Two-Dimensional Multi-Target Detection: An Autocorrelation Analysis Approach

We consider the two-dimensional multi-target detection problem of recovering a target image from a noisy measurement that contains multiple copies of the image, each randomly rotated and translated. Motivated by the structure reconstruction problem in single-particle cryo-electron microscopy, we focus on the high noise regime, where the noise hampers accurate detection of the image occurrences. We develop an autocorrelation analysis framework to estimate the image directly from a measurement with an arbitrary spacing distribution of image occurrences, bypassing the estimation of individual locations and rotations. We conduct extensive numerical experiments, and demonstrate image recovery in highly noisy environments. The code to reproduce all numerical experiments is publicly available at https://github.com/krshay/MTD-2D.

[1]  Aggelos K. Katsaggelos,et al.  Using Deep Neural Networks for Inverse Problems in Imaging: Beyond Analytical Methods , 2018, IEEE Signal Processing Magazine.

[2]  Afonso S. Bandeira,et al.  Optimal rates of estimation for multi-reference alignment , 2017, Mathematical Statistics and Learning.

[3]  W. Kühlbrandt,et al.  Current limitations to high-resolution structure determination by single-particle cryoEM , 2021, Quarterly Reviews of Biophysics.

[4]  A. KLUG,et al.  Three-dimensional Image Reconstruction from the Viewpoint of information Theory , 1972, Nature.

[5]  Amit Singer,et al.  Estimation in the Group Action Channel , 2018, 2018 IEEE International Symposium on Information Theory (ISIT).

[6]  Anat Levin,et al.  Blind Motion Deblurring Using Image Statistics , 2006, NIPS.

[7]  Amit Singer,et al.  Single-Particle Cryo-Electron Microscopy: Mathematical Theory, Computational Challenges, and Opportunities , 2019, IEEE Signal Processing Magazine.

[8]  Timothy J. Schulz,et al.  Multiframe blind deconvolution of astronomical images , 1993 .

[9]  Amit Singer,et al.  Multi-target detection with application to cryo-electron microscopy , 2019, Inverse Problems.

[10]  Zhizhen Zhao,et al.  Fourier-Bessel rotational invariant eigenimages , 2012, Journal of the Optical Society of America. A, Optics, image science, and vision.

[11]  R. Henderson.  The potential and limitations of neutrons, electrons and X-rays for atomic resolution microscopy of unstained biological molecules , 1995, Quarterly Reviews of Biophysics.

[12]  Afonso S. Bandeira,et al.  Estimation under group actions: recovering orbits from invariants , 2017, Applied and Computational Harmonic Analysis.

[13]  Amit Singer,et al.  Image Recovery from Rotational And Translational Invariants , 2020, ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[14]  Amit Singer,et al.  Multi-Target Detection With an Arbitrary Spacing Distribution , 2019, IEEE Transactions on Signal Processing.

[15]  Amit Singer,et al.  Super-resolution multi-reference alignment , 2020, Information and inference : a journal of the IMA.

[16]  Ellen D. Zhong,et al.  CryoDRGN: Reconstruction of heterogeneous cryo-EM structures using neural networks , 2021, Nature Methods.

[17]  Karim Abed-Meraim,et al.  Blind system identification , 1997, Proc. IEEE.

[18]  Zhizhen Zhao,et al.  Fast Steerable Principal Component Analysis , 2014, IEEE Transactions on Computational Imaging.

[19]  David M. Mount,et al.  Analysis of approximate nearest neighbor searching with clustered point sets , 1999, Data Structures, Near Neighbor Searches, and Methodology.

[20]  Amit Singer,et al.  Multi-target detection with rotations , 2021, ArXiv.

[21]  A. Singer,et al.  Reducing bias and variance for CTF estimation in single particle cryo-EM. , 2020, Ultramicroscopy.

[22]  Sjors H.W. Scheres,et al.  RELION: Implementation of a Bayesian approach to cryo-EM structure determination , 2012, Journal of structural biology.

[23]  A. Klug,et al.  Measurement and compensation of defocusing and aberrations by Fourier processing of electron micrographs , 1971 .

[24]  Amit Singer,et al.  Toward single particle reconstruction without particle picking: Breaking the detection limit , 2018, bioRxiv.

[25]  Raj Rao Nadakuditi,et al.  Passive Radar Detection With Noisy Reference Channel Using Principal Subspace Similarity , 2018, IEEE Transactions on Aerospace and Electronic Systems.

[26]  S. Scheres,et al.  How cryo-EM is revolutionizing structural biology. , 2015, Trends in biochemical sciences.

[27]  David J. Fleet,et al.  cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination , 2017, Nature Methods.

[28]  Michael Unser,et al.  CryoGAN: A New Reconstruction Paradigm for Single-Particle Cryo-EM via Deep Adversarial Learning , 2020, bioRxiv.

[29]  E. Nogales.  The development of cryo-EM into a mainstream structural biology technique , 2015, Nature Methods.

[30]  Tamir Bendory,et al.  Multi-reference alignment in high dimensions: sample complexity and phase transition , 2020, SIAM J. Math. Data Sci..

[31]  Guillermo Sapiro,et al.  Fundamental Limits in Multi-Image Alignment , 2016, IEEE Transactions on Signal Processing.

[32]  J. Frank.  Three-Dimensional Electron Microscopy of Macromolecular Assemblies: Visualization of Biological Molecules in Their Native State , 1996 .