Quantitative Bioimaging An Introduction to Biology,

Instrumentation, Experiments, and Data Analysis for

Scientists and Engineers, 1st Edition

Visit the link below to download the full version of this book:

https://medipdf.com/product/quantitative-bioimaging-an-introduction-to-biology-i
nstrumentation-experiments-and-data-analysis-for-scientists-and-engineers-1st-ed
ition/

Click Download Now

First edition published 2020
by CRC Press
6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742

and by CRC Press

2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

© 2021 Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, LLC

Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have
attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders
if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please
write and let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized
in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying,
microfilming, and recording, or in any information storage or retrieval system, without written permission from the
publishers.

For permission to photocopy or use material electronically from this work, access www.copyright.com or contact the
Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. For works that are
not available on CCC please contact [email protected]

Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for
identification and explanation without intent to infringe.

ISBN: 9781138598980 (hbk)

ISBN: 9780429469893 (ebk)

Typeset in Computer Modern font

by KnowledgeWorks Global Ltd.

Visit the eResources: www.routledge.com/9781138598980

To our families.
Contents

Preface xvii

Acknowledgments xix

I Introduction 1
Overview 3

1 Then and Now 5

2 Introduction to Two Problems in Cellular Biology 7

2.1 Antibody trafficking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Localization experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.3 Association experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 Dynamic studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.5 Iron transport, transferrin, and the transferrin receptor . . . . . . . . . 13

3 Basics of Microscopy Techniques 17

3.1 Optical microscopy for cell biology . . . . . . . . . . . . . . . . . . . . . 17
3.2 Transmitted light microscopy . . . . . . . . . . . . . . . . . . . . . . . . 17
3.3 Fluorescence microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3.1 Fluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.3.2 Layout of an epifluorescence widefield microscope . . . . . . . . 21
3.4 Inverted versus upright microscope . . . . . . . . . . . . . . . . . . . . . 22
3.5 Components of commercial microscopes . . . . . . . . . . . . . . . . . . 23
3.5.1 Light sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.5.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.6 Fixed and live cell experiments . . . . . . . . . . . . . . . . . . . . . . . 27
3.7 Sample preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.8 A note regarding safety . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

4 Introduction to Image Formation and Analysis 29

4.1 Image formation and point spread functions . . . . . . . . . . . . . . . . 29
4.2 Resolution: an elementary introduction . . . . . . . . . . . . . . . . . . . 31
4.3 Modeling and analyzing the data . . . . . . . . . . . . . . . . . . . . . . 34

Notes 37

Exercises 39

II Biology and Chemistry 41

Overview 43

vii
viii Contents

5 From genes to proteins 45

5.1 Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.2 DNA and genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.3 How are proteins made? . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5.4 Structures of proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.5 Protein structure determination . . . . . . . . . . . . . . . . . . . . . . . 58

6 Antibodies 65
6.1 Structure of antibodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
6.2 Variable regions and binding activity . . . . . . . . . . . . . . . . . . . . 67
6.3 Constant regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.4 Antibody production for laboratory and clinical use . . . . . . . . . . . 71
6.4.1 The classical method: hybridoma technology . . . . . . . . . . . 71
6.5 Diagnostic techniques using antibody detection methods . . . . . . . . . 73
6.5.1 Enzyme-linked immunosorbent assay . . . . . . . . . . . . . . . 73
6.5.2 Surface plasmon resonance for the quantitation of the affinity of an
interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

7 Cloning of genes for protein expression 77

7.1 Features of expression constructs . . . . . . . . . . . . . . . . . . . . . . 77
7.2 Methods for generating expression plasmids . . . . . . . . . . . . . . . . 78
7.2.1 Restriction enzymes . . . . . . . . . . . . . . . . . . . . . . . . . 78
7.2.2 Polymerase chain reaction . . . . . . . . . . . . . . . . . . . . . . 79
7.2.3 Details of approaches for generating expression plasmids . . . . 79
7.2.4 Transfection of mammalian cells for expression . . . . . . . . . . 86
7.3 Antibody engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
7.3.1 Chimeric antibodies . . . . . . . . . . . . . . . . . . . . . . . . . 87
7.3.2 Humanized antibodies . . . . . . . . . . . . . . . . . . . . . . . . 88
7.3.3 Isolation of V regions . . . . . . . . . . . . . . . . . . . . . . . . 88

8 Principles of Fluorescence 91
8.1 Wave and particle description of light . . . . . . . . . . . . . . . . . . . 91
8.2 Jablonski diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
8.3 Stokes shift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
8.4 Photobleaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
8.5 Photophysical characterization of fluorophores . . . . . . . . . . . . . . 94
8.5.1 Quantum yield . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
8.5.2 Beer-Lambert law, effective absorption cross section and molar ex-
tinction coefficient . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.5.3 Brightness of a fluorophore . . . . . . . . . . . . . . . . . . . . . 96
8.6 Excitation and emission spectra . . . . . . . . . . . . . . . . . . . . . . . 96
8.7 Fluorophores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
8.7.1 Chemical fluorescent dyes . . . . . . . . . . . . . . . . . . . . . . 97
8.7.1.1 Labeling of proteins via cysteine or lysine residues . . . 98
8.7.1.2 Labeling of proteins with fluorophore-conjugated strepta-
vidin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
8.7.1.3 In situ labeling of proteins in cells using peptide tags . 100
8.7.2 Quantum dots . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
8.7.2.1 Labeling of proteins with quantum dots . . . . . . . . . 101
8.7.3 Fluorescent proteins . . . . . . . . . . . . . . . . . . . . . . . . . 102
8.7.4 Photoactivatable and photoswitchable fluorescent probes . . . . 105
Contents ix

