Hybrid deep learning architecture for skin disease classification

Skin diseases affect 30-70% of the world population, and globally, skin cancer rates are continuously increasing. In this respect, prevention programs and early detection of skin diseases are of particular importance. To screen sewer workers for skin diseases with regard to their work-related risk. Employees of the municipal utilities in Munich (Münchner Stadtentwässerung) underwent awhole-body examination of the skin, conducted by two dermatologists. In addition, all employees completed apaper-based questionnaire on risk behavior and preventive measures. We examined 81employees (79men, 2women, mean age 45.7 ± 9.5years). Skin lesions in need of treatment were found in 30.9% (n = 25): the most frequent diagnosis was mycosis pedis (16.1%). In addition, one employee was diagnosed with basal cell carcinoma and two with actinic keratoses. According to the questionnaire, 43.5% of the employees had undergone aphysician-led skin cancer screening in the past, whereas sun-protection practices were rarely applied. According to our findings, employee skin cancer screening seems to be beneficial for the detection of work-related skin diseases and is associated with a high participation rate. Furthermore, the study suggests that sewer workers have ahigh rate of mycosis pedis, possibly awork-related effect.

BackgroundThere is a high prevalence of skin diseases sub-Saharan Africa, including skin neglected tropical diseases (NTDs) that could lead to lifelong disabilities and deformities if not diagnosed and treated early. To achieve early detection and early treatment of these skin diseases, we developed a mobile health app, eSkinHealth.ObjectiveThis paper outlines a protocol for evaluating the effect of our eSkinHealth app in the early detection and effective management of skin diseases in Côte d’Ivoire.MethodsA mixed methods pilot trial will be conducted in Côte d’Ivoire and will consist of 3 phases: (1) the development and improvement of the eSkinHealth app, (2) a pilot trial to evaluate the usability of the eSkinHealth app for local medical staff in Côte d’Ivoire, and (3) a pilot trial to evaluate the effectiveness of early detection and case management of targeted skin NTDs (Buruli ulcer, leprosy, yaws, and lymphatic filariasis) with the eSkinHealth app in Côte d’Ivoire. The pilot study will be implemented as a 2-arm trial with local health care providers and patients with skin NTDs over a 3-month follow-up period. The local health care providers will be assigned to an intervention group receiving the eSkinHealth app to be used in their daily practices or a control group. Training will be provided on the use and implementation of the app and the diagnostic pipeline to the intervention group only, while both groups will receive training on skin diseases. Our primary outcome is to evaluate the early detection and effective management of skin diseases using the eSkinHealth app in Côte d’Ivoire by the number of cases diagnosed and managed. Additionally, we will evaluate the eSkinHealth app with validated questionnaires and in-depth interviews. Procedures of our methods have been reviewed and approved by the Institutional Review Board of the Ministry of Health, Côte d’Ivoire and by Tulane University in 2021.ResultsThis study was funded in 2021. We started the enrollment of patients in February 2022, and data collection is currently underway. We expect the first results to be submitted for publication in 2023.ConclusionsOur eSkinHealth app is a field-adapted platform that could provide both direct diagnostic and management assistance to health workers in remote settings. The study will provide evidence for the usability and the effectiveness of the eSkinHealth app to improve the early detection and case management of skin NTDs in Côte d’Ivoire and, furthermore, is expected to contribute to knowledge on mobile health approaches in the control of skin NTDs.Trial RegistrationClinicalTrials.gov NCT05300399; https://clinicaltrials.gov/ct2/show/study/NCT05300399International Registered Report Identifier (IRRID)DERR1-10.2196/39867

ObjectivesSkin diseases account for 1.79% of the global disease burden, though their impact may be underreported due to limited research. Raising awareness about early detection is essential, but tools to measure this are scarce. This study aimed to develop and validate a questionnaire assessing community awareness and early detection of common skin diseases, including leprosy, in Indonesia.MethodsThe questionnaire was drafted in English, translated into Bahasa Indonesia, and tested for face validity, content validity, and internal consistency. A pilot test with 25 participants and a field test with 680 participants were conducted. Results were analyzed using Cronbach alpha and descriptive methods.ResultsThe final questionnaire comprised 17 questions on skin disease: common knowledge, intentions on prevention and health-seeking behavior. Validity and internal consistency were confirmed during pilot testing, and no participants in the field test reported confusion. The Cronbach alpha score exceeded 0.70, confirming strong internal consistency.ConclusionThis validated questionnaire can assess public awareness of early skin disease detection. It is available for international adaptation and may help improve early detection and prevention in Indonesia’s healthcare system.

