Multimedia ResearchISSN:2582-547X

Lightweight Optimisation-Driven Deep Learning for Cardiac Disease from MRI (CMRI)

Abstract

Accurate diagnosis of cardiac diseases from Magnetic Resonance Imaging (MRI) remains a challenging task due to the need for high predictive performance while maintaining computational efficiency. This paper presents a computing-centric and efficiency-aware deep learning framework for cardiac MRI analysis, explicitly designed for diagnostic and clinical research. The proposed model is based on a two-layer GhostNet architecture enhanced with a combined channel-attention mechanism that integrates squeeze-and-excitation (SE) operations, enabling effective recalibration of feature channels with minimal parameter overhead. Network training is performed using an improved Parrot Optimisation Algorithm, which is algorithmically and mathematically formulated to adaptively optimise model parameters, resulting in improved convergence behaviour and generalisation capability. To further reduce computational complexity, structured pruning is systematically applied and quantitatively analysed, with explicit evaluation of parameter count, floating-point operations (FLOPs), memory footprint, and accuracy trade-offs. This pruning strategy enables the development of a compact diagnostic model suitable for deployment on resource-constrained computing platforms without substantial degradation in predictive performance. The framework is evaluated on benchmark cardiac MRI datasets using accuracy, precision, recall, and F1-score, alongside comprehensive computing-efficiency metrics. Experimental results demonstrate a favourable accuracy–complexity balance and significant reductions in model size and computational cost compared to conventional convolutional neural network architectures.

References

  • Alasmari, Sultan, Rayed AlGhamdi, Ghanshyam G. Tejani, Sunil Kumar Sharma, and Seyed Jalaleddin Mousavirad. 2025. Federated learning-based multimodal approach for early detection and personalised care in cardiac disease. Frontiers in Physiology, 16, p.1563185.
  • Qiu, Yuting, James Meng, and Baihua Li. 2025. Semi-supervised Strong-Teacher Consistency Learning for few-shot cardiac MRI image segmentation. Computer Methods and Programs in Biomedicine, 261, p.108613.
  • Qian, Xianling, Yali Wu, Peter Speier, Caixia Fu, Yunzhu Wu, Lude Cheng, Yinyin Chen et al. 2025. Comparison of Pilot Tone-Triggered and ECG-Triggered Cardiac MRI: A Prospective Clinical Feasibility Study. Journal of Cardiovascular Magnetic Resonance, p.101925.
  • Vink, Caitlin EM, Sonia Borodzicz-Jazdzyk, Elize AM de Jong, Janneke Woudstra, Tim P. van de Hoef, Steven AJ Chamuleau, Etto C. Eringa, Marco JW Götte, and Yolande Appelman. 2025. Quantitative perfusion by cardiac MRI reveals compromised myocardial perfusion in patients with angina with non-obstructive coronary artery disease. Clinical Research in Cardiology, pp.1-11.
  • Zheng, Weizeng, Xia Ying, Yuan Chen, Le Wang, Peiyue Jiang, Ying Jiang, Guohui Yan et al. 2025. Association Between Fetal Myocardial Alterations and Congenital Heart Disease Based on Post-Mortem Myocardial MRI. Iradiology, 3(3), pp.222-230.
  • Udomluck, P., Cholamjiak, W., Inpun, J. and Woratamrongpatai, W., 2026. Multi-Task Deep Learning Model for Automated Detection and Severity Grading of Lumbar Spinal Stenosis on MRI: Multi-Centre External Validation. Diseases, 14(1), p.32.
  • Turgut, Özgün, Philip Müller, Paul Hager, Suprosanna Shit, Sophie Starck, Martin J. Menten, Eimo Martens, and Daniel Rueckert. 2025. Unlocking the diagnostic potential of electrocardiograms through information transfer from cardiac magnetic resonance imaging. Medical Image Analysis, 101, p.103451.
  • Alsekait, Deema Mohammed, Ahmed Younes Shdefat, Ayman Nabil, Asif Nawaz, Muhammad Rizwan Rashid Rana, Zohair Ahmed, Hanaa Fathi, and Diaa Salama AbdElminaam. 2024. Heart-Net: A Multi-Modal Deep Learning Approach for Diagnosing Cardiovascular Diseases. Computers, Materials & Continua, 80(3).
  • Paciorek, Aleksandra M., Claudio E. von Schacky, Sarah C. Foreman, Felix G. Gassert, Florian T. Gassert, Jan S. Kirschke, Karl-Ludwig Laugwitz, Tobias Geith, Martin Hadamitzky, and Jonathan Nadjiri. 2024. Automated assessment of cardiac pathologies on cardiac MRI using T1-mapping and late gadolinium phase sensitive inversion recovery sequences with deep learning. BMC Medical Imaging, 24(1), p.43.
  • Iqbal, Talha, Aaleen Khalid, and Ihsan Ullah. 2024. Explaining decisions of a light-weight deep neural network for real-time coronary artery disease classification in magnetic resonance imaging. Journal of Real-Time Image Processing, 21(2), p.31.
  • Koechli, Melina, Fraser M. Callaghan, Barbara EU Burkhardt, Maélène Lohézic, Xucheng Zhu, Beate Rücker, Emanuela R. Valsangiacomo Buechel, Christian J. Kellenberger, and Julia Geiger. 2024. Accelerated cardiac magnetic resonance imaging using deep learning for volumetric assessment in children. Pediatric Radiology, 54(10), pp.1674-1685.
  • Phair, Andrew, Anastasia Fotaki, Lina Felsner, Thomas J. Fletcher, Haikun Qi, René M. Botnar, and Claudia Prieto. 2024. A motion-corrected deep-learning reconstruction framework for accelerating whole-heart magnetic resonance imaging in patients with congenital heart disease. Journal of Cardiovascular Magnetic Resonance, 26(1), p.101039.
  • Zhao, Bo, Jianjun Peng, Ce Chen, Yongyan Fan, Kai Zhang, and Yang Zhang. 2025. Diagnosis of Coronary Heart Disease through Deep Learning-Based Segmentation and Localization in Computed Tomography Angiography. IEEE Access.
  • Lyu, Donghang, Marius Staring, Matthias JP van Osch, Mariya Doneva, Hildo J. Lamb, and Nicola Pezzotti. 2026. Convolutional recurrent U-net for cardiac cine MRI reconstruction via effective spatio-temporal feature exploitation. Medical Physics, 53(1), p.e70245.
  • Yang, Yang, Maosheng Fu, Xiancun Zhou, Chaochuan Jia, and Peng Wei. 2025. A Multi-Strategy Parrot Optimization Algorithm and Its Application. Biomimetics, 10(3), p.153.