Honors & Awards

  • Early Career Fellowship, The European Society for Magnetic Resonance in Medicine and Biology (ESMRMB) (2020)

Boards, Advisory Committees, Professional Organizations

  • Member of The Membership, Marketing, and Media Committee, The European Society for Magnetic Resonance in Medicine and Biology (ESMRMB) (2020 - Present)
  • Member of The Early Career Investigators Committee, The European Society for Magnetic Resonance in Medicine and Biology (ESMRMB) (2019 - Present)
  • Member of EU COST Action Glioma MR Imaging 2.0, European Cooperation in Science and Technology (COST) (2019 - Present)

Professional Education

  • Bachelor of Science, Hong Kong Baptist University - Ho Sin Hang Campus (2012)
  • Doctor of Philosophy, University of Oxford, Engineering Science (Biomedical Engineering) (2019)
  • Master of Science, The Chinese University of Hong Kong, Biomedical Engineering (2013)
  • Bachelor of Science, Hong Kong Baptist University, Computer Science (2012)

Stanford Advisors

All Publications

  • Predicting PET Cerebrovascular Reserve with Deep Learning by Using Baseline MRI: A Pilot Investigation of a Drug-Free Brain Stress Test. Radiology Chen, D. Y., Ishii, Y., Fan, A. P., Guo, J., Zhao, M. Y., Steinberg, G. K., Zaharchuk, G. 2020: 192793


    Background Cerebrovascular reserve (CVR) may be measured by using an acetazolamide test to clinically evaluate patients with cerebrovascular disease. However, acetazolamide use may be contraindicated and/or undesirable in certain clinical settings. Purpose To predict CVR images generated from acetazolamide vasodilation with a deep learning network by using only images before acetazolamide administration. Materials and Methods Simultaneous oxygen 15 (15O)-labeled water PET/MRI before and after acetazolamide injection were retrospectively analyzed for patients with Moyamoya disease and healthy control participants from April 2017 to May 2019. Inputs to deep learning models were perfusion-based images (arterial spin labeling [ASL]), structural scans (T2 fluid-attenuated inversion-recovery, T1), and brain location. Two models, that is, 15O-labeled water PET cerebral blood flow (CBF) and MRI (PET-plus-MRI model) before acetazolamide administration and only MRI (MRI-only model) before acetazolamide administration, were trained and tested with sixfold cross-validation. The models learned to predict a voxelwise relative CBF change (rDeltaCBF) map by using rDeltaCBF measured with PET due to acetazolamide as ground truth. Quantitative analysis included image quality metrics (peak signal-to-noise ratio, root mean square error, and structural similarity index), as well as comparison between the various methods by using correlation and Bland-Altman analyses. Identification of vascular territories with impaired rDeltaCBF was evaluated by using receiver operating characteristic metrics. Results Thirty-six participants were included: 24 patients with Moyamoya disease (mean age ± standard deviation, 41 years ± 12; 17 women) and 12 age-matched healthy control participants (mean age, 39 years ± 16; nine women). The rDeltaCBF maps predicted by both deep learning models demonstrated better image quality metrics than did ASL (all P < .001 in patients) and higher correlation coefficient with PET than with ASL (PET-plus-MRI model, 0.704; MRI-only model, 0.690 vs ASL, 0.432; both P < .001 in patients). Both models also achieved high diagnostic performance in identifying territories with impaired rDeltaCBF (area under receiver operating characteristic curve, 0.95 for PET-plus-MRI model [95% confidence interval: 0.90, 0.99] and 0.95 for MRI-only model [95% confidence interval: 0.91, 0.98]). Conclusion By using only images before acetazolamide administration, PET-plus-MRI and MRI-only deep learning models predicted cerebrovascular reserve images without the need for vasodilator injection. © RSNA, 2020 Online supplemental material is available for this article.

    View details for DOI 10.1148/radiol.2020192793

    View details for PubMedID 32662761

  • Quantification of cerebral perfusion and cerebrovascular reserve using Turbo-QUASAR arterial spin labeling MRI. Magnetic resonance in medicine Zhao, M. Y., Václavů, L., Petersen, E. T., Biemond, B. J., Sokolska, M. J., Suzuki, Y., Thomas, D. L., Nederveen, A. J., Chappell, M. A. 2019


