Dr Ali Gooya

  • Senior Lecturer in Machine Learning (School of Computing Science)

Biography

Ali Gooya is currently a Senior Lecturer in the School of Computing Science (IDA-Section) at the University of Glasgow, Scotland, UK. Before joining the University of Glasgow in 2022, he was an Associate Professor in the School of Computing at the University of Leeds (2018-2022) and the University of Sheffield (2016-2018). He has won multiple prestigious fellowships, including Allen Touring Institute (2022), JSPS short-term invitational (2020), FP7 Marie-Curie IIF fellowship (2014), and Japan Society for Promotion of Science JSPS-PDRA (2008). He earned his PhD from the University of Tokyo in medical image analysis (2007). He joined the University of Pennsylvania as a post-doc research associate to develop machine learning methods for medical vision (2008-2011).

Research interests

My research interest broadly lies in machine learning, computer vision and medical imaging. I am particularly keen on deep probabilistic learning with the target applications in cancer and cardiac image analysis, computer-aided decision support systems, prediction and marker discovery, statistical inference on populations, and computational anatomy.

My research vision is to aspire to unsupervised machine learning for AI in healthcare, as expert annotations in this particular field are sparse. I am experienced in creating methodologically innovative deep Bayesian frameworks, often involving rigorous mathematical modelling.

Publications

List by: Type | Date

Jump to: 2023 | 2022 | 2020 | 2019
Number of items: 6.

2023

Elhaminia, B., Gilbert, A., Lilley, J., Abdar, M., Frangi, A. F., Scarsbrook, A., Appelt, A. and Gooya, A. (2023) Toxicity prediction in pelvic radiotherapy using multiple instance learning and cascaded attention layers. IEEE Journal of Biomedical and Health Informatics, (doi: 10.1109/JBHI.2023.3238825) (Early Online Publication)

Zakeri, A., Hokmabadi, A., Bi, N., Wijesinghe, I., Nix, M. G., Petersen, S. E., Frangi, A. F., Taylor, Z. A. and Gooya, A. (2023) DragNet: learning-based deformable registration for realistic cardiac MR sequence generation from a single frame. Medical Image Analysis, 83, 102678. (doi: 10.1016/j.media.2022.102678) (PMID:36403308)

2022

Appelt, A.L., Elhaminia, B., Gooya, A. , Gilbert, A. and Nix, M. (2022) Deep learning for radiotherapy outcome prediction using dose data – a review. Clinical Oncology, 34(2), e87-e96. (doi: 10.1016/j.clon.2021.12.002) (PMID:34924256)

Zakeri, A., Hokmabadi, A., Ravikumar, N., Frangi, A. F. and Gooya, A. (2022) A probabilistic deep motion model for unsupervised cardiac shape anomaly assessment. Medical Image Analysis, 75, 102276. (doi: 10.1016/j.media.2021.102276) (PMID:34753021)

2020

Kumar, N. et al. (2020) A multi-organ nucleus segmentation challenge. IEEE Transactions on Medical Imaging, 39(5), pp. 1380-1391. (doi: 10.1109/TMI.2019.2947628) (PMID:31647422)

2019

Zhang, L., Gooya, A. , Pereanez, M., Dong, B., Piechnik, S.K., Neubauer, S., Petersen, S.E. and Frangi, A.F. (2019) Automatic assessment of full left ventricular coverage in cardiac cine magnetic resonance imaging with Fisher-discriminative 3-D CNN. IEEE Transactions on Biomedical Engineering, 66(7), pp. 1975-1986. (doi: 10.1109/TBME.2018.2881952)

This list was generated on Mon Mar 27 14:47:11 2023 BST.
Jump to: Articles
Number of items: 6.

Articles

Elhaminia, B., Gilbert, A., Lilley, J., Abdar, M., Frangi, A. F., Scarsbrook, A., Appelt, A. and Gooya, A. (2023) Toxicity prediction in pelvic radiotherapy using multiple instance learning and cascaded attention layers. IEEE Journal of Biomedical and Health Informatics, (doi: 10.1109/JBHI.2023.3238825) (Early Online Publication)

Zakeri, A., Hokmabadi, A., Bi, N., Wijesinghe, I., Nix, M. G., Petersen, S. E., Frangi, A. F., Taylor, Z. A. and Gooya, A. (2023) DragNet: learning-based deformable registration for realistic cardiac MR sequence generation from a single frame. Medical Image Analysis, 83, 102678. (doi: 10.1016/j.media.2022.102678) (PMID:36403308)

Appelt, A.L., Elhaminia, B., Gooya, A. , Gilbert, A. and Nix, M. (2022) Deep learning for radiotherapy outcome prediction using dose data – a review. Clinical Oncology, 34(2), e87-e96. (doi: 10.1016/j.clon.2021.12.002) (PMID:34924256)

Zakeri, A., Hokmabadi, A., Ravikumar, N., Frangi, A. F. and Gooya, A. (2022) A probabilistic deep motion model for unsupervised cardiac shape anomaly assessment. Medical Image Analysis, 75, 102276. (doi: 10.1016/j.media.2021.102276) (PMID:34753021)