Rationale: Care of emergency department (ED) patients with pneumonia can be challenging. Clinical decision support may decrease unnecessary variation and improve care. Objectives: To report patient outcomes and processes of care after deployment of electronic pneumonia clinical decision support (ePNa): a comprehensive, open loop, real-time clinical decision support embedded within the electronic health record. Methods: We conducted a pragmatic, stepped-wedge, cluster-controlled trial with deployment at 2-month intervals in 16 community hospitals. ePNa extracts real-time and historical data to guide diagnosis, risk stratification, microbiological studies, site of care, and antibiotic therapy. We included all adult ED patients with pneumonia over the course of 3 years identified by International Classification of Diseases, 10th Revision discharge coding confirmed by chest imaging. Measurements and Main Results: The median age of the 6,848 patients was 67 years (interquartile range, 50-79), and 48% were female; 64.8% were hospital admitted. Unadjusted mortality was 8.6% before and 4.8% after deployment. A mixed effects logistic regression model adjusting for severity of illness with hospital cluster as the random effect showed an adjusted odds ratio of 0.62 (0.49-0.79; P < 0.001) for 30-day all-cause mortality after deployment. Lower mortality was consistent across hospital clusters. ePNa-concordant antibiotic prescribing increased from 83.5% to 90.2% (P < 0.001). The mean time from ED admission to first antibiotic was 159.4 (156.9-161.9) minutes at baseline and 150.9 (144.1-157.8) minutes after deployment (P < 0.001). Outpatient disposition from the ED increased from 29.2% to 46.9%, whereas 7-day secondary hospital admission was unchanged (5.2% vs. 6.1%). ePNa was used by ED clinicians in 67% of eligible patients. Conclusions: ePNa deployment was associated with improved processes of care and lower mortality. Clinical trial registered with www.clinicaltrials.gov (NCT03358342).

Recent advances in deep learning have significantly transformed medical diagnostics, particularly in dermatology. Accurate skin disease detection and classification areessential for effective treatment and improved patient outcomes. This systematic review examines deep learning approaches, including Convolutional Neural Networks (CNNs) and transfer learning, for automated dermatological diagnosis. Public datasets such as HAM10000 and ISIC play a key role in training robust models; however, challenges including dataset imbalance, disease heterogeneity, and overfitting remain. Techniques such as ensemble learning, attention mechanisms, explainable artificial intelligence, data augmentation, hybrid models, and task-specific loss functions have been shown to enhance accuracy, robustness, and interpretability. This study follows a systematic review methodology in accordance with the PRISMA guidelines. The review synthesizes 17 studies published between 2021 and 2024, highlighting the potential of deep learning to support scalable and reliable dermatological diagnostic systems.

BACKGROUND Various media are used to enhance public understanding about diseases. While mobile health (mHealth) applications are widely used, there is little proof for using such apps to raise awareness of skin diseases. OBJECTIVE We intend to develop an app (DEDIKASI-apps) to raise awareness of skin diseases, including leprosy. The study will explore baseline awareness, assess the app’s acceptance among community and healthcare workers, and evaluate its effectiveness in enhancing awareness about skin diseases. METHODS The study will be conducted in 4 phases; (1) development of the DEDIKASI-app; (2) questionnaire development for awareness study; (3) acceptability testing; and (4) effect measurement of the DEDIKASI-app. We’ll adopt Design Thinking Methodology to develop the app, involving systematic-reviews, expert consultations, FGD, and validation process for questionnaire on skin diseases awareness. We’ll recruit 50 members of community for awareness and acceptability study and one healthcare worker per Puskesmas to assess their perception regarding the app. A randomized controlled trial will be conducted to measure the effectiveness of DEDIKASI-app. Result of awareness level will be categorized into low, moderate, and adequate level, each will be presented with proportions of community. RESULTS NA CONCLUSIONS Innovative approaches are required to enhance awareness in the community. The study will introduce new tools and insights to address the limited knowledge about detecting skin problems. CLINICALTRIAL The ethical clearance of the study has been sought from the Health Research Ethics Committee Faculty of Public Health Universitas Airlangga Surabaya with registration number 160/EA/KEPK/2023. The trial in phase four has been registered to the INA-CRR, the trial study registry of the Ministry of Health, Republic of Indonesia with number INA-O8EX276.

