Ahmadreza Argha
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Ahmadreza Argha has a total of 85 co-authors. Here are the top co-authors by number of shared publications:
- Branko George Celler
Faculty of Engineering, University of New South Wales, Sydney, Australia
27 shared publications - Hamid Alinejad-Rokny
UNSW Sydney
11 shared publications - Nigel Hamilton Lovell
UNSW Sydney
9 shared publications - Steven Weidong Su
Shandong First Medical University
9 shared publications - Siaw-Teng Liaw
Discipline of General Practice, School of Clinical Medicine, UNSW Sydney, Australia
7 shared publications - Andrey V. Savkin
School of Electrical Engineering and Telecommunications, University of New South Wales, Sydney, NSW 2052, Australia
5 shared publications
Latest Publications
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Engineering
- Vol. 73 (10) , 3400-3411
ECG-Adapt: A Novel Framework for Robust Electrocardiogram Classification Across Diverse Populations and Recording Conditions
The electrocardiogram (ECG) is a vital diagnostic tool used to monitor and diagnose a wide range of cardiac conditions. However, ECG signals can exhibit significant variability across different patient populations, recording devices, and environmental conditions, creating challenges in developing universally robust and accurate classification models. This research addresses these challenges by exploring and advancing domain adaptation techniques to enhance the robustness and generalizability of ECG classification models. By leveraging unsupervised domain adaptation (UDA), we aim to mitigate the performance degradation that typically occurs when models trained on one dataset are applied to another, thereby improving diagnostic accuracy and reliability across diverse clinical settings. We introduce ECG-Adapt, an integrated approach that aligns features both within classes and across domains. Unlike existing methods that rely on clustering as a preprocessing step, ECG-Adapt does not require clustering, simplifying the workflow. It further incorporates weakly supervised learning to prevent overfitting of the discriminator to pseudo-labels generated by the classifier, enhancing robustness and generalizability. Applying our novel unsupervised domain adaptation framework led to substantial performance gains. For instance, ECG-Adapt improved the average $F_{1}$-score by 8% on single-lead problems and 7% on 12-lead problems. By leveraging ECG-Adapt, performance degradation when applying models across datasets can be mitigated, enhancing diagnostic accuracy and reliability in diverse clinical settings and demonstrating strong potential for real-world deployment.
- 1 October 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Engineering
- Vol. 73 (10) , 3400-3411
ECG-Adapt: A Novel Framework for Robust Electrocardiogram Classification Across Diverse Populations and Recording Conditions
The electrocardiogram (ECG) is a vital diagnostic tool used to monitor and diagnose a wide range of cardiac conditions. However, ECG signals can exhibit significant variability across different patient populations, recording devices, and environmental conditions, creating challenges in developing universally robust and accurate classification models. This research addresses these challenges by exploring and advancing domain adaptation techniques to enhance the robustness and generalizability of ECG classification models. By leveraging unsupervised domain adaptation (UDA), we aim to mitigate the performance degradation that typically occurs when models trained on one dataset are applied to another, thereby improving diagnostic accuracy and reliability across diverse clinical settings. We introduce ECG-Adapt, an integrated approach that aligns features both within classes and across domains. Unlike existing methods that rely on clustering as a preprocessing step, ECG-Adapt does not require clustering, simplifying the workflow. It further incorporates weakly supervised learning to prevent overfitting of the discriminator to pseudo-labels generated by the classifier, enhancing robustness and generalizability. Applying our novel unsupervised domain adaptation framework led to substantial performance gains. For instance, ECG-Adapt improved the average $F_{1}$-score by 8% on single-lead problems and 7% on 12-lead problems. By leveraging ECG-Adapt, performance degradation when applying models across datasets can be mitigated, enhancing diagnostic accuracy and reliability in diverse clinical settings and demonstrating strong potential for real-world deployment.
