my photo

Research Interests

Machine Learning in Medicine Data Analysis, Computational Healthcare, Internet of Things

I focus on frontier technology of machine learning applicable to medical data analysis and clinical decision making. This involves object detection, segmentation and quantification in medical images, and merging multi-modal (e.g., image, genetic and molecular data), simulation, cross-domain synthesis, life-long learning for automatic diagnosis and treatment planning with imperfect data. I have expertise in deep learning with multi-modal and high-dimensional medical data.

Recently, I widely investigate the problems of interpretability and generalizability of deep learning models that reached a status of Pareto Optimality, and applicability of deep neural networks in the complete clinical decision pipeline.

News

  • [01/2020] Our paper Low-attenuation non-calcified plaque on coronary CT angiography predicts myocardial infarction Results from the multicenter SCOT-HEART trial accepted in Circulation (IF: 23.054)
  • [01/2020] The project A Digital Platform of an AI-aided pipeline for drug development in targeted therapy of cancer, rewarded the fund of Hangzhou 521 Talent programme, project launching in IIAT, Hangzhou, China
  • [01/2020] Launching a project about B-cell lymphoma collaborating with Shuren University and Shulan Hospital in March, Hangzhou

Selected Publications and Patents

Sorry that for works developed in Toshiba, and work directly applied to real on-going clinical trials, code and dataset can not be released due to confidential or consent agreements. Some papers/results can be available on requests. Other research output are available here.

Recently Accepted/Major Revision/Submitted

Michelle C William et al., Low-attenuation non-calcified plaque on coronary CT angiography predicts myocardial infarction, results from the Multicenter SCOT-HEART trial. Circulation. (Accepted)

C Wang et al., FIRE: unsupervised bi-directional inter-modality registration using deep networks. CMIG. (Minor Revision)

1 paper about zero-shot image segmentation. IEEE TMI. (Major Revision)

1 paper about brain aging synthesis. IEEE TMI. (Submitted)

Published

Chengjia Wang, Giorgos Papanastasiou, Sotirios Tsaftaris, Guang Yang, Calum Gray, David Newby, Gillian Macnaught, Tom MacGillivray, TPSDicyc: Improved Deformation Invariant Cross-domain Medical Image Synthesis, MICCAI-MLMIR 2019.

Chengjia Wang, Shizhou Dong, Xiaofeng Zhao, Giorgos Papanastasiou, Heye Zhang, Guang Yang, SaliencyGAN: Deep Learning Semi-supervised Salient Object Detection in the Fog of IoT, IEEE TII, 2019

Xiahai Zhuang, Lei Li, Christian Payer, Darko Stern, Martin Urschler, Mattias P Heinrich, Julien Oster, Chunliang Wang, Orjan Smedby, Cheng Bian, Xin Yang, Pheng-Ann Heng, Aliasghar Mortazi, Ulas Bagci, Guanyu Yang. Evaluation of Algorithms for Multi-Modality Whole Heart Segmentation: An Open-Access Grand Challenge, Medical Image Analysis 2019.

Chengjia Wang, Keith A Goatman, James Boardman, Erin Beveridge, David Newby, Scott Semple. Distance Oriented Particle Swarm Optimizer for Brain Image Registration. IEEE Access, 2019. [paper]

Giorgos Papanastasiou, Mark Rodrigues, Chengjia Wang, Kerstin Heurling, Rustam Salman, Gillian Macnaught. Pharmacokinetic analysis of 18F-flutemetamol uptake in cerebral amyloid angiopathy using PET-MR imaging. PSMR 8th conference on PET/MR and SPECT/MR, 2019

Giorgos Papanastasiou, Mark Rodrigues, Chengjia Wang, Kerstin Heurling, Rustam Salman, Gillian Macnaught. Quantitative assessment of 18F-flutemetamol uptake in cerebral amyloid angiopathy using a reduced PET-MR acquisition time frame. (oral) PSMR 8th conference on PET/MR and SPECT/MR, 2019

Chengjia Wang, Tom MacGillivray, Gillian Macnaught, Guang Yang, David Newby. A Two-Stage U-Net Model for 3D Multi-class Segmentation on Full-Resolution Cardiac Data. Statistical Atlases and Computational Models of the Heart (STACOM) workshop of MICCAI. 2018

Chengjia Wang, Gillian Macnaught, Giorgos Papanastasiou, Tom MacGillivray, David Newby. Unsupervised learning for cross-domain medical image synthesis using deformation invariant cycle consistency networks. (oral) Simulation and Synthesis in Medical Imaging (SASHIMI) workshop of MICCAI. 2018

