstrates superior performance in both utilized datasets. Last but not least, we combined the strength of Grad-CAM++ with our proposed model to enhance the interpretability and localization of disease areas, providing valuable insights for agricultural practitioners and researchers to make informed decisions and optimize disease management strategies.

Evaluation and analysis of visual perception using attention-enhanced computation in multimedia affective computing.

Facial expression recognition (FER) plays a crucial role in affective computing, enhancing human-computer interaction by enabling machines to understand and respond to human emotions. Despite advancements in deep learning, current FER systems often struggle with challenges such as occlusions, head pose variations, and motion blur in natural environments. These challenges highlight the need for more robust FER solutions. To address these issues, we propose the Attention-Enhanced Multi-Layer Transformer (AEMT) model, which integrates a dual-branch Convolutional Neural Network (CNN), an Attentional Selective Fusion (ASF) module, and a Multi-Layer Transformer Encoder (MTE) with transfer learning. The dual-branch CNN captures detailed texture and color information by processing RGB and Local Binary Pattern (LBP) features separately. The ASF module selectively enhances relevant features by applying global and local attention mechanisms to the extracted features. The MTE captures long-range dependencies and models the complex relationships between features, collectively improving feature representation and classification accuracy. Our model was evaluated on the RAF-DB and AffectNet datasets. Experimental results demonstrate that the AEMT model achieved an accuracy of 81.45% on RAF-DB and 71.23% on AffectNet, significantly outperforming existing state-of-the-art methods. These results indicate that our model effectively addresses the challenges of FER in natural environments, providing a more robust and accurate solution. The AEMT model significantly advances the field of FER by improving the robustness and accuracy of emotion recognition in complex real-world scenarios. This work not only enhances the capabilities of affective computing systems but also opens new avenues for future research in improving model efficiency and expanding multimodal data integration.

Foundation models in gastrointestinal endoscopic AI: Impact of architecture, pre-training approach and data efficiency.

Pre-training deep learning models with large data sets of natural images, such as ImageNet, has become the standard for endoscopic image analysis. This approach is generally superior to training from scratch, due to the scarcity of high-quality medical imagery and labels. However, it is still unknown whether the learned features on natural imagery provide an optimal starting point for the downstream medical endoscopic imaging tasks. Intuitively, pre-training with imagery closer to the target domain could lead to better-suited feature representations. This study evaluates whether leveraging in-domain pre-training in gastrointestinal endoscopic image analysis has potential benefits compared to pre-training on natural images. To this end, we present a dataset comprising of 5,014,174 gastrointestinal endoscopic images from eight different medical centers (GastroNet-5M), and exploit self-supervised learning with SimCLRv2, MoCov2 and DINO to learn relevant features for in-domain downstream tasks. The learned features are compared to features learned on natural images derived with multiple methods, and variable amounts of data and/or labels (e.g. Billion-scale semi-weakly supervised learning and supervised learning on ImageNet-21k). The effects of the evaluation is performed on five downstream data sets, particularly designed for a variety of gastrointestinal tasks, for example, GIANA for angiodyplsia detection and Kvasir-SEG for polyp segmentation. The findings indicate that self-supervised domain-specific pre-training, specifically using the DINO framework, results into better performing models compared to any supervised pre-training on natural images. On the ResNet50 and Vision-Transformer-small architectures, utilizing self-supervised in-domain pre-training with DINO leads to an average performance boost of 1.63% and 4.62%, respectively, on the downstream datasets. This improvement is measured against the best performance achieved through pre-training on natural images within any of the evaluated frameworks. Moreover, the in-domain pre-trained models also exhibit increased robustness against distortion perturbations (noise, contrast, blur, etc.), where the in-domain pre-trained ResNet50 and Vision-Transformer-small with DINO achieved on average 1.28% and 3.55% higher on the performance metrics, compared to the best performance found for pre-trained models on natural images. Overall, this study highlights the importance of in-domain pre-training for improving the generic nature, scalability and performance of deep learning for medical image analysis. The GastroNet-5M pre-trained weights are made publicly available in our repository: huggingface.co/tgwboers/GastroNet-5M_Pretrained_Weights.

Development of a CNN for Adult Brain Tumour Characterisation: Implications and Future Directions for Transfer Learning.

Brain tumours are the most commonly occurring solid tumours in children, albeit with lower incidence rates compared to adults. However, their inherent heterogeneity, ethical considerations regarding paediatric patients, and difficulty in long-term follow-up make it challenging to gather large homogenous datasets for analysis. This study focuses on the development of a Convolutional Neural Network (CNN) for brain tumour characterisation using the adult BraTS 2020 dataset. We propose to transfer knowledge, from models pre-trained on extensive adult brain tumour datasets to smaller cohort datasets (e.g., paediatric brain tumours) in future studies, by leveraging Transfer Learning (TL). This approach aims to extract relevant features from pre-trained models, addressing the limited availability of annotated paediatric datasets and enhancing tumour characterisation in children. The implications and potential applications of this methodology in paediatric neuro-oncology are discussed.

Novel Deep CNNs Explore Regions, Boundaries, and Residual Learning for COVID-19 Infection Analysis in Lung CT.

COVID-19 poses a global health crisis, necessitating precise diagnostic methods for timely containment. However, accurately delineating COVID-19-affected regions in lung CT scans is challenging due to contrast variations and significant texture diversity. In this regard, this study introduces a novel two-stage classification and segmentation CNN approach for COVID-19 lung radiological pattern analysis. A novel Residual-BRNet is developed to integrate boundary and regional operations with residual learning, capturing key COVID-19 radiological homogeneous regions, texture variations, and structural contrast patterns in the classification stage. Subsequently, infectious CT images undergo lesion segmentation using the newly proposed RESeg segmentation CNN in the second stage. The RESeg leverages both average and max-pooling implementations to simultaneously learn region homogeneity and boundary-related patterns. Furthermore, novel pixel attention (PA) blocks are integrated into RESeg to effectively address mildly COVID-19-infected regions. The evaluation of the proposed Residual-BRNet CNN in the classification stage demonstrates promising performance metrics, achieving an accuracy of 97.97%, F1-score of 98.01%, sensitivity of 98.42%, and MCC of 96.81%. Meanwhile, PA-RESeg in the segmentation phase achieves an optimal segmentation performance with an IoU score of 98.43% and a dice similarity score of 95.96% of the lesion region. The framework's effectiveness in detecting and segmenting COVID-19 lesions highlights its potential for clinical applications.