- Open Access
- 31 August 2026
- Published by Wiley in Advanced Science
- Vol. 13 (64)
SemanticST: A Scalable Multi‐Contextual Graph Learning Framework for Uncovering Spatial Niches and Robust Multi‐Sample Integration in Spatial Transcriptomics
Spatial transcriptomics (ST) analysis is often hindered by technical limitations and methodological biases that let dominant signals overshadow subtle but crucial biological patterns, such as rare cell types and fine‐grained heterogeneity. This is especially true for high‐complexity datasets from platforms like Xenium. We present SemanticST, a graph neural network (GNN) framework that addresses this challenge through a fundamentally different design. SemanticST is the first GNN method to implement mini‐batch training, enabling scalable analysis of massive ST datasets (validated on Xenium). Crucially, it employs a multi‐semantic graph fusion strategy that learns disentangled biological representations across tissue, using a min‐cut loss that requires neither graph corruption nor contrastive sampling. Benchmarking across diverse tissues (e.g., brain, embryo, tumor) confirms consistent superiority. It achieves up to 20% higher ARI/NMI on the gold‐standard brain cortex and uniquely delineates all mouse olfactory bulb layers and hippocampal sub‐regions. In high‐resolution breast cancer data, SemanticST identifies computationally plausible spatial domains, including a candidate rare triple receptor‐positive region and a FOXC2‐enriched EMT‐associated domain, from Xenium alone. Furthermore, SemanticST provides superior, robust multi‐sample integration on established benchmarks. SemanticST offers an essential, scalable framework for translating spatial complexity into biologically informative and testable hypotheses.
- 31 August 2026
- Published by Wiley in Advanced Science
- Vol. 13 (64)
Open AccessSemanticST: A Scalable Multi‐Contextual Graph Learning Framework for Uncovering Spatial Niches and Robust Multi‐Sample Integration in Spatial Transcriptomics
Spatial transcriptomics (ST) analysis is often hindered by technical limitations and methodological biases that let dominant signals overshadow subtle but crucial biological patterns, such as rare cell types and fine‐grained heterogeneity. This is especially true for high‐complexity datasets from platforms like Xenium. We present SemanticST, a graph neural network (GNN) framework that addresses this challenge through a fundamentally different design. SemanticST is the first GNN method to implement mini‐batch training, enabling scalable analysis of massive ST datasets (validated on Xenium). Crucially, it employs a multi‐semantic graph fusion strategy that learns disentangled biological representations across tissue, using a min‐cut loss that requires neither graph corruption nor contrastive sampling. Benchmarking across diverse tissues (e.g., brain, embryo, tumor) confirms consistent superiority. It achieves up to 20% higher ARI/NMI on the gold‐standard brain cortex and uniquely delineates all mouse olfactory bulb layers and hippocampal sub‐regions. In high‐resolution breast cancer data, SemanticST identifies computationally plausible spatial domains, including a candidate rare triple receptor‐positive region and a FOXC2‐enriched EMT‐associated domain, from Xenium alone. Furthermore, SemanticST provides superior, robust multi‐sample integration on established benchmarks. SemanticST offers an essential, scalable framework for translating spatial complexity into biologically informative and testable hypotheses.
- Open Access
- 16 June 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1) , 5405
Interpretable graph-based models on multimodal biomedical data integration: a technical review and benchmarking
Integrating diverse biomedical modalities is essential for robust healthcare insights, and graph-based models are increasingly used to capture complex relational structures. Yet, their clinical translation hinges on interpretability. This review surveys interpretable graph-based models applied to multimodal biomedical data, highlighting dominant trends in disease classification, static graph construction, and post-hoc explainability. We categorize explainable artificial intelligence (XAI) techniques, benchmark SHAP, saliency, sensitivity, and graph masking on Alzheimer’s disease data, and reveal complementary strengths. A development flowchart and future directions, such as dynamic graphs, knowledge integration, and LLM-based explainability, position this work as a key reference for trustworthy biomedical AI.
- 16 June 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1) , 5405
Open AccessInterpretable graph-based models on multimodal biomedical data integration: a technical review and benchmarking
Integrating diverse biomedical modalities is essential for robust healthcare insights, and graph-based models are increasingly used to capture complex relational structures. Yet, their clinical translation hinges on interpretability. This review surveys interpretable graph-based models applied to multimodal biomedical data, highlighting dominant trends in disease classification, static graph construction, and post-hoc explainability. We categorize explainable artificial intelligence (XAI) techniques, benchmark SHAP, saliency, sensitivity, and graph masking on Alzheimer’s disease data, and reveal complementary strengths. A development flowchart and future directions, such as dynamic graphs, knowledge integration, and LLM-based explainability, position this work as a key reference for trustworthy biomedical AI.
