HCHS-Net: A Multimodal Handcrafted Feature and Metadata Framework for Interpretable Skin Lesion Classification.

According to Skin Cancer Foundation, skin cancer is by far the most common type of cancer in the United States and worldwide. Early diagnosis of skin cancer is critical because proper treatment at early stages can increase the chance of cure and recovery. However, visual inspection of dermoscopic images by dermatologists is error-prone and time-consuming. To ensure accurate diagnosis and faster treatment of skin cancer, deep learning techniques have been utilized to conduct automated skin lesion segmentation and classification. In this paper, after image processing, a Mask R-CNN model is built for lesion segmentation, where transfer learning is utilized by using the pre-trained weights from Microsoft COCO dataset. The weights of the trained Mask R-CNN model are saved and transferred to the next task - skin lesion classification, to train a Mask R-CNN model for classification. Our experiments are conducted on the benchmark datasets from the International Skin Imaging Collaboration 2018 (ISIC 2018) and evaluated by the same metrics used in ISIC 2018. The lesion boundary segmentation and lesion classification have achieved an accuracy of 96% and a balanced multiclass accuracy of 80%, respectively.

In dermatology, accurate classification and segmentation of skin lesions remain important challenges due to variations in lesion appearance, low contrast in images, and overlapping features across disease types. Various traditional models pose some major difficulties in simultaneously capturing fine-grained local features and global environmental dependencies. Hence, this paper proposes a novel model that integrates a Residual Attention DeepLabV3 + Network for the segmentation, and a dual attention conformer for the classification, employing deep learning approaches to enhance both segmentation and classification performance of skin lesions. Publicly available skin disease image datasets are employed, with the pre-processing steps including Image rescaling and normalization, Histogram equalization, Gaussian blurring, edge detection, noise remover, and Contrast Limited Adaptive Histogram Equalization. To evaluate the performance of the model, the dataset is split using the holdout method, with a ratio of 80:10:10 for the training, testing, and validation phases. Based on the labeled ISIC dermatology image dataset, the simulation results show the excellent performance of the proposed model with 99.18%, 99.06%, and 98.92% accuracy, precision and recall. Overall, the proposed model achieves superior results in skin lesion classification and segmentation and offers promising potential for integration into medical diagnostic systems.

Skin lesion classification has recently attracted significant attention. Regularly, physicians take much time to analyze the skin lesions because of the high similarity between these skin lesions. An automated classification system using deep learning can assist physicians in detecting the skin lesion type and enhance the patient’s health. The skin lesion classification has become a hot research area with the evolution of deep learning architecture. In this study, we propose a novel method using a new segmentation approach and wide-ShuffleNet for skin lesion classification. First, we calculate the entropy-based weighting and first-order cumulative moment (EW-FCM) of the skin image. These values are used to separate the lesion from the background. Then, we input the segmentation result into a new deep learning structure wide-ShuffleNet and determine the skin lesion type. We evaluated the proposed method on two large datasets: HAM10000 and ISIC2019. Based on our numerical results, EW-FCM and wide-ShuffleNet achieve more accuracy than state-of-the-art approaches. Additionally, the proposed method is superior lightweight and suitable with a small system like a mobile healthcare system.

Multiclass skin lesion localization and classification using deep learning based features fusion and selection framework for smart healthcare

Diagnosis of brain tumor gliomas is a challenging task in medical image analysis due to its complexity, the less regularity of tumor structures, and the diversity of tissue textures and shapes. Semantic segmentation approaches using deep learning have consistently outperformed the previous methods in this challenging task. However, deep learning is insufficient to provide the required local features related to tissue texture changes due to tumor growth. This paper designs a hybrid method arising from this need, which incorporates machine-learned and hand-crafted features. A semantic segmentation network (SegNet) is used to generate the machine-learned features, while the grey-level co-occurrence matrix (GLCM)-based texture features construct the hand-crafted features. In addition, the proposed approach only takes the region of interest (ROI), which represents the extension of the complete tumor structure, as input, and suppresses the intensity of other irrelevant area. A decision tree (DT) is used to classify the pixels of ROI MRI images into different parts of tumors, i.e. edema, necrosis and enhanced tumor. The method was evaluated on BRATS 2017 dataset. The results demonstrate that the proposed model provides promising segmentation in brain tumor structure. The F-measures for automatic brain tumor segmentation against ground truth are 0.98, 0.75 and 0.69 for whole tumor, core and enhanced tumor, respectively.

