Khandaker Mohammad Mohi Uddin

Background and Aims: Diabetic foot ulcers (DFUs), which heal slowly because of inadequate blood supply, are among the serious infections and chronic foot problems that can result from diabetes mellitus (DM). The primary consequence of DFU, which can result in amputation if left untreated. Methods: The paper proposes an intelligent and automated approach for classifying foot images as either healthy or DFU images. At the initial stage, the proposed system presents a novel dataset containing 5500 foot images collected from diverse individuals with healthy and DFU conditions. The preprocessing stage involves three steps: the Region of Interest (ROI) method removes unnecessary portions of the foot images, the RGB images are converted to grayscale, and Non‐Local Means (NLM) filtering is applied to reduce noise and remove unwanted information. Two neural feature extractors, ResNet50 and Faster R‐CNN, are used to independently extract features from the foot images, and the resulting feature vectors are integrated into a single fused feature. The softmax function of Faster RCNN method classifies DFU or normal image using fused features, and bounding box method regressor function localize the ulcer region from DFU image. Results: Compared with relevant state‐of‐the‐art methods, the proposed Faster RCNN‐based deep learning approach with feature fusion demonstrates superior performance in DFU recognition, achieving a testing accuracy of 99.85%, specificity of 99.37%, and precision of 99.50%. With a detection accuracy of 98.83%, this fusion‐based approach demonstrated competitive performance through the use of a generalization validation mechanism. Conclusion: For automated diabetic foot ulcer identification and localization, the suggested fusion‐based Faster R‐CNN framework shows very accurate and dependable performance. These results point to its significant potential as a helpful clinical decision‐making tool for the early identification and treatment of DFU.

Background and Aims: Polycystic ovary syndrome (PCOS) is the most prevalent endocrine ailment impacting women of reproductive age, distinguished by ongoing imbalances of hormones that lead to the growth of various ovarian cysts alongside other health issues. Infrequent or delayed menstrual periods, and also frequently excessive amounts of androgen, a male hormone, are characteristics of women with PCOS. Despite other consequences, this disease may lead to weight gain, undesirable body hair, type 2 diabetes, and gestational diabetes. The number of cases of PCOS has shockingly ascended in the past few years, according to statistics. But the clinical diagnosis procedure for PCOS in the real world becomes critical since the precision of interpretations greatly relies on the knowledge of the doctor. Therefore, a PCOS prediction model using artificial intelligence might be an appropriate supplement to the lengthy and error‐prone diagnosis method. Methods: To strictly prevent data leakage, the dataset underwent an 80/20 train‐test split before applying SMOTE and two‐stage feature selection exclusively to the training data. This framework combined Chi‐Square filtering with Recursive Feature Elimination to identify the most pertinent features. A hybrid XGBoost (XGB) and Multi‐Layer Perceptron (MLP) ensemble was selected from multiple machine learning models following rigorous evaluation using 10‐fold stratified cross‐validation. Finally, SHapley Additive exPlanations and Local Interpretable Model‐Agnostic Explanations were integrated to significantly enhance overall model transparency. Results: The hybrid ensemble model (XGB + MLP) outperformed conventional single ML models, with the greatest classification accuracy of 96.33% among all evaluated models. The model's interpretability and performance were greatly enhanced by the 16 characteristics that were chosen. Conclusion: Combining hybrid ensemble learning, two‐stage feature selection, and XAI approaches provides a computationally efficient, dependable, and interpretable method for PCOS diagnosis. Practitioners may find this model to be a useful decision‐support tool that improves the accuracy of diagnosis and lessens the need for human interpretation.

Nowadays, respiratory illnesses are common among children around the world, which is contributing significantly to child mortality due to rapid growth. These diseases appear with several symptoms, and several may share similar symptoms; the most well-known include asthma, bronchiolitis, and pneumonia. The mortality rate associated with chronic pediatric respiratory conditions is increasing annually, underscoring the need to assess the severity of these diseases. This research aims to improve the mortality prediction in pediatric respiratory disorders by using a robust machine learning architecture and a comprehensive dataset from the Children’s Hospital of Rabat, Morocco. Ten machine learning algorithms were implemented after thorough preprocessing, which included encoding, feature selection, oversampling, and hyperparameter tuning. Random Forest, Logistic Regression, AdaBoost, Multilayer Perceptron, K-Nearest Neighbors, Decision Trees, CatBoost, XGBoost, Naïve Bayes, and Gradient Boosting have been applied to get better results. With an accuracy rate of 99.29%, a precision of 92.68%, a recall rate of 92.55%, and an F1-score of 92.58%, Random Forest performed more effectively than the other methods. Explainable artificial intelligence techniques, such as Shapley Additive Values and Local Interpretable Model-agnostic Explanations, ensure clinician trust and model interpretability. Received: 22 November 2025 | Revised: 3 April 2026 | Accepted: 3 June 2026 Conflicts of Interest The authors declare that they have no conflicts of interest to this work. Data Availability Statement The data that support the findings of this study are openly available in the Mendeley Data repository at https://data.mendeley.com/datasets/2d9dvnycjw/1. Author Contribution Statement Khandaker Mohammad Mohi Uddin: Conceptualization, Methodology, Writing – review & editing, Supervision. Anjuman Ansary: Software, Validation, Formal analysis, Investigation, Writing – original draft. Nazim Uddin: Visualization, Project administration. Md. Tofael Ahmed Bhuiyan: Resources, Data curation.