    To compare cerebral blood flow (CBF) and cerebrovascular reserve (CVR) quantification from Turbo-QUASAR (quantitative signal targeting with alternating radiofrequency labeling of arterial regions) arterial spin labeling (ASL) and single post-labeling delay pseudo-continuous ASL (PCASL).A model-based method was developed to quantify CBF and arterial transit time (ATT) from Turbo-QUASAR, including a correction for magnetization transfer effects caused by the repeated labeling pulses. Simulations were performed to assess the accuracy of the model-based method. Data from an in vivo experiment conducted on a healthy cohort were retrospectively analyzed to compare the CBF and CVR (induced by acetazolamide) measurement from Turbo-QUASAR and PCASL on the basis of global and regional differences. The quality of the two ASL data sets was examined using the coefficient of variation (CoV).The model-based method for Turbo-QUASAR was accurate for CBF estimation (relative error was 8% for signal-to-noise ratio = 5) in simulations if the bolus duration was known. In the in vivo experiment, the mean global CVR estimated by Turbo-QUASAR and PCASL was between 63% and 64% and not significantly different. Although global CBF values of the two ASL techniques were not significantly different, regional CBF differences were found in deep gray matter in both pre- and postacetazolamide conditions. The CoV of Turbo-QUASAR data was significantly higher than PCASL.Both ASL techniques were effective for quantifying CBF and CVR, despite the regional differences observed. Although CBF estimated from Turbo-QUASAR demonstrated a higher variability than PCASL, Turbo-QUASAR offers the advantage of being able to measure and control for variation in ATT.

    View details for DOI 10.1002/mrm.27956

    View details for PubMedID 31513311

  • A systematic study of the sensitivity of partial volume correction methods for the quantification of perfusion from pseudo-continuous arterial spin labeling MRI NEUROIMAGE Zhao, M. Y., Mezue, M., Segerdahl, A. R., Okell, T. W., Tracey, I., Xiao, Y., Chappell, M. A. 2017; 162: 384–97


    Arterial spin labeling (ASL) MRI is a non-invasive technique for the quantification of cerebral perfusion, and pseudo-continuous arterial spin labeling (PCASL) has been recommended as the standard implementation by a recent consensus of the community. Due to the low spatial resolution of ASL images, perfusion quantification is biased by partial volume effects. Consequently, several partial volume correction (PVEc) methods have been developed to reduce the bias in gray matter (GM) perfusion quantification. The efficacy of these methods relies on both the quality of the ASL data and the accuracy of partial volume estimates. Here we systematically investigate the sensitivity of different PVEc methods to variability in both the ASL data and partial volume estimates using simulated PCASL data and in vivo PCASL data from a reproducibility study. We examined the PVEc methods in two ways: the ability to preserve spatial details and the accuracy of GM perfusion estimation. Judging by the root-mean-square error (RMSE) between simulated and estimated GM CBF, the spatially regularized method was superior in preserving spatial details compared to the linear regression method (RMSE of 1.2 vs 5.1 in simulation of GM CBF with short scale spatial variations). The linear regression method was generally less sensitive than the spatially regularized method to noise in data and errors in the partial volume estimates (RMSE 6.3 vs 23.4 for SNR = 5 simulated data), but this could be attributed to the greater smoothing introduced by the method. Analysis of a healthy cohort dataset indicates that PVEc, using either method, improves the repeatability of perfusion quantification (within-subject coefficient of variation reduced by 5% after PVEc).

    View details for DOI 10.1016/j.neuroimage.2017.08.072

    View details for Web of Science ID 000416502800034

    View details for PubMedID 28887087

  • Disease-Specific Target Gene Expression Profiling of Molecular Imaging Probes: Database Development and Clinical Validation MOLECULAR IMAGING Chan, L., Ngo, C., Wang, F., Zhao, M. Y., Zhao, M., Law, H., Wong, S., Yung, B. 2014; 13


    Molecular imaging probes can target abnormal gene expression patterns in patients and allow early diagnosis of disease. For selecting a suitable imaging probe, the current Molecular Imaging and Contrast Agent Database (MICAD) provides descriptive and qualitative information on imaging probe characteristics and properties. However, MICAD does not support linkage with the expression profiles of target genes. The proposed Disease-specific Imaging Probe Profiling (DIPP) database quantitatively archives and presents the gene expression profiles of targets across different diseases, anatomic regions, and subcellular locations, providing an objective reference for selecting imaging probes. The DIPP database was validated with a clinical positron emission tomography (PET) study on lung cancer and an in vitro study on neuroendocrine cancer. The retrieved records show that choline kinase beta and glucose transporters were positively and significantly associated with lung cancer among the targets of 11C-choline and [18F]fluoro-2-deoxy-2-d-glucose (FDG), respectively. Their significant overexpressions corresponded to the findings that the uptake rate of FDG increased with tumor size but that of 11C-choline remained constant. Validated with the in vitro study, the expression profiles of disease-associated targets can indicate the eligibility of patients for clinical trials of the treatment probe. A Web search tool of the DIPP database is available at http://www.polyu.edu.hk/bmi/dipp/.

    View details for DOI 10.2310/7290.2014.00017

    View details for Web of Science ID 000344215700002

    View details for PubMedID 25022454