Kumar, N. et al. (2020) A multi-organ nucleus segmentation challenge. IEEE Transactions on Medical Imaging, 39(5), pp. 1380-1391. (doi: 10.1109/TMI.2019.2947628) (PMID:31647422)

Zhang, L., Gooya, A. , Pereanez, M., Dong, B., Piechnik, S.K., Neubauer, S., Petersen, S.E. and Frangi, A.F. (2019) Automatic assessment of full left ventricular coverage in cardiac cine magnetic resonance imaging with Fisher-discriminative 3-D CNN. IEEE Transactions on Biomedical Engineering, 66(7), pp. 1975-1986. (doi: 10.1109/TBME.2018.2881952)

This list was generated on Mon Mar 27 14:47:11 2023 BST.

Grants

  • EPSRC New Investigator Grant (as PI, EP/S012796/1)
  • 2 x Cancer Research UK sandpit awards on early detection of cancer
  • Innovate UK KTP application (as Co-PI)
  • Marie-Curie IIF Fellowship
  • JSPS Short Invitational Fellowship
  • JSPS PDRA Fellowship
  • 2x NVIDIA GPU grant in aid

Supervision

I am looking for PhD students interested in Deep Learning, Probabilistic and Generative Modelling applied to Medical Imaging and patient meta-data. Potential project directions are listed below, and other projects are due. The candidate is expected to have strong analytical and math skills, good programming experience, some prior experience in machine learning and visual computing, and good English communication skills. Please get in touch with me (ali.gooya@glasgow.ac.uk) for further information.

Bayesian Deep Atlases for Cardiac Motion Abnormality Assessment by Integrating  Imaging and Metadata 

Cardiovascular diseases (CVDs) are the second biggest killer in the UK; currently, more than 7 million people live with CVD in the country. Early identification of individuals with significant risk is critical to improve the patient's quality of life and reduce the financial burden on the social and healthcare systems. Many CVDs lead to a shortage of blood supply to the heart muscle, and abnormal motion is diagnosed non-invasively by analysing the patient's dynamic cardiac imaging data. Manual assessment of these images is subjective, non-reproducible, limited to the left ventricle, and time-consuming. Statistical atlases, describing the 'average' pattern of the heart motion over a sizeable healthy population, can be potentially helpful in identifying deviations from normality in individuals. However, the integration of the existing atlases into clinical practice is inhibited by three fundamental limitations: (i) the derived motion statistics are often independent of the patient's age, gender, weight, etc. (metadata) that are essential for precise diagnosis, (ii) the detected abnormalities due to failure of heart segmentation could not be disentangled from the underlying clinical conditions.

To alleviate these fundamental limitations, this proposal aims, for the first time, to develop a complete probabilistic atlas to evaluate bi-ventricular motion abnormalities accurately.

  • Holistically integrating imaging and metadata from a large population cardiac imaging study.
  • Disentangling the algorithmic segmentation failures from underlying clinical conditions
  • Addressing the computational challenge of extending deep transformer models to motion data

The framework will be a novel Bayesian approach extending the recent developments in deep recurrent neural networks (e.g. Vision Transformers). These networks provide a natural mechanism to model sequential data such as 2D video. Yet, using Transformers to model the complex dynamics of the heart motion is conceptually new and powerful. The motion will be modelled as the spatiotemporal (3D+t) sequence of the heart shapes across the cardiac cycle, extracted from cine Cardiac Magnetic Resonance (CMR) images. The atlas will be a recurrent model that, given a sequence, will predict a probabilistic distribution function (pdf) for the following heart status. The critical aspect is that the pdf will be conditioned on the patient's metadata (age, gender, ethnicity, etc.). Thus by measuring the spatial deviations from the expected shape at each phase, the atlas will allow very accurate quantification of anatomical and functional cardiac abnormalities (and variances showing uncertainties) specific to the patients.

We have extensive experience developing Bayesian and non-Gaussian statistical atlases from cardiac shapes and motion. However, the previous work discarded the patient metadata (such as age, gender, ethnicity, etc.). Therefore, the atlas was not clinically deployable to study cardiac motion abnormalities, which are relevant to various CVDs. 

The atlas will be derived from the UK Biobank CMR study aiming to scan 100,000 patients by 2022. The training of the atlas will be pursued as the new releases of the data sets from the UK Biobank becomes available. We have established collaboration with this study's clinical advisor and have full access to the CMR data sets. 

Teaching

COMPSCI1020 How to learn a new language? 

 

COMPSCI4015 Professional Software Development 

Professional activities & recognition

Prizes, awards & distinctions

  • 2022: Short-Term, Tokyo Women Medical University (JSPS Invitational Fellowship)
  • 2019: Early Detection Seed Grants (Cancer Research UK)
  • 2020: Vet-AI/University of Leeds (Innovate-UK)

Research fellowships

  • 2021 - 2022: Allen Touring Institue
  • 2014 - 2016: Marie Skłodowska-Curie Individual Fellowship
  • 2008 - 2010: JSPS Fellowship

Editorial boards

  • 2017 - 2019: MICCAI-SASHIMI Workshop