In Bangladesh, crop leaf diseases create a serious risk to food security and production from agriculture. Timely identification of leaf diseases in rice, wheat, and bean crops is considered crucial for the implementation of effective disease detection and classification strategies. To address this challenge, a MobilenetV2-based disease identification and classification system is proposed in this research. Previous studies focus on classifying diseases of a single species, leaving the need to train models separately for each species. This research focuses on forming a single standard model to perform leaf disease classification for multiple crop species including rice, wheat, and beans. The approach makes use of transfer learning with the MobilenetV2 model, which is fine-tuned using a dataset of annotated crop leaf images specific to Bangladesh. Following a comprehensive evaluation, an overall accuracy of 97.87% was achieved in the classification of crop leaf diseases, which surpasses the accuracy of a number of previous studies focusing on leaf disease detection of a single crop. The system demonstrates the capability to rapidly diagnose diseases in real time by enabling the users to prompt intervention to mitigate potential crop losses, ultimately leading to amplified crop yield and food security. Overall, the research highlights the promise of AI-powered solutions in tackling crop leaf disease detection, which in turn encourages greater research and technology adoption to support sustainable farming methods especially in the crop disease classification domain in Bangladesh and throughout the world. Received: 24 May 2025 | Revised: 9 March 2026 | Accepted: 14 April 2026 Conflicts of Interest The authors declare that they have no conflicts of interest to this work. Data Availability Statement The data that support the findings of this study are openly available in the Bangladeshi Crops Disease Dataset at https://www.kaggle.com/datasets/nafishamoin/bangladeshi-crops-disease-dataset and the Bean Disease Dataset at https://www.kaggle.com/datasets/therealoise/bean-disease-dataset. Author Contribution Statement Md. Mahmudul Hasan: Conceptualization, Methodology, Visualization, Supervision. Md. Omar Faruq: Software, Validation, Writing – original draft. Mahadi Hasan Musa: Formal analysis, Investigation. Mohammad Mamunur Rashid: Resources, Data curation, Writing – review & editing. Khandaker Mohammad Mohi Uddin: Writing – review & editing, Project administration, Supervision.

Objective: Asthma, a chronic respiratory condition characterized by airway inflammation and constriction, affects millions of individuals worldwide, resulting in high healthcare expenses and a lower quality of life. Early prediction and control of asthma risk are critical for avoiding exacerbations and improving outcomes. Methods: In this study, we describe a comprehensive asthma prediction model that uses machine learning and deep learning techniques to estimate asthma risk based on a variety of health and environmental parameters. Recursive feature elimination and Extra Trees Classifier were used to choose features, and the synthetic minority over-sampling approach was used to balance the dataset to overcome class imbalance. Hyperparameter tuning was used to optimize performance for 12 machine learning models such as extreme gradient boosting, Random Forest, and support vector machine as well as deep learning models, including multilayer perceptrons, convolutional neural networks, recurrent neural network, and artificial neural network. Results: After hyperparameter adjustment, ensemble approaches that used both hard and soft voting were evaluated. When hyperparameter adjustment was used, the soft voting ensemble that combined XGBoost and CatBoost achieved the highest accuracy (93.61%). Shapley additive explanations and local interpretable model-agnostic explanations were employed to make predictions interpretable, providing information on feature contributions and boosting clinician confidence. A Flask server and web interface were also deployed, enabling real-time user interaction where patients and medical professionals could enter data and obtain asthma risk estimations immediately. Conclusions: This study presents an accurate and explainable asthma risk prediction framework using ensemble machine and deep learning models, achieving 93.61% accuracy with real-time clinical applicability.

Objectives: Heart disease is currently one of the most serious diseases, contributing to a large share of patient deaths. Diagnosing it is a challenging procedure that needs to be done both quickly and accurately. Methods: This research introduces a practical machine learning framework aimed at improving heart disease prediction using multiple datasets. To build a more reliable model, three separate datasets were first combined into a single, enriched source of data. Since the data was imbalanced, a hybrid balancing technique using the Instance Hardness Threshold (IHT) and Tomek Links was applied to ensure fairer learning across classes. All preprocessing (scaling, balancing, hyperparameter tuning) was done exclusively within the training set of each fold in stratified 5-fold cross validation, with the hyperparameter tuning further embedded in an inner 3-fold cross validation. A total of 12 machine learning models were trained and compared, with the top-performing models ensembled using soft- and hard-voting. Results: The most successful model, a soft voting combination of XGBoost, Extra Trees, Random Forest, MLP and Decision Tree, outperformed the others with an accuracy of 96.02% (95% CI: 95.47 – 96.56%) with a precision of 97.83% and a recall of 94.37% using the fold-wise protocol designed to prevent preprocessing leakage. In addition to accuracy, calibration analysis and clinical decision-curve analysis showed that the model’s predicted probabilities are suitable for a clinical context, and a sensitivity–specificity analysis revealed an optimal decision threshold of 0.323. Finally, SHAP was applied to clarify the impact of each feature; chest pain type and thalassemia status were found to be the most influential predictors. Conclusion: The proposed ensemble model, based on a hybrid-balanced SHAP interpretable model holds great promise for clinical applications and transparency in heart disease prediction.

Objective: Esophageal cancer is among the most rapidly spreading malignancies worldwide. Early detection of esophageal cancer is critical for disease prevention and for improving overall population health. Most studies have used statistical methodologies to assess the esophageal cancer risk, and only a few studies have used prediction models. Methods: The esophageal cancer dataset, comprising 3985 patient records with 85 demographic, pathological, and follow-up features, was obtained from Kaggle. A comprehensive data-engineering pipeline was implemented, including the removal of null and low-variance features, elimination of identifier variables to prevent data leakage, mode-based imputation, label encoding, and data standardization. Feature relevance was assessed using Mutual Information, and the top 31 clinically meaningful features were retained for model development. Six machine learning classifiers—Support Vector Machine, Gaussian Naïve Bayes, k-nearest neighbors, AdaBoost, Multilayer Perceptron, and LightGBM (Gradient Boosting Machine)—were trained and evaluated. A stratified 10-fold cross-validation was applied to maintain class balance, and GridSearchCV was used for hyperparameter optimization. Model interpretability was assessed using Shapley Additive Explanations (SHAP) for global and local feature attribution and Local Interpretable Model-Agnostic Explanations (LIME) for instance-level explanations. Furthermore, the top features identified by SHAP and LIME were used to retrain the LightGBM model to evaluate performance under reduced dimensionality. Results: Among all evaluated classifiers, LightGBM exhibited the highest and most stable performance, achieving an accuracy of 99.87% prior to hyperparameter tuning and 99.74% following stratified cross-validated tuning, with near-perfect precision, recall, F1-score, and area under the curve values. Explainability analyses indicated that clinically relevant variables, including tumor staging, smoking-related factors, and follow-up indicators, played a significant role in model predictions. The SHAP-selected top-20 feature model maintained high predictive performance (99.76%), demonstrating that the classifier remained robust despite dimensionality reduction. Conclusions: The proposed LightGBM-based model demonstrates exceptional predictive accuracy and strong interpretability, suggesting its potential utility for the early detection of esophageal cancer using machine learning approaches.

Celiac disease is a chronic autoimmune disorder that primarily affects individuals genetically predisposed to it, causing damage to the small intestine upon gluten consumption. Although it affects around one person out of every hundred worldwide, celiac disease is frequently misdiagnosed. Eliminating serious health problems from celiac disease requires an early and precise diagnosis, and modern diagnostic techniques hold great promise for increasing diagnostic precision. Still, there is a dearth of comprehensive review literature that provides insights into upcoming developments and summarizes the state of the subject now. The diagnostic utility of machine learning algorithms, biomarker analysis, and imaging methods for celiac disease is evaluated critically in this study. The focus is on the remarkable 99.72% accuracy obtained with a random forest–based stacking ensemble method and the 98.5% diagnostic accuracy obtained with the GEO dataset (GDS3646, celiac disease: primary leukocytes, GPL6104, GSE11501). Additionally, the investigate methods to enhance model interpretability by integrating SHapley filter, embedded, and wrapper techniques are combined with a hybrid feature selection strategy improved using GridSearchCV and K‐fold cross‐validation and SHapley Additive exPlanations (SHAP), as well as Local Interpretable Model‐agnostic Explanations (LIME). The noninvasive attraction of biomarker evaluation, the merits of sophisticated diagnostic procedures, and the potential of machine learning in expediting celiac disease diagnosis are all highlighted in this paper. The argumentation also includes challenges related to integration into medical protocols and standards. Future studies will concentrate on improving testing techniques for widespread clinical use, integrating protein and genetic information to improve early diagnosis and tailoring therapies to improve the lives of patients.