The early detection and diagnosis of skin diseases are crucial for effective treatment and management. This paper presents DermaSense AI, an innovative platform that leverages Convolutional Neural Networks (CNNs), specifically the InceptionV3 architecture, to automate the detection of skin diseases with high accuracy. By integrating with Telegram, a popular messaging platform, DermaSense AI offers a user-friendly interface for preliminary diagnosis, enabling users to upload images for analysis. The system addresses challenges in dermatological care, such as limited access, high costs, and subjective diagnostics, by providing a scalable, cost-effective, and efficient solution. This study outlines the development, implementation, and testing of the DermaSense AI platform, demonstrating its potential to enhance dermatological care globally through accessible and reliable diagnostics.

Intelligent fault diagnosis methods based on Transformer variants have shown excellent performance. However, these methods generally model features globally, which means a poor anti-noise ability and greater computational complexity. Therefore, a lightweight framework for bearing fault diagnosis based on depthwise separable convolution (DSC) and sparse attention mechanism named DSC-SparseFormer is proposed. First, a guided-and-sparse self-attention (GSSA) module, which can adaptively adjust the region of interest for features, is proposed to extract features, thereby reducing sensitivity to redundant information and downgrading computational complexity. This is potentially the first time such a methodology has been proposed and applied. Second, discrete cosine transform (DCT) is applied to map time-domain signals into frequency domain to enrich feature space, and a time-frequency coupling attention (TFCA) module is developed to dynamically integrate time-domain and frequency-domain features. Finally, extensive experiments on the HUSTBearing dataset, DHU dataset, and HIT dataset demonstrate that DSC-SparseFormer achieves high-accuracy (HUSTBearing: 99.67%, DHU: 99.94%, and HIT: 99.93%) and lightweight (Params: 0.003 M, FLOPs: 6.07 M) fault diagnosis. The effectiveness of the GSSA and TFCA is validated by ablation studies, and feature visualisation is employed to improve the interpretability of the DSC-SparseFormer.

Aimed to reduce the excessive cost of neural network, this paper proposes a lightweight neural network combining dilated convolution and depthwise separable convolution. Firstly, the dilated convolution is used to expand the receptive field during the convolution process while maintaining the number of convolution parameters, which can extract more high-level global semantic features and improve the classification accuracy of the network. Second, the use of the depthwise separable convolution reduces the network parameters and computational complexity in convolution operations. The experimental results on the CIFAR-10 dataset show that the proposed method improves the classification accuracy of the network while effectively compressing the network size.KeywordsLightweight neural networkDilated convolutionDepthwise separable convolutionClassification accuracyCloud computing

Real-time clinical decision support (CDS) integrated with anesthesia information management systems (AIMS) can generate point of care reminders to improve quality of care. To develop, implement and evaluate a real-time clinical decision support system for anesthetic management of pediatric traumatic brain injury (TBI) patients undergoing urgent neurosurgery. We iteratively developed a CDS system for pediatric TBI patients undergoing urgent neurosurgery. The system automatically detects eligible cases and evidence-based key performance indicators (KPIs). Unwanted clinical events trigger and display real-time messages on the AIMS computer screen. Main outcomes were feasibility of detecting eligible cases and KPIs, and user acceptance. The CDS system was triggered in 22 out of 28 (79%) patients. The sensitivity of detecting continuously sampled KPIs reached 93.8%. For intermittently sampled KPIs, sensitivity and specificity reached 90.9% and 100%, respectively. 88% of providers reported that CDS helped with TBI anesthesia care. CDS implementation is feasible and acceptable with a high rate of case capture and appropriate generation of alert and guidance messages for TBI anesthesia care.

Dogs are beloved pets and loyal companions to millions of people worldwide. Unfortunately, dogs can also suffer from a variety of skin diseases that can cause discomfort, pain, and even life-threatening complications. Some dog skin diseases can be transmitted to humans through direct contact. Early detection and treatment of these skin diseases are crucial for the health and well-being of dogs and humans. Our project aims to provide a quick and precise approach to identifying various types of skin diseases in dogs. To expedite the process of identifying and diagnosing infections related to canine interaction, we plan to utilize a machine-learning model. This approach aims to reduce the time and expertise required for accurate and consistent diagnosis, which can otherwise be challenging and time-consuming. Two models, InceptionV3 and MobileNetV2, were utilized and compared in our implementation. In the case of InceptionV3, a training accuracy of 0.99 and a validation accuracy of 0.98 were achieved. For MobileNetV2, we attained a validation accuracy of 96 and a categorical accuracy of 97. Key Words: Dermatophytosis, zoonosis, image classification, deep learning, Transfer learning, InceptionV3, MobileNetV2, CNN, DNN.