Martin Lyngby Lassen, Jacek Kwiecinski, Sebastien Cadet, Damini Dey, Chengjia Wang, Marc R Dweck, Daniel Berman, Guido Germano, David E Newby, Piotr J Slomka, Data-driven gross patient motion detection and compensation: Implications for coronary 18F-NaF PET imaging. Journal of Nuclear Medicine, 2018

Noel Conlisk, Rachael O Forsythe, Lyam Hollis, Barry J Doyle, Olivia MB McBride, Jennifer MJ Robson, Chengjia Wang, Calum D Gray, Scott IK Semple, Tom MacGillivray, Edwin JR van Beek, David E Newby, Peter R Hoskins. Exploring the biological and mechanical properties of abdominal aortic aneurysms using USPIO MRI and peak tissue stress: a combined clinical and finite element study. Journal of cardiovascular translational research. 2017

MA3RS Study Investigators, Aortic wall inflammation predicts abdominal aortic aneurysm expansion, rupture, and need for surgical repair, Circulation, 2017

Chengjia Wang, Keith Goatman, Scott Semple, Method of, and apparatus for, registration of medical images, US Patent, 2016

Chengjia Wang, Keith Goatman, Tom MacGillivray, Erin Beveridge, Yolanda Koutraki, James P Boardman, Colin Stirrat, Sarah Sparrow, Emma Moore, Rozalia Pataky, Shah Alam, Marc Dweck, Calvin Chin, Calum Gray, David Newby, Scott Semple. Automatic multi-parametric MR registration method using mutual information based on adaptive asymmetric k-means binning. IEEE ISBI, 2015

Chengjia Wang, Colin Stirrat, Shirjel Alam, Marc Dweck, Calvin Chin, Thomas MacGillivray, Calum Gray, David Newby, Scott Semple. Robust registration software of multi-parametric cardiac MR data with presence of motion-related artefacts and intensity non-homogeneity. ESMRMB. 2015

Chengjia Wang, Tom MacGillivray , Yolanda Koutraki , James P Boardman , Sarah Sparrow , Rozalia Pataky , and Scott Semple. A robust automated multi-parametric registration software applied to neonatal MR neuro data, ESMRMB. 2015.

Yolanda Koutraki, Olivia Mcbride, J. Robson, Racheal Forsythe, Chengjia Wang, Tom Macgillivray, Calum Gray, Keith Goatman, J. Camilleri-Brennan3, J. Jegadeeson, David Newby, Scott Semple. Automatic classification of abdominal aortic aneurysms to identify patients at risk of aneurysm expansion and rupture. ESMRMB. 2015.

Yolanda Koutraki, Chengjia Wang, Jennifer Robson, Olivia Mcbride, Rachael Forsythe, Tom MacGillivray, Calum Gray, Keith Goatman, J. Camilleri-Brennan, David E Newby, and Scott I Semple. 3D visualization of ‘Hotspots’ of inflammation in abdominal aortic aneurysms (AAA) ISMRM. 2015.

Scott Semple, Shirjel R Alam, Tom J MacGillivray, Marc R Dweck, Anoop S Shah, Jenny Richards, Chengjia Wang, Ninian Lang, Graham McKillop, Saeed Mirsadraee, Renzo Pessotto, Vipin Zamvar, Peter Henriksen, David Newby. Quantitative myocardial inflammation assessed using a novel USPIO-Magnetic Resonance Imaging Acquisition and Analysis Protocol, Journal of cardiovascular magnetic resonance: official journal of the Society for Cardiovascular Magnetic Resonance. 2013.

Fundings & Awards

– Contributions in funding applications and projects

  • Hangzhou 521 Talent, “A Digital Platform of an AI-aided pipeline for drug development in targeted therapy of cancer “ Hangzhou City Council, £110,000-220,000
  • BHF Programme Grant no. RG/16/10/32375 ,“Non-invasive Imaging of Human Coronary Atherothrombosis”
  • Welcome trust, NHS “Prediction of Recurrent Events With 18F-Fluoride” (PREFFIR)
  • NHS “Scottish COmputed Tomography of the HEART Trial (SCOT-HEART)”
  • INSPIRE-EDINBURGH-2

– Other Awards

  • AWS Cloud Credits Award for Research 2019
  • Erasmus+ Travel Award 2017, 2019
  • Runner-up for the Best Project Award of The Hamlyn Winter School on Surgical Imaging and Vision (two winners per year)
  • Microsoft Microsoft Azure Sponsorship 2017 (CRM:0518003)
  • Amazon AWS Research Grant 2017
  • Magna Cum Laude Merit award ISMRM 2015
  • Certificate of Merit Award ESMRMB 2015
  • Centre 4 Cardiovascular Science Symposium–EBQ Award, 2014
  • SUPA INSPIRE Prize

