Dr Hamid Abbasi

PhD, MSE, BSE

Profile Image
Research Fellow

Biography

Hamid is a Research Fellow in the Auckland Bioengineering Institute (ABI) and the Department of Physiology at the University of Auckland with a longer-standing interest in advanced signal and image processing of biomedical data. He received his Ph.D. in 2017 in Engineering Science on developing automated signal processing algorithms for early detection of EEG biomarkers of hypoxic-ischemic brain injury at birth. He also holds M.E and B.E degrees in Control, Electronics and Instrumentation engineering. 

His main area of interest is the application of advanced signal processing strategies and innovative machine learning approaches such as deep-learning platforms in the biomedical field and neuroscience. His research interests expands to finding translations from lab experiments to the clinic, brain development, pattern recognition and classification, medical image processing and data science. Hamid is an active member of the IEEE and the IEEE Engineering in Medicine and Biology Society and the IEEE Brain community.

Research | Current

Hypoxic-Ischemic (HI) brain injury due to lack of cerebral oxygen at or around the time of birth leads to severe neurodevelopmental impairments, disability, and death. The injury significantly contributes in debilitating morbidity and lifelong disabilities such as cerebral palsy, epilepsy, cognitive disorders and learning difficulties.

In term neonates, the window of opportunity for treatments such as therapeutic hypothermia (or brain cooling) is critically short and has been shown to be optimal if the treatment is initiated within the first 3-6 hours of birth, only. Conditions are often more complicated for premature babies whom hypothermia is not applicable to and whom seldom present symptoms due to their frailty. Currently, diagnosis of the injury in high-risk preterm infants is limited by the reliance on manual clinical observation. Thus, there is an urgent need to find ways to better assess and automatically accelerate the detection of high-risk preterms with HI.

Hamid has been focused on developing automated advanced signal and image processing algorithms based on deep machine-learning approaches for the accurate identification of translational Electroencephalogram (EEG) patterns, known to be the HI biomarkers in animal models, from an exclusive data bank of clinical EEG recordings of preterm infants, straight after birth. As a result, the developed pattern recognition strategies are able to precisely classify EEG biomarker from background activity and artifact, in real-time, with highly competitive accuracies compared to a clinical expert. The key outcomes of his study will hopefully help with early diagnosis of HI brain injury, at bed-side, within the most critical first hours of life, where the current therapeutic protocols have been shown to be optimally neuroprotective in.

Hamid is interested in developing further insights for future expansions, including further capabilities of deep-learning strategies in biomedical engineering and neuroscience, along with the importance of employing high-tech facilities to achieve ultimate outcomes.

 

Research interests

  • Advanced signal and image processing,
  • Machine learning, deep-learning, and convolutional neural networks (CNN)
  • Biomedical and medical sciences
  • Brain, neurophysiology, neuroscience, and cardiology
  • Neural networks, wavelets and wavelet-based neural networks, fuzzy systems
  • Control and instrumentation,
  • Data monitoring and analysis,
  • Non-linear model predictive control.
     

Google scholar profile

 

Research group

Signal processing

Teaching | Current

  • ENGSCI 313 - Mathematical Modelling 3ECE
  • ENGGEN 140 - Engineering Biology and Chemistry
  • ENGSCI211
  • ENGSCI313
  • ENGSCI712

Distinctions/Honours

  • New Zealand Health Research Council (HRC) Scholarship 
  • UniServices Commercialization Prize, 2014 Spark $100k Challenge (with BabyAware)
  • Engineering Postgraduate Poster Competition 2014, 2nd place in Engineering Science
  • University of Auckland Doctoral Scholarship - 2016

Areas of expertise

  • Advanced Signal Processing,
  • Biomedical Engineering
  • Machine learning and Deep Convolutional Neural Networks (CNN)
  • Neural Networks, Wavelets and Wavelet-Based Neural Networks, Fuzzy systems
  • Data Monitoring and Analysis,
  • Electrical, Control and Instrumentation Engineering,