Amid increasing concerns about skin diseases exacerbated by climate change or lifestyle, some diseases are undiagnosed or misdiagnosed due to limited healthcare facilities. The worldwide health burden emphasizes the need for innovative diagnostics. This study explores the evolutionary role of deep learning in skin disease detection, providing the most advanced and effective research approaches, model achievements, and dataset usage exclusively. The review adapts data from 30 research papers and many datasets to address imbalanced class and various efficiency factors. The developments in CNN models like MobileNet or EfficientNet, have strengthened computational potential, while hybrid models have accommodated local and global features. Furthermore, Explainable AI (EXI) and augmented datasets have overcome the challenges including noisy, biased datasets and the less interpretable AI models. This study declares the innovative capacity of deep learning in dermatological analysis, highlighting its scalability and performance. Future research is required to consider dataset diversity, interpretability, and incorporating medical metadata to enhance model performances.

Abstract: Skin infections are more common than many other diseases, and they can be caused by various factors such as infectious pollution, microorganisms, awareness, and infections. With the advancement of lasers and photonics-based medical technology, the diagnosis of skin infections has become faster and more accurate. However, the cost of such diagnosis is still limited and expensive. Therefore, image processing techniques are used to develop an automated assessment system for dermatology at an early stage. The extraction of features plays a crucial role in accurately and quickly diagnosing skin diseases. PC vision plays an important role in the detection of skin diseases in various ways. This study focuses on four skin diseases: ringworm, nail parasite, psoriasis, and atopic dermatitis. Convolutional neural networks have achieved close to or even better performance than humans in the imaging field. The skin diseases are classified using a machine learning algorithm, i.e., random forest, which achieves an accuracy of 98.23% after 100 epochs.

Autism spectrum disorder (ASD) encompasses a wide range of complex neurodevelopment conditions, posing significant challenges to early diagnosis. In fact, other than early diagnosis (i.e., before normal development) it is extremely challenging to adequately classify these subjects as there are no behavioral markers that are unique to ASD vs. non-ASD subjects. But the early detection is very important in helping with some of the developmental trajectories as the timely intervention which they are getting and the access which the kids and families have to specially designed therapies and support systems will be better. Neuroimaging in autism: a comprehensive review from a machine learning perspective. Autism Research- We investigate the effectiveness of CNN models and a hybrid CNN-RNN architecture in classifying ASD based on the neuroimaging data with an aim of improving feature extraction from the neuroimaging data using the Automated Anatomical Labeling (AAL) map. A key challenge in this task is how to build representational feature maps that can elucidate the intricate patterns across brain regions linked with ASD. Various kinds of models are tested VGG16 2D CNN, VGG16 3D CNN, EfficientNetV2, Inception v3, ResNet50 2D CNN and Hybrid model (CNN-RNN). While going for traditional CNN models gave at least 60% to 69% accuracies, the accuracy from hybrid CNN-RNN model, which combines spatial and temporal features, far outperformed other models with an accuracy of 100%. Thus, hybrid deep learning architectures have a substantial role in the classification of ASD. An involvement of AAL mapping and deep learning methods in this study proposed that the semi-automated approach could be implemented in forensic precision for the early diagnosis of ASD. These findings guide future neuroimaging-based diagnostic research and support the development of more sophisticated hybrid models for classifying autism spectrum disorder.

- Skin diseases are among the most prevalent health problems worldwide, affecting millions of people across different age groups and regions. Early and accurate diagnosis is essential to prevent severe complications, particularly in life-threatening cases such as melanoma. However, manual diagnosis through clinical examination or biopsy is often time-consuming, costly, and subjective, leading to inconsistent outcomes. To overcome these limitations, this study proposes an automated Skin Disease Diagnosis System using Machine Learning, which leverages Deep Learning (DL) and Transfer Learning (TL) techniques for precise classification of dermatological images. The proposed system employs a Convolutional Neural Network (CNN) architecture using pre-trained MobileNetV2 and EfficientNetB0 models, fine-tuned on a curated dataset of multiple skin disease classes including benign, malignant (melanoma), eczema, psoriasis, fungal, and normal skin. The images are preprocessed through resizing, normalization, and augmentation to enhance generalization and minimize overfitting. Keywords: Skin Disease Detection, Deep Learning, Machine Learning, MobileNetV2, EfficientNetB0, Transfer Learning, CNN, Image Classification.