- 9 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Robotics and Bionics
- Vol. 8 (2) , 785-797
Practical Multiple Model Switching sEMG Decoders for Motion Intention Description in Rehabilitation Robot Control
Accurate decoding of human motion intention from surface electromyography (sEMG) is crucial for smooth and responsive rehabilitation robot control. This study develops a multiple-model sEMG decoder incorporating a practical bumpless switching strategy that enables seamless transitions among submodels for complex multi-joint movements. The method identifies a minimal common state representation across submodels with different dimensions, ensuring stable and continuous output without abrupt changes during task transitions. Several state-resetting schemes are proposed and experimentally compared in both numerical simulations and real-time human–robot cooperative arm motion tasks. Results demonstrate that the proposed decoder maintains smooth force estimation and robust performance across diverse movement conditions, supporting its applicability to real-time rehabilitation control. Due to safety constraints regarding potential abrupt switching in human-in-the-loop trials, a limited cohort was used; consequently, these findings should be interpreted as exploratory.
- 9 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Robotics and Bionics
- Vol. 8 (2) , 785-797
Practical Multiple Model Switching sEMG Decoders for Motion Intention Description in Rehabilitation Robot Control
Accurate decoding of human motion intention from surface electromyography (sEMG) is crucial for smooth and responsive rehabilitation robot control. This study develops a multiple-model sEMG decoder incorporating a practical bumpless switching strategy that enables seamless transitions among submodels for complex multi-joint movements. The method identifies a minimal common state representation across submodels with different dimensions, ensuring stable and continuous output without abrupt changes during task transitions. Several state-resetting schemes are proposed and experimentally compared in both numerical simulations and real-time human–robot cooperative arm motion tasks. Results demonstrate that the proposed decoder maintains smooth force estimation and robust performance across diverse movement conditions, supporting its applicability to real-time rehabilitation control. Due to safety constraints regarding potential abrupt switching in human-in-the-loop trials, a limited cohort was used; consequently, these findings should be interpreted as exploratory.
- Open Access
- 28 January 2026
- Published by Wiley in Advanced Science
- Vol. 13 (12) , e12842
CLinNET: An Interpretable and Uncertainty‐Aware Deep Learning Framework for Multi‐Modal Clinical Genomics
Identifying molecular drivers and diagnostic genes for neurocognitive disorders (NDs) remains a major challenge due to the prevalence of variants of uncertain significance (VUS) and limitations in current diagnostic platforms. While artificial intelligence (AI) offers potential solutions, existing models often lack interpretability and fail to address uncertainty, limiting clinical utility. CLinNET, a multi‐modal deep neural network with a dual‐branch design integrating sequencing data, gene expression, biological pathways, and gene ontology (GO) is introduced to enhance gene curation and VUS interpretation. CLinNET employs a biologically informed architecture, confidence‐based uncertainty quantification, and layer‐wise SHapley Additive exPlanations (SHAP) for robust interpretability. Its sparse networks, enriched with pathway and GO data, prioritize tissue‐expressed genes to improve prediction accuracy and biological relevance. Trained on ND datasets, CLinNET outperformed existing methods with an F1‐score of 76.4%, accuracy of 77.2%, and area under the precision‐recall curve (AUC‐PR) of 84%. Incorporating uncertainty filtering further improved precision to 87% while retaining 73% of predictions as high‐confidence. CLinNET identified significantly more ND‐associated genes than random permutations, with minimal overlap with cardiovascular‐associated genes, confirming specificity. Among the top decile of ranked genes, 78 were linked to NDs ( p ‐value = 1.2e‐11), and 372 to rare diseases involving nervous system abnormalities, highlighting their diagnostic potential. CLinNET's validation in prostate cancer datasets underscores its adaptability, positioning it as a robust tool for individualized medicine.