(Full list available on CV)

Talks

  • “AI aided targeted therapy for cancers based medical imaging and genetic data analysis”
    at Imperial Institute of Advanced Technology (IIAT), Hangzhou 11/2019

  • “AI in Medical Imaging: Brain, Heart, and more”
    at China-Britain Artificial Intelligence Association, 11/2018
    at Imperial Institute of Advanced Technology (IIAT), Hangzhou 10/2019

  • “Deep learning applied to multi-modality cardiovascular images: towards fully automatic analysis and interpretation”
    at Institute of Digital Communications, University of Edinburgh, 08/2018
    at Cedars Sinai Medical Centre, UCLA, 03/2018

  • “Face recognition for special imaging devices: multi-channel and multi-spectral” (confidential)
    at Oxford Multi Spectral Ltd, 12/2017
    at 3D Engine Ltd, Oxford, 12/2017

  • “AUTOBrain.ai: Imaging for Smart Security and HealthCare” [Short Version Slides]
    at RTCInnovation Ltd, Birmingham, 11/2017

  • “Deep learning for automatic medical image analysis: multi-organ segmentation, lesion detection, and pathology detection”
    at BHF Centre for Cardiovascular Science, University of Edinburgh, 07/2017
    at Qingdao University of Science and Technology, Qingdao, 12/2016

  • “A “Registration-oriented” Particle Swarm Optimizer” (Confidential)
    at Toshiba Medical Visualization System - Europe, 03/2016
    at Clinical Research Imaging Centre, 12/2015

  • “Commercialisation of a multi-modality medical image analysis software”
    at Edinburgh Research and Innovation, 10/2015

  • CPR talk: “Towards Fully Automatic Multi-modal and Multi-parametric Medical Image Analysis”
    at Queens Medical Research Institute (QMRI), University of Edinburgh, 07/2014

  • “Introduction of Deep Neural Networks, A Comparison with Manually defined features” (Confidential)
    at Toshiba Medical Visualization System - Europe, 12/2013

  • “Commercialization of Medical Image Analysis Software” [slides]
    at SUPA INSPIRE annual meeting, 07/2013

  • “Automated Robust Multi-modality Registration Tool” [slides]
    at Toshiba Medical Visualization System - Europe, 09/2013
    at Qingdao University of Science and Technology, 08/2013
    at Clinical Research Imaging Centre, University of Edinburgh, 06/2013

  • “Machine Learning based Analysis Techniques for Coronary Arteries and Cardiac Functions using Advanced MRI and CT Imaging”[slides]
    at Queen’s Medical Research Institute, University of Edinburgh, 03/2013 at Toshiba Medical Visualization System - Europe, 07/2012
    at SUPA INSPIRE Annual Meeting, Heriot-watt University, 06/2012

Professional Activities

Memberships

IEEE, IEEE EMBS, MICCAI Society, ARUK, ISMRM, ESMRMB

Conference & Journal Reviews

IEEE Transactions on Evolutionary Computing (TEVC)
IEEE Transactions on Medical Imaging (TMI)
IEEE Transactions on on Neural Networks and Learning System (TNNLS)
IEEE Access
Medical Image Analysis (MedIA)
Computers in Biology and Medicine
Future Generation Computer Systems (FGCS)
IEEE Transactions on Biomedical Engineering (TBME)
IEEE Transactions on Image Processing (TIP)
IEEE Transactions on Cybernetics (CYB)
IEEE Journal of Biomedical and Health Informatics (JBHI)
Circulation
NeuroImage
Neurocomputing
Pattern Recognition
Medical Physics
Computerized Medical Imaging and Graphics
Neuroradiology
Journal of Magnetic Resonance Imaging
SPIE Journal of Medical Imaging

MICCAI, ISBI, ISMRM, ESMRMB

Selected Established collaborations

Internal collaborations:

  • Usher Institute, Edinburgh Medical School
  • IDCOM, School of Engineering

External collaborations:

  • UCLA Cedars Sinai Medical Centre and University of Manchester (Multi-centre Study)
  • Canon Medical Systems Europe, Siemens (Cross-modality image registration, Medical data management)
  • Imperial Institute of Advanced Technology (IIAT)
  • Medicine, National Heart & Lung Institute, Imperial College London (Joint project)
  • School of Biomedical Engineering, Sun Yat-Sen University (Multi-centre study)

Details of specific collaborators and more collaborations are available on request.

Others: