MAMumtahina Ahmed

Mumtahina Ahmed

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Mumtahina Ahmed has a total of 35 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    FEDI-CODE: A federated and causally-informed framework for dementia risk prediction using multi-site patient data

    by Mohammad Moniruzzaman, Md Shahab Uddin, Ahsan Ahmed, Md Aktarujjaman, Mumtahina Ahmed, Ashifur Rahman

    Early detection of dementia is critical for timely intervention and disease management, yet it remains a challenging task due to the fragmented nature of healthcare data and the need for privacy-preserving solutions. This paper proposes FEDI-CODE, a Federated and Causally Informed Dementia Estimation framework that integrates deep learning, federated learning, and counterfactual inference to predict dementia risk across distributed patient data sources. FEDI-CODE is designed to operate without centralizing sensitive medical data, enabling collaborative training across institutions while preserving privacy. It combines temporal modeling of longitudinal imaging and clinical data with individualized treatment effect estimation for modifiable risk factors such as alcohol consumption, weight, and cardiovascular indicators. A fusion module aggregates representations from each site to form a global prediction head. Extensive experiments on simulated multi-site dementia datasets demonstrate that FEDI-CODE achieves an accuracy of 83.7%, a precision of 83%, a recall of 81%, an F1-score of 82%, and an AUC-ROC of 0.86, outperforming standard federated models and deep learning baselines by notable margins. The model also generalizes well to external datasets, achieving 79.2% accuracy and 0.80 AUC-ROC, confirming its robustness. Furthermore, FEDI-CODE produces interpretable causal insights by estimating individual treatment effects, offering actionable clinical value. These results highlight FEDI-CODE as a scalable, interpretable, and privacy-aware solution for early dementia screening and personalized risk assessment.


  • Open Access

    Cross-Attention Multi-Modal Transformer for Early Diagnosis of Parkinson's Disease

    by Jungpil Shin, Jannatul Maua, Mir Nafiul Nagib, Ahsan Ahmed, Mumtahina Ahmed, M. F. Mridha

    Early and accurate diagnosis of Parkinson's Disease (PD) is essential for timely intervention and management. While prior studies have used unimodal signals like voice, gait, or handwriting separately, real-world screening requires integrating diverse biomedical data. This work proposes XAMT++, a Cross-Attention Multi-Modal Transformer that jointly processes voice (Oxford Telemonitoring), gait (Daphnet FoG), and handwriting (Spiral Drawings) data for improved PD diagnosis. XAMT++ includes (i) modality-specific encoders, (ii) a cross-attention fusion module, and (iii) a hierarchical transformer backbone. It is trained with a hybrid supervised and contrastive alignment loss to remain robust even when some modalities are missing. On the three public datasets, XAMT++ achieves 0.87 classification accuracy, 0.88 F1 score, 0.89 AUROC, and 0.90 AUPRC, outperforming unimodal and standard fusion baselines. Notably, it maintains strong early-stage detection performance, achieving 0.60 accuracy within 0–30 days post-diagnosis, with further improvements as the disease progresses. Inter-modal alignment metrics confirm high representation consistency, with a cosine similarity of 0.87 and a mutual information of 0.89. These results demonstrate XAMT++ as a practical and effective multi-modal diagnostic framework for early PD detection.


  • Open Access

    Physics-informed self-supervised diagnosis of rotating machinery using latent ODEs and transformer encoders

    by Al Amin, Mohammad Shafat Ahsan, Jannatul Maua, Mumtahina Ahmed, Kamruddin Nur

    This paper proposes a novel Physics-Informed Self-Supervised Diagnosis (PI-SSD) framework for rotating machinery fault detection, combining physical modeling, self-supervised representation learning, and uncertainty-aware classification. The architecture integrates a multi-resolution convolutional encoder, a windowed Transformer for temporal context modeling, and a latent neural ordinary differential equation (ODE) module that embeds mechanical priors, such as Jeffcott rotor dynamics, directly into the learning process. A masked segment reconstruction objective enables self-supervised pretraining using unlabeled healthy signals, while an evidential classifier head produces fault probabilities with calibrated uncertainty. We evaluate PI-SSD on two publicly available datasets, the NASA PHM’09 Gearbox dataset and the Aalto Rotor dataset, covering 6 fault types and over 5,500 multichannel vibration recordings. Compared to seven strong baselines, PI-SSD achieves the highest Macro-F1 score (0.91) and lowest Expected Calibration Error (ECE = 0.022) on the NASA dataset, while maintaining strong domain transfer performance on Aalto (Macro-F1 = 0.81, PR-MSE = 0.067) without fine-tuning. Ablation studies confirm the contribution of each component, and physics consistency analysis demonstrates low violation rates under varying speeds. These results highlight the potential of embedding physics knowledge into self-supervised neural systems for robust, interpretable, and transferable fault diagnosis in industrial applications.


  • Open Access

    Two-Stage Hybrid Deep Learning Architecture for Cross-Domain Anomaly Detection and Failure Prediction

    by Rahomotul Islam, Nafiza Humauara Hasan Neha, Mumtahina Ahmed, Jannatul Maua, Neha Chaudhary, M. F. Mridha, Nilanjan Dey, Satyabrata Roy

    Industrial systems require effective anomaly detection and predictive maintenance strategies to prevent unexpected failures and optimize operational efficiency. However, traditional approaches often treat anomaly detection and failure prediction as independent tasks, limiting their ability to leverage operational patterns for early fault diagnosis. This paper proposes a novel two-stage hybrid deep learning framework that integrates unsupervised anomaly detection with supervised failure prediction through cross-domain transfer learning. The framework first employs a variational autoencoder (VAE) to extract latent anomaly embeddings from HVAC system operational data without requiring labeled anomalies. These embeddings are transferred and combined with sensor readings from a predictive maintenance dataset to enrich the failure prediction model. Experimental evaluations on two publicly available datasets demonstrate that the proposed hybrid model achieves superior performance compared to conventional methods, with notable improvements in accuracy (90.2%), precision (0.89), recall (0.90), F1-score (0.90), and AUC-ROC (0.873). Additionally, the approach exhibits strong cross-domain generalization ability, highlighting its potential applicability across different industrial environments. This work introduces a scalable and flexible anomaly-driven transfer learning strategy that addresses the scarcity of labeled failure data and enhances the robustness of predictive maintenance systems. The findings suggest that integrating unsupervised operational insights into supervised learning frameworks is a promising direction for advancing intelligent industrial monitoring.


  • Open Access

    Dual-Encoder Adversarial Learning for Cloud-Based Cyber Intrusion Detection

    by Al Amin, Mohammad Shafat Ahsan, Rahomotul Islam, Mumtahina Ahmed, M. F. Mridha, Jakir Hossen

    Cloud-based systems have become prime targets for increasingly sophisticated cyberattacks due to their distributed, multi-tenant, and dynamic nature. Traditional intrusion detection systems often fail to generalize across diverse cloud environments, particularly when dealing with heterogeneous data sources such as network traffic and user behavior. In this paper, we propose a novel Dual-Encoder Adversarial Learning framework designed to detect cyber intrusions in cloud systems through a unified representation of multi-domain features. The architecture integrates private encoders for domain-specific signal extraction, a shared encoder for domain-invariant representation learning, and adversarial training implemented via a Gradient Reversal Layer to enforce cross-domain generalization. A reconstruction decoder is also employed to retain semantic fidelity in latent embeddings. Extensive experiments conducted on two benchmark datasets, UNSW-NB15 and the Cybersecurity Intrusion Detection dataset, demonstrate that the proposed model consistently outperforms state-of-the-art baselines in terms of accuracy (0.917), precision (0.903), recall (0.910), F1-score (0.906), and AUC (0.947). The model also exhibits strong robustness under noisy conditions and achieves competitive efficiency in training time and memory usage. These results validate the effectiveness of the proposed architecture for efficient and robust intrusion detection in cloud environments.


  • Open Access

    Lead-Aware Multi-Resolution Transformer With Domain Adaptation for Beat-Level ECG Arrhythmia Classification

    by Masuduzzaman Niloy, Tanzimul Islam, Shafiq Ullah, Jobayar Alom, Mumtahina Ahmed, M. F. Mridha, Jakir Hossen

    Accurate and generalizable ECG classification remains a critical challenge in automated cardiac diagnostics, particularly when dealing with multi-lead signals from heterogeneous sources. In this paper, we propose MuRe-LAT, a Multi-Resolution Lead-Aware Transformer architecture with domain-adaptive normalization designed for robust arrhythmia detection. The model integrates a convolutional stem for local feature extraction, a lead attention module to adaptively weigh channel contributions, dual-scale transformer encoders for capturing both short-term and long-term dependencies, and a domain-specific batch normalization mechanism to mitigate dataset shift. To enhance generalization, MuRe-LAT is pretrained using a masked signal modeling objective. We evaluate the model on two benchmark datasets MIT-BIH Arrhythmia and PTB-XL spanning both ambulatory and clinical ECG recordings. On MIT-BIH, MuRe-LAT achieves an accuracy of 94.0%, a macro-F1 of 88.4, an AUROC of 96.2, and a PR-AUC of 91.6; on PTB-XL, it reaches 88.7% accuracy, 78.1 macro-F1, and 90.5 AUROC. The model also demonstrates strong cross-domain generalization, achieving up to 91.7% accuracy when transferred from PTB-XL to MIT-BIH without retraining. Compared to strong baselines including ResNet-1D, BiLSTM, ECG-BERT, and CardioFormer, MuRe-LAT outperforms or closely matches state-of-the-art results in over 90% of evaluation settings. With only 12.3M parameters and an inference time of 8.4 ms per 10-second ECG window, MuRe-LAT offers a compelling balance between accuracy and efficiency, making it suitable for deployment in real-time clinical and wearable applications.


  • Open Access

    A Dual-Encoder Approach to Bankruptcy Prediction With Financial Ratios and Corporate Disclosures

    by Tohidul Islam, Jannatul Maua, Rahomotul Islam, Mumtahina Ahmed, M. F. Mridha, Jakir Hossen

    Accurate and early prediction of corporate bankruptcy is critical for financial risk assessment and decision-making. Traditional models rely solely on structured financial ratios, overlooking valuable qualitative information embedded in corporate disclosures. This paper proposes a multimodal deep learning framework that integrates structured accounting data with unstructured Management Discussion and Analysis (MD&A) text to improve bankruptcy prediction. The model employs a dual-encoder architecture with a gated fusion mechanism to dynamically combine numerical and textual representations. We evaluate our approach on three publicly available datasets: the U.S. Corporate Bankruptcy dataset, the Polish Companies Bankruptcy dataset, and the ECL dataset. Experimental results demonstrate that the proposed model consistently outperforms classical statistical methods and unimodal deep learning baselines across multiple metrics, including accuracy, F1-score, ROC-AUC, and PR-AUC. The model also generalizes well across domains and time horizons, offering a robust and interpretable tool for financial risk analysis. This work highlights the importance of leveraging both quantitative and qualitative signals for more reliable bankruptcy forecasting.


  • Open Access

    A systematic review of AI-driven business models for advancing Sustainable Development Goals

    by Azam Khan, Arifur Rahman, Farhad Uddin Mahmud, Kanchon Kumar Bishnu, Mumtahina Ahmed, M.F. Mridha, Zeyar Aung


  • Open Access

    Optimizing Magnetic Resonance Imaging Analysis for Brain Tumors: A Lightweight Neural Network Approach

    by Zakir Hossain Zamil, Raisul Islam, Anisur Rahman, Mumtahina Ahmed, Nahid Hasan, Mohsin Kabir