Committees/Professional groups/Services

  • Member of Early Career Researchers committee at the Center for Brain Research

  • Member of Faculty of Medical and Health Sciences Post-doctoral Society at the University of Auckland

  • Member of the Institute of Electrical and Electronics Engineers (IEEE)

  • Member of IEEE Signal Processing Society

  • Member of Journal of Biomedical and Health Informatics, IEEE

  • Member of IEEE Young Professionals
  • Member of the Exposure organizing committee, the University of Auckland, 2014 and 2015 (Posters Team Lead)
  • Member of the executive committee of the Velocity/Spark, the University of Auckland, 2015 (Research Team Lead)

Selected publications and creative works (Research Outputs)

  • Abbasi, H., Saberi, S., Zarvani, M., Amiri, P., & Azmi, R. (2020). Deep Learning Classification Schemes for the Identification of COVID-19 Infected Patients using Large Chest X-ray Image Dataset. IEEE Engineering in Medicine and Biology Society (EMBC) Montreal, Canada. Related URL.
    URL: http://hdl.handle.net/2292/51786
  • Abbasi, H., Gunn, A. J., Bennet, L., & Unsworth, C. P. (2020). Wavelet Spectral Deep-training of Convolutional Neural Networks for Accurate Identification of High-Frequency Micro-Scale Spike Transients in the Post-Hypoxic-Ischemic EEG of Preterm Sheep. Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS, 1011-1014. Montreal, Canada: IEEE. 10.1109/EMBC44109.2020.9176397
    Other University of Auckland co-authors: Alistair Gunn, Laura Bennet
  • Abbasi, H., Gunn, A. J., Bennet, L., & Unsworth, C. P. (2020). Deep Convolutional Neural Network and Reverse Biorthogonal Wavelet Scalograms for Automatic Identification of High Frequency Micro-Scale Spike Transients in the Post-Hypoxic-Ischemic EEG. Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS, 1015-1018. Montreal, Canada: IEEE. 10.1109/EMBC44109.2020.9176499
    Other University of Auckland co-authors: Alistair Gunn, Laura Bennet
  • Abbasi, H., Gunn, A. J., Unsworth, C. P., & Bennet, L. (2020). Wavelet Spectral Time-Frequency Training of Deep Convolutional Neural Networks for Accurate Identification of Micro-Scale Sharp Wave Biomarkers in the Post-Hypoxic-Ischemic EEG of Preterm Sheep. Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS Montreal, Canada: IEEE. 10.1109/EMBC44109.2020.9176057
    Other University of Auckland co-authors: Alistair Gunn, Laura Bennet
  • Abbasi, H., Gunn, A. J., Unsworth, C. P., & Bennet, L. (2020). Deep Convolutional Neural Networks for the Accurate Identification of High-Amplitude Stereotypic Epileptiform Seizures in the Post-Hypoxic-Ischemic EEG of Preterm Fetal Sheep. Paper presented at 42nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC'20), Montreal, Canada. 20 July - 24 July 2020. Related URL.
    Other University of Auckland co-authors: Alistair Gunn, Laura Bennet
  • Abbasi, H., & Unsworth, C. P. (2020). Electroencephalogram studies of hypoxic ischemia in fetal and neonatal animal models. Neural Regeneration Research, 15 (5), 828-837. 10.4103/1673-5374.268892
  • Abbasi, H., Gunn, A. J., Bennet, L., & Unsworth, C. P. (2020). Latent Phase Identification of High-Frequency Micro-Scale Gamma Spike Transients in the Hypoxic Ischemic EEG of Preterm Fetal Sheep Using Spectral Analysis and Fuzzy Classifiers. Sensors, 20 (5)10.3390/s20051424
    Other University of Auckland co-authors: Alistair Gunn, Laura Bennet
  • Abbasi, H., & Unsworth, C. P. (2020). Applications of advanced signal processing and machine learning in the neonatal hypoxic-ischemic electroencephalogram. Neural Regeneration Research, 15 (2), 222-231. 10.4103/1673-5374.265542
    Other University of Auckland co-authors: Charles Unsworth