Skin problems are common in day-to-day living. Skin lesions cause patients to be emotionally and psychologically burdened, which may be worse than the physical disorders. Skin lesions must be detected early to receive effective treatment. The diagnostic method must be precise and performed within a reasonable time. Many skin lesions look similar, which increases the potential of human error when classifying them. Deep learning's use in disease diagnosis has been a key focus of dermatological research. Skin lesion classification based on deep learning aids in the automatic skin lesion classification by dermoscopy, removing errors caused by naked eye inspection. The goal of this paper is to provide a quick survey on deep learning based skin lesions categorization, and outline the features of skin lesions. The essential stages of skin lesions and elements that affect dermatological diagnosis are reviewed, and the current difficulties and prospects of classification are discussed. The findings suggest that a deep learning-based skin lesion detection technique may beat expert dermatologists in some scenarios, and that there is a lot of room for further study.

Investigating the learning approaches of students in nursing education

BackgroundSkin cancer is one of the life threating diseases in the world. So, millions of lives can be saved by early detection of skin cancer. In addition, automating the computer‐aided system of skin lesion detection and classification (SLDC) will assist the medical practitioners to ensure more efficacious treatment of skin lesion disease.Material and MethodIn this article, a hybrid preprocessing‐based transfer learning model for SLDC is proposed, which is named as SLDCNet. Initially, the hybrid Gaussian filter (HGF) with connected component label (CCL) based fast march inpainting procedure is used for hair removal and denoising of skin lesions. Next, full resolution convolutional networks (FrCN) based segmentation method is adapted for detecting the cancer region. Then, feature extraction is performed using deep residual learning and finally, transfer learning mechanism is applied for classification of eight skin lesions.ResultsThe extensive simulation results shows that proposed SLDCNet resulted in a classification accuracy of 99.92%, sensitivity of 99%, and specificity of 99.36%, respectively.ConclusionFrom the obtained results, it is proven that proposed SLDCNet provides better performance as compared to state‐of‐art SLDC approaches, and even the standard ISIC‐2019 public challenge.

<span lang="EN-US">One type of skin cancer that is considered a malignant tumor is melanoma. Such a dangerous disease can cause a lot of death in the world. The early detection of skin lesions becomes an important task in the diagnosis of skin cancer. Recently, a machine learning paradigm emerged known as deep learning (DL) utilized for skin lesions classification. However, in some previous studies by using seven class images diagnostic of skin lesions classification based on a single DL approach with CNNs architecture does not produce a satisfying performance. The DL approach allows the development of a medical image analysis system for improving performance, such as the deep convolutional neural networks (DCNNs) method. In this study, we propose an ensemble learning approach that combines three DCNNs architectures such as Inception V3, Inception ResNet V2 and DenseNet 201 for improving the performance in terms of accuracy, sensitivity, specificity, precision, and F1-score. Seven classes of dermoscopy image categories of skin lesions are utilized with 10015 dermoscopy images from well-known the HAM10000 dataset. The proposed model produces good classification performance with 97.23% accuracy, 90.12% sensitivity, 97.73% specificity, 82.01% precision, and 85.01% F1-Score. This method gives promising results in classifying skin lesions for cancer diagnosis.</span>

