Research Interests

My major study area is signal & image processing and machine learning. My research interests include medical imaging, image reconstruction, computational imaging, inverse problems, optimization, machine learning, and deep learning.

Journal Publications

• Deep Learning Denoising of Digital Breast Tomosynthesis: Observer Study of The Effect on Microcalcification Detection in Breast Phantom Images, H.-P. Chan, M. A. Helvie, M. Gao, L. M. Hadjiyski, C. Zhou, K. Garver, K. A. Klein, C. McLaughlin, R. Oudsema, W. T. Rahman, and M. A. Roubidoux, Medical Physics, March 2023. Accepted.

• Deep Convolutional Neural Network With Adversarial Training for Denoising Digital Breast Tomosynthesis Images, M. Gao, J. A. Fessler, and H.-P. Chan, IEEE Transactions on Medical Imaging, Vol. 40, No. 7, pp. 1805-1816, July 2021. Paper Supplement

Preprints

• Fast, Precise Myelin Water Quantification Using DESS MRI and Kernel Learning, G. Nataraj, J.-F. Nielsen, M. Gao, and J. A. Fessler, 2018. Paper

Conference Proceedings and Abstracts

• Deep CNN Task-based Image Quality Assessment: Application to Digital Breast Tomosynthesis Reconstruction and Denoising, M. Gao, M. A. Helvie, R. K. Samala, L. M. Hadjiyski, J. A. Fessler, and H.-P. Chan, in Proceedings of SPIE, 2023. Accepted.

• Deep Learning Denoising and Assessment of Detectability of Microcalcifications in Digital Breast Tomosynthesis: A Task-based Image Evaluation Approach Using CNN, M. Gao, M. A. Helvie, R. K. Samala, J. A. Fessler, and H.-P. Chan, in RSNA Annual Meeting, Chicago, 2022.

• Deep Convolutional Neural Network Regularized Digital Breast Tomosynthesis Reconstruction With Detector Blur and Correlated Noise Modeling, M. Gao, J. A. Fessler, and H.-P. Chan, in Proceedings of SPIE, 12031, 1203108, 2022. Paper

• Plug-and-play Reconstruction With Deep Learning Denoising for Improving Detectability of Microcalcifications in Digital Breast Tomosynthesis Images, M. Gao, J. A. Fessler, and H.-P. Chan, in RSNA Annual Meeting, Chicago, 2021.

• Digital Breast Tomosynthesis Denoising Using Deep Convolutional Neural Network: Effects of Dose Level of Training Target Images, M. Gao, J. A. Fessler, and H.-P. Chan, in Proceedings of SPIE, 11595, 115951K, 2021. Paper

• Training Deep Convolutional Neural Network With In Silico Data for Denoising Digital Breast Tomosynthesis Images, M. Gao, J. A. Fessler, and H.-P. Chan, in RSNA Annual Meeting, virtual, 2020.

• Deep Convolutional Neural Network Denoising for Digital Breast Tomosynthesis Reconstruction, M. Gao, R. K. Samala, J. A. Fessler, and H.-P. Chan, in Proceedings of SPIE, 11312, 113120Q, 2020. Paper

• Myelin Water Fraction Estimation Using Small-tip Fast Recovery MRI, S. T. Whitaker, G. Nataraj, M. Gao, J.-F. Nielsen, and J. A. Fessler, in ISMRM Annual Conference, Montréal, 2019.

• Kernel Regression for Fast Myelin Water Imaging, G. Nataraj, M. Gao, J.-F. Nielsen, and J. A. Fessler, in ISMRM Workshop on Machine Learning Part II, Washington D.C., 2018 (2nd-place poster award). Paper

• Shallow Learning With Kernels for Dictionary-free Magnetic Resonance Fingerprinting, G. Nataraj, M. Gao, J. Assländer, C. Scott, and J. A. Fessler, in ISMRM Workshop on MR Fingerprinting, Cleveland, 2017. Paper

Working Experience

I was a summer intern at Apple Inc. working with Farhan Baqai and Hao Sun on low-light image denoising in the Camera Algorithms team in 2022.