- 28 January 2026
- Published by Wiley in Advanced Science
- Vol. 13 (12) , e12842
Open AccessCLinNET: An Interpretable and Uncertainty‐Aware Deep Learning Framework for Multi‐Modal Clinical Genomics
Identifying molecular drivers and diagnostic genes for neurocognitive disorders (NDs) remains a major challenge due to the prevalence of variants of uncertain significance (VUS) and limitations in current diagnostic platforms. While artificial intelligence (AI) offers potential solutions, existing models often lack interpretability and fail to address uncertainty, limiting clinical utility. CLinNET, a multi‐modal deep neural network with a dual‐branch design integrating sequencing data, gene expression, biological pathways, and gene ontology (GO) is introduced to enhance gene curation and VUS interpretation. CLinNET employs a biologically informed architecture, confidence‐based uncertainty quantification, and layer‐wise SHapley Additive exPlanations (SHAP) for robust interpretability. Its sparse networks, enriched with pathway and GO data, prioritize tissue‐expressed genes to improve prediction accuracy and biological relevance. Trained on ND datasets, CLinNET outperformed existing methods with an F1‐score of 76.4%, accuracy of 77.2%, and area under the precision‐recall curve (AUC‐PR) of 84%. Incorporating uncertainty filtering further improved precision to 87% while retaining 73% of predictions as high‐confidence. CLinNET identified significantly more ND‐associated genes than random permutations, with minimal overlap with cardiovascular‐associated genes, confirming specificity. Among the top decile of ranked genes, 78 were linked to NDs ( p ‐value = 1.2e‐11), and 372 to rare diseases involving nervous system abnormalities, highlighting their diagnostic potential. CLinNET's validation in prostate cancer datasets underscores its adaptability, positioning it as a robust tool for individualized medicine.
- 6 December 2025
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 275, 109202
Cluster-Based Insights into Cardiovascular and Autonomic Responses to Head-Up Tilt in Hypertension
- 6 December 2025
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 275, 109202
Cluster-Based Insights into Cardiovascular and Autonomic Responses to Head-Up Tilt in Hypertension
- Open Access
RxSafeBench: Identifying Medication Safety Issues of Large Language Models in Simulated Consultation
Numerous medical systems powered by Large Language Models (LLMs) have achieved remarkable progress in diverse healthcare tasks. However, research on their medication safety remains limited due to the lack of real world datasets, constrained by privacy and accessibility issues. Moreover, evaluation of LLMs in realistic clinical consultation settings, particularly regarding medication safety, is still underexplored. To address these gaps, we propose a framework that simulates and evaluates clinical consultations to systematically assess the medication safety capabilities of LLMs. Within this framework, we generate inquiry diagnosis dialogues with embedded medication risks and construct a dedicated medication safety database, RxRisk DB, containing 6,725 contraindications, 28,781 drug interactions, and 14,906 indication-drug pairs. A two-stage filtering strategy ensures clinical realism and professional quality, resulting in the benchmark RxSafeBench with 2,443 high-quality consultation scenarios. We evaluate leading open-source and proprietary LLMs using structured multiple choice questions that test their ability to recommend safe medications under simulated patient contexts. Results show that current LLMs struggle to integrate contraindication and interaction knowledge, especially when risks are implied rather than explicit. Our findings highlight key challenges in ensuring medication safety in LLM-based systems and provide insights into improving reliability through better prompting and task-specific tuning. RxSafeBench offers the first comprehensive benchmark for evaluating medication safety in LLMs, advancing safer and more trustworthy AI-driven clinical decision support.