Purpose:There has been burgeoning interest in applying machine learning methods for predicting radiotherapy outcomes. However, the imbalanced ratio of a large number of variables to a limited sample size in radiation oncology constitutes a major challenge. Therefore, dimensionality reduction methods can be a key to success. The study investigates and contrasts the application of traditional machine learning methods and deep learning approaches for outcome modeling in radiotherapy. In particular, new joint architectures based on variational autoencoder (VAE) for dimensionality reduction are presented and their application is demonstrated for the prediction of lung radiation pneumonitis (RP) from a large-scale heterogeneous dataset.Methods:A large-scale heterogeneous dataset containing a pool of 230 variables including clinical factors (e.g., dose, KPS, stage) and biomarkers (e.g., single nucleotide polymorphisms (SNPs), cytokines and micro-RNAs) in a population of 106 non-small cell lung cancer (NSCLC) patients who received radiotherapy was used for modeling RP. 22 patients had grade 2 or higher RP. Four methods were investigated, including feature selection (case A) and feature extraction (case B) with traditional machine learning methods, a VAE-MLP joint architecture (case C) with deep learning and lastly, the combination of feature selection and joint-architecture (case D). For feature selection, Random forest (RF), Support Vector Machine (SVM) and multi-layer perceptron (MLP) were implemented to select relevant features. Specifically, each method was run for multiple times to rank features within several cross-validated (CV) resampled sets. A collection of ranking lists were then aggregated by top 5% and Kemeny graph methods to identify the final ranking for prediction. A synthetic minority oversampling technique (SMOTE) was applied to correct for class imbalance during this process. For deep learning, a VAE-MLP joint architecture where a VAE aimed for dimensionality reduction and an MLP aimed for classification was developed. In this architecture, reconstruction loss and prediction loss were combined into a single loss function to realize simultaneous training and weights were assigned to different classes to mitigate class imbalance. To evaluate the prediction performance and conduct comparisons, the area under receiver operating characteristic (ROC) curves (AUCs) were performed for nested CVs for both handcrafted feature selections and the deep learning approach. The significance of differences in AUCs was assessed using the DeLong test of U-statistics.Results:An MLP-based method using weight pruning (WP) feature selection yielded the best performance among the different hand-crafted feature selection methods (case A), reaching an AUC of 0.804 (95% CI: 0.761–0.823) with 29 top features. A VAE-MLP joint architecture (case C) achieved a comparable but slightly lower AUC of 0.781 (95% CI: 0.737–0.808) with the size of latent dimension being 2. The combination of handcrafted features (case A) and latent representation (case D) achieved a significant AUC improvement of 0.831 (95% CI: 0.805–0.863) with 22 features (p-value = 0.000642 compared with handcrafted features only (Case A) and p-value=0.000453 compared to VAE alone (Case C)) with an MLP classifier.Conclusion:The potential for combination of traditional machine learning methods and deep learning VAE techniques has been demonstrated for dealing with limited datasets in modeling radiotherapy toxicities. Specifically, latent variables from a VAE-MLP joint architecture are able to complement handcrafted features for the prediction of RP and improve prediction over either method alone.

Recently biometric systems have shown improved capabilities because of the remarkable success of deep learning in solving various computer vision tasks. In ear recognition, the use of deep learning techniques is seldom due to training data scarcity. The existing work has shown poor performance as the majority of techniques are based on either handcraft features or pre-trained models. Besides this, transfer-learning has also shown poor performance because of the diversity among the tasks. To circumvent the existing issues, in this work, we have presented an end-to-end framework for ear recognition. It consist of the Ear Mask Extraction (EME) network to segment the ear, a normalization algorithm to align the ear, and a novel siamese-based CNN (CG-ERNet) for deep ear feature learning. CG-ERNet exploits domain-specific knowledge by using Curvature Gabor filters and uses triplet loss, triplet selection, and adaptive margin for better convergence of the loss. For comparative analysis, we trained state-of-the-art deep learning models like Face-Net, VGG19, ResNet50, Inception, Exception, and Mobile-Net for ear-recognition. The performance is assessed using five well-known evaluation metrics. In the extensive experimentation, our proposed model (CG-ERNet) outperformed the deep learning models and handcrafted feature based methods on four different, publicly available, benchmark datasets. To make the results more interpretable, we employ the t-SNE visualization of learned features. Additionally, our proposed method has shown robustness to various environmental challenges like Gaussian noise, Gaussian blur, up to ± 30 degrees of rotation, and 20% of occlusion.

