Jul 2026· Annual International Computer Software and Applications Conference· pp. 2290-2295· 0 citations· 20 references
Abstract
Bacterial pharyngitis requires timely antibiotic treatment, whereas most non-bacterial cases are self-limited; diagnostic errors may therefore lead to missed infections or unnecessary antibiotic use. This study proposes PharyTriFuse, a multimodal framework that integrates throat-image analysis with large language model (LLM) reasoning and a medical knowledge graph (KG) to classify bacterial versus non-bacterial pharyngitis from smartphone-acquired oropharyngeal images. Experiments were conducted on the public PGUPharyngitis dataset over 742 images using a stratified 72%/8%/20% train/validation/test split. Images were standardized using CLAHE and redness enhancement to reduce acquisition variability. Two visual backbones (EfficientNet-B4 and ConvNeXt-Base) were evaluated under four configurations: AI-only, AI+LLM, AI+KG, and AI+LLM+KG. Performance was assessed using standard classification metrics and inference efficiency. Results show that incorporating LLM reasoning and structured medical knowledge improves classification performance over vision-only baselines while maintaining real-time inference capability under certain configurations. These findings suggest that multimodal AI systems can enhance smartphone-based decision support for pharyngitis assessment.
An advanced system capable of automatically detecting complicated appendicitis from ultrasound images was developed and was explained with gradient-weighted class activation mapping (Grad-CAM), which creates a heatmap of the regions responsible for the model's prediction of the infected areas.
Fahad Ahammed, Omar Faruq Shikdar, Navid Zaman et al.· 0 citations
Skin cancer cases are still increasing in the world at a very fast rate and early diagnostic intervention is pivotal in improving patient survival. While traditional clinical interpretation of skin lesion is a slow and labour-intensive subjective process which heavily relies on dermatologic expertise, this process can be greatly enhanced by automated image analysis via deep learning framework. In this study, a deep learning model was designed and tested for skin lesion classification with multi-classes. The dataset HAM10000 contains 10,015 dermoscopic images from seven different diagnostic categories. A random oversampling technique was used to overcome the basic imbalances of the datasets, and a strong image enhancement pipeline was implemented to improve the visibility of features. Five different transfer learning architectures (Xception, InceptionResNetV2, MobileNetV3Large, DenseNet121 and NASNetMobile) were used with the same custom classification head and trained under the same experimental settings. A challenging-to-classify minority sample problem was effectively solved by using a categorical focal loss. The empirical results showed that Xception architecture obtained the best performance with an F1-score of 0.98 and a rate of accuracy of 97.77%. This was followed sequentially by InceptionResNetV2 (94.40% accuracy, 0.94 F1-score), MobileNetV3Large (93.86% accuracy, 0.94 F1-score), DenseNet121 (87.35% accuracy, 0.87 F1-score), and NASNetMobile (75.12% accuracy, 0.75 F1-score). A detailed analysis was done for each model to calculate F1-scores, per-class accuracy and a confusion matrix, as well as training and validation trajectories were created. Based on the results of the experiments, the proposed pipeline is highly competitive and outperforms the recent best results obtained on the same data set. In conclusion, this research proves that transfer learning, systematic class balancing, and focal loss functions provide a computationally viable approach and highly effective method for automatic skin cancer classification, while also highlighting that the backbone architecture is the key factor that determines the effectiveness of classification under similar training conditions.