Open AccessRxSafeBench: Identifying Medication Safety Issues of Large Language Models in Simulated Consultation
Numerous medical systems powered by Large Language Models (LLMs) have achieved remarkable progress in diverse healthcare tasks. However, research on their medication safety remains limited due to the lack of real world datasets, constrained by privacy and accessibility issues. Moreover, evaluation of LLMs in realistic clinical consultation settings, particularly regarding medication safety, is still underexplored. To address these gaps, we propose a framework that simulates and evaluates clinical consultations to systematically assess the medication safety capabilities of LLMs. Within this framework, we generate inquiry diagnosis dialogues with embedded medication risks and construct a dedicated medication safety database, RxRisk DB, containing 6,725 contraindications, 28,781 drug interactions, and 14,906 indication-drug pairs. A two-stage filtering strategy ensures clinical realism and professional quality, resulting in the benchmark RxSafeBench with 2,443 high-quality consultation scenarios. We evaluate leading open-source and proprietary LLMs using structured multiple choice questions that test their ability to recommend safe medications under simulated patient contexts. Results show that current LLMs struggle to integrate contraindication and interaction knowledge, especially when risks are implied rather than explicit. Our findings highlight key challenges in ensuring medication safety in LLM-based systems and provide insights into improving reliability through better prompting and task-specific tuning. RxSafeBench offers the first comprehensive benchmark for evaluating medication safety in LLMs, advancing safer and more trustworthy AI-driven clinical decision support.
- Open Access
- 29 August 2025
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 161
Gradient surgery: A necessity for robust test-time adaptation for detecting casting defects
- 29 August 2025
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 161
Open AccessGradient surgery: A necessity for robust test-time adaptation for detecting casting defects
- Open Access
MultiECGNet: A novel deep learning-based multi-format ensemble method for image-based electrocardiographic diagnosis of atrial fibrillation
Aim To evaluate the performance of an ensemble classifier, MultiECGNet, using multi-format electrocardiographic (ECG) images for the diagnosis of atrial fibrillation (AF), and to compare its performance with a signal-based deep learning model. Methods An ensemble of ECG classifiers was developed using four models derived by truncating the pre-trained EfficientNet B3 model at different feature extraction layers. Transfer learning was employed to train the ensemble on the publicly available PTB-XL dataset for AF detection. External validation was performed using the Chinese Physiological Society Signal Challenge 2018 dataset. ECG samples were converted into two image formats (2x6 and 4x3), and performance was evaluated across same-format, cross-format, and mixed-format classification tasks. Results The ensemble classifier detected AF in the external validation dataset with an accuracy of 0.95 and F1 score of 0.87, which was comparable to a signal-based model (F1-score: 0.87 vs 0.83) and outperformed a single EfficientNet-B3 model (F1-score: 0.87 vs 0.71). Training on one ECG format and testing on a different format resulted in reduced performance (F1 score: 0.66-0.71). However, training on a dataset containing a balanced mix of both formats improved performance compared to same-format training alone (F1-score: 0.86 vs 0.87). Conclusion The proposed image-based ensemble classifier demonstrated comparable performance to a strong signal-based model for AF detection. While cross-format generalization posed a challenge, incorporating multiple ECG formats during training mitigated this limitation and improved model robustness.
Open AccessMultiECGNet: A novel deep learning-based multi-format ensemble method for image-based electrocardiographic diagnosis of atrial fibrillation
Aim To evaluate the performance of an ensemble classifier, MultiECGNet, using multi-format electrocardiographic (ECG) images for the diagnosis of atrial fibrillation (AF), and to compare its performance with a signal-based deep learning model. Methods An ensemble of ECG classifiers was developed using four models derived by truncating the pre-trained EfficientNet B3 model at different feature extraction layers. Transfer learning was employed to train the ensemble on the publicly available PTB-XL dataset for AF detection. External validation was performed using the Chinese Physiological Society Signal Challenge 2018 dataset. ECG samples were converted into two image formats (2x6 and 4x3), and performance was evaluated across same-format, cross-format, and mixed-format classification tasks. Results The ensemble classifier detected AF in the external validation dataset with an accuracy of 0.95 and F1 score of 0.87, which was comparable to a signal-based model (F1-score: 0.87 vs 0.83) and outperformed a single EfficientNet-B3 model (F1-score: 0.87 vs 0.71). Training on one ECG format and testing on a different format resulted in reduced performance (F1 score: 0.66-0.71). However, training on a dataset containing a balanced mix of both formats improved performance compared to same-format training alone (F1-score: 0.86 vs 0.87). Conclusion The proposed image-based ensemble classifier demonstrated comparable performance to a strong signal-based model for AF detection. While cross-format generalization posed a challenge, incorporating multiple ECG formats during training mitigated this limitation and improved model robustness.