Heuristics are one of the most important tools to guide search to solve combinatorial problems. They are often specifically designed for one single problem and require both expertise and implementation work. Generic frameworks like SAT or CSP have developed heuristics that obey general principles like first fail or are able to learn and adapt from the exploration of the search tree like Dom/wDeg. In SAT, the classic VSIDS heuristic falls into both categories. The question of whether it is possible to learn from solving existing problems has been addressed for a long time by portfolio solvers where the best heuristic is chosen by Machine Learning from hand-crafted features, and more recently with Deep Learning by embedding this knowledge into a Graph Neural Network (GNN). In this paper, we build upon the latter category by proposing a new heuristic based on Deep Reinforcement Learning using two GNNs with adversarial rewards. We show that our method reduces the number of fails to get the first solution by more than 50% compared to MiniSat. This work shows the advantages of this type of techniques to extract structural and contextual knowledge from past solving experience.

Deep convolutional neural networks for automatic classification of gastric carcinoma using whole slide images in digital histopathology

Accurate sleep stage scoring is crucial for diagnosing and treating sleep disorders, yet traditional manual methods are time-consuming and susceptible to variability. While recent advancements in machine learning and deep learning have enhanced automated sleep stage detection, many approaches still rely on handcrafted features and encounter limitations when processing full-night data. In this paper, we introduce a novel many-to-many classification framework that leverages a temporal focal modulation network for efficient and accurate sleep stage scoring. Our model, SleepFocalNet, processes full-night single-channel EEG signals and predicts sleep stages for all epochs simultaneously. SleepFocalNet is composed of three key components: a convolution block for local feature extraction, a focal modulation block for long-range temporal modeling, and a classification block for final predictions. We evaluated SleepFocalNet on Sleep Heart Health Study (SHHS), SleepEDF-20, and SleepEDF-78 datasets, achieving state-of-the-art performance. On SHHS, SleepFocalNet attained an accuracy of 0.888 and an F1-score of 0.815. On SleepEDF-20, it obtained an accuracy of 0.885 and an F1-score of 0.836. On SleepEDF-78, it outperformed other models with an accuracy of 0.855 and an F1-score of 0.800. This study represents the first application of temporal focal modulation networks in sleep stage scoring. Additionally, we conducted an extensive analysis of various network configurations to assess the impact of different architectural choices on performance. The results validate the potential of our approach to enhance the reliability and scalability of automated sleep stage scoring, offering a robust alternative to existing methods.

Accurate classification of mitotic figures in glioma histopathology images is critical for reliable tumor grading and informed treatment planning. In this study, we introduce HybridNet, a hybrid deep learning framework that combines the high-level semantic capabilities of a pre-trained Vision Transformer (ViT) with the domain-specific discriminative strength of handcrafted features. Specifically, four handcrafted feature sets, local binary patterns, texture descriptors, morphological attributes, and color statistics, are extracted to complement the ViT’s learned representations. These engineered features are concatenated with the partially fine-tuned ViT features to enhance classification performance and model interpretability. An ablation study was conducted to evaluate the contribution of each handcrafted feature group, revealing consistent performance gains with each module addition. In addition to HybridNet, we benchmark several conventional machine learning classifiers trained solely on handcrafted features, as well as standalone deep learning architectures including ResNet50, DenseNet121, EfficientNet-B0, and ViT, to establish robust performance baselines. The proposed method is evaluated on a curated dataset of glioma cell crops using five-fold cross-validation, achieving a mean classification accuracy of 91.75% with consistent performance across folds. The results underscore the complementary value of integrating handcrafted and deep features and highlight HybridNet’s potential as an interpretable and computationally efficient solution for mitotic figure classification in digital pathology workflows.