Yusra Shafiq, H. M. Shahzad· Journal of innovative resear...· 0 citations
Background: Deep-learning systems for endoscopic image classification are commonly evaluated with random data splits, which may overestimate performance under acquisition shift; uncertainty-aware selective prediction may improve reliability by allowing a model to abstain on uncertain cases. Methods: We evaluated binary abnormality detection in two endoscopic imaging domains: laryngeal contact-endoscopy narrow-band imaging (CE-NBI; 210 patients, patient-level) and gastrointestinal endoscopy (HyperKvasir; 6746 images, image-level, predominantly white-light). ImageNet-pretrained ResNet-50 deep ensembles were assessed under a resolution-defined acquisition-shift stress test; for the gastrointestinal data, a random stratified split was additionally used as an in-distribution reference. We evaluated discrimination, calibration, decision-curve analysis, and entropy-based selective prediction. Results: In the gastrointestinal dataset, random-split evaluation produced high performance (AUROC 0.994, 95% CI 0.990–0.996; AUPRC 0.991). Under resolution shift on the same data, performance fell to AUROC 0.723 (95% CI 0.710–0.736; AUPRC 0.690; sensitivity 0.417; specificity 0.874); the two intervals do not overlap. Selective prediction improved reliability among retained cases: under resolution shift, accuracy rose from 0.683 at full coverage to 0.852 (95% CI 0.835–0.871) at 25% coverage (balanced accuracy 0.646 → 0.798). Predictive entropy was significantly higher for incorrect than for correct predictions in both regimes (Mann–Whitney p = 7.1 × 10−77 with rank-biserial |r| = 0.30 under shift). In the laryngeal cohort, no statistically significant differences were detected among four architectures (ROC-AUC 0.844–0.901; all pairwise DeLong p > 0.05). Conclusions: Random-split evaluation substantially overestimated performance relative to a resolution-defined acquisition-shift stress test, and entropy-based selective prediction improved reliability by identifying a high-confidence subset for automated prediction while deferring the remainder to human review. Target-domain recalibration substantially restores calibration under shift (ECE 0.172 → 0.036 with temperature scaling; → 0.017 with isotonic regression) but does not recover discrimination; selective prediction is complementary, mitigating residual confident-wrong predictions. An encoder-transfer experiment showed asymmetric cross-domain utility; features learned on the larger gastrointestinal cohort transferred to the laryngeal cohort (AUROC 0.80 vs. in-domain 0.89), whereas the reverse direction did not transfer (0.53 vs. 0.72). Prospective multi-center validation remains required before clinical deployment. All code, fold definitions, random seeds, and a reproducible protocol are publicly released.
Behnam Kiani Kalejahi, S. Khan, Murodbek Akhrorov et al.· Biomedicines· 0 citations
The feasibility of developing a deep learning model for the detection of intussusception using a smaller dataset of POCUS images is demonstrated and fine-tuning models were best adapted to the screening nature of POCUS images.
A. Thyagachandran, Brian Lefchak, H. Murthy et al.· Frontiers in Radiology· 0 citations
Diagnosing skin diseases remains a clinical challenge due to the visual similarity among diverse dermatological conditions. This study presents a prototype deep learning–powered system for multi-class dermatological screening, implemented through a cloud-based architecture and accessed via a smartphone interface. A pre-trained Convolutional Neural Network (CNN), EfficientNetV2B3, was fine-tuned on a composite dataset encompassing nine disease categories. The model achieved promising performance, with an accuracy of 0.87, precision of 0.87, recall of 0.87, and an F1 score of 0.86, indicating its potential reliability for automated classification. Prototype validation was conducted using a cloud API (Google Cloud Storage + PostMan) to verify the inference pipeline and user interaction. While the current implementation demonstrates the feasibility of cloud-based dermatological screening, real-device mobile performance metrics such as latency, model size, and memory consumption remain future work. Users can capture or upload skin images, which are processed to generate preliminary diagnostic feedback, including symptom descriptions and general treatment information. While not intended to replace professional medical evaluation, the prototype serves as a proof-of-concept tool for initial screening and early intervention. This work illustrates how artificial intelligence (AI) can be harnessed in mobile health applications to expand access to dermatological care and supports broader initiatives to integrate AI into healthcare delivery.
Theetach Rabablert, Amonnat Kaewnok, C. Sirisathitkul et al.· The Scientist· 0 citations
Mpox lesions can resemble other dermatological conditions, motivating image-based screening, yet published studies remain difficult to compare owing to differences in dataset construction, augmentation policy, and evaluation design. This study provides a leakage-aware benchmark for binary mpox classification using a unified dataset assembled from MSLD v1.0 and v2.0. Seven pretrained backbones and a weighted ensemble were compared under group-stratified five-fold cross-validation with original-only test evaluation, validation-based threshold selection, and temperature scaling. The weighted ensemble achieved mean accuracy 0.8729, F1-score 0.8334, and AUC 0.9388; ConvNeXt-Tiny was the strongest single model (F1 0.8159, AUC 0.9284). These grouped original-only results are intentionally conservative relative to augmentation-heavy or single-split designs and should be interpreted as deflated but more trustworthy reference values. Post hoc calibration analysis, content-level near-duplicate auditing, and a test-time augmentation ablation are provided to substantiate the methodological claims. The contribution is methodological: a transparent benchmark emphasizing reproducible dataset curation, grouped evaluation, and calibrated comparison, while highlighting the limitations of current public skin-image data. Accordingly, these results should be interpreted as a reproducible reference benchmark rather than a clinically validated diagnostic tool, and external clinical validation remains necessary before deployment.
Bassam W. Aboshosha, Shafiq Ul Rehman, L. N. Mahmoud et al.· PLoS ONE· 1 citation