8.7.5 Other labeling modalities . . . . . . . . . . . . . . . . . . . . . . 105

9 Cells 107
9.1 Cellular structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
9.2 Receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
9.3 Typical biological systems . . . . . . . . . . . . . . . . . . . . . . . . . . 112
9.3.1 Subcellular trafficking of the Fc receptor, FcRn . . . . . . . . . . 112
9.3.2 Subcellular trafficking of the transferrin receptor . . . . . . . . . 112
9.4 Sample preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
9.4.1 Labeling of proteins in fixed cells . . . . . . . . . . . . . . . . . . 113
9.4.2 Sample preparation for typical fixed cell experiments . . . . . . 114
9.4.3 Sample preparation for typical live cell imaging experiments . . 115

Notes 121

Exercises 125

III Optics and Microscopy 127

Overview 129

10 Microscope Designs 131

10.1 Light path for widefield fluorescence microscopy . . . . . . . . . . . . . 131
10.1.1 Infinity-corrected light path . . . . . . . . . . . . . . . . . . . . . 131
10.2 Imaging in three dimensions . . . . . . . . . . . . . . . . . . . . . . . . . 132
10.2.1 Focus control and acquisition of z-stacks . . . . . . . . . . . . . 132
10.2.2 Multifocal plane microscopy . . . . . . . . . . . . . . . . . . . . 133
10.3 Imaging of multiple colors . . . . . . . . . . . . . . . . . . . . . . . . . . 133
10.4 Light path for confocal microscopy . . . . . . . . . . . . . . . . . . . . . 135
10.5 Two-photon excitation microscopy . . . . . . . . . . . . . . . . . . . . . 137
10.6 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
10.6.1 Numerical aperture and immersion medium . . . . . . . . . . . . 140
10.6.2 Corrections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
10.6.3 Transmission efficiency . . . . . . . . . . . . . . . . . . . . . . . 141
10.7 Optical filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
10.7.1 Example: a filter set for a GFP-labeled protein . . . . . . . . . . 145
10.7.2 Imaging of multiple fluorophores . . . . . . . . . . . . . . . . . . 147
10.8 Transmitted light microscopy . . . . . . . . . . . . . . . . . . . . . . . . 151

11 Microscopy Experiments 155

11.1 Fixed cell experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
11.1.1 Localization of FcRn . . . . . . . . . . . . . . . . . . . . . . . . 155
11.1.2 Association experiments with FcRn, EEA1, LAMP1, and transfer-
rin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
11.1.3 Pulse-chase verification of fate of mutated IgG . . . . . . . . . . 158
11.2 Imaging a 3D sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
11.2.1 Acquisition of z-stacks . . . . . . . . . . . . . . . . . . . . . . . . 160
11.2.2 Out-of-focus haze . . . . . . . . . . . . . . . . . . . . . . . . . . 160
11.3 Live cell experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
11.3.1 Example: FcRn-mediated IgG trafficking . . . . . . . . . . . . . 163
11.4 Total internal reflection fluorescence microscopy (TIRFM) . . . . . . . . 164
11.4.1 Objective-based total internal reflection fluorescence microscopy 164
x Contents

11.4.2 Exocytosis imaged by total internal reflection fluorescence mi-

croscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
11.5 pH measurement and ratiometric imaging . . . . . . . . . . . . . . . . . 167
11.6 Single molecule microscopy . . . . . . . . . . . . . . . . . . . . . . . . . 170
11.6.1 Bulk versus single molecule experiments . . . . . . . . . . . . . . 170
11.6.2 Single molecule tracking experiments . . . . . . . . . . . . . . . 172
11.6.3 Localization-based super-resolution microscopy . . . . . . . . . . 173
11.6.3.1 Photophysics of the stochastic excitation of organic fluo-
rophores . . . . . . . . . . . . . . . . . . . . . . . . . . 175
11.6.4 A localization-based super-resolution experiment . . . . . . . . . 177
11.7 Multifocal plane microscopy . . . . . . . . . . . . . . . . . . . . . . . . . 177
11.7.1 Focal plane spacing and magnification . . . . . . . . . . . . . . . 179
11.7.2 Transferrin trafficking in epithelial cells . . . . . . . . . . . . . . 180
11.7.3 Imaging the pathway preceding exocytosis . . . . . . . . . . . . 181

12 Detectors 187
12.1 Photoelectric effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
12.2 Point detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
12.3 Image detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
12.3.1 Charge-coupled device (CCD) detectors . . . . . . . . . . . . . . 190
12.3.2 Complementary metal-oxide-semiconductor (CMOS) detectors . 191
12.3.3 Electron-multiplying charge-coupled device (EMCCD) detectors 192
12.4 Randomness of photon detection and detector noise sources . . . . . . . 193
12.5 Grayscale and color cameras . . . . . . . . . . . . . . . . . . . . . . . . . 194
12.6 Specifications of image detectors . . . . . . . . . . . . . . . . . . . . . . 196
12.7 Measurements of detector specifications . . . . . . . . . . . . . . . . . . 199
12.7.1 Determination of CCD and CMOS detector specifications . . . . 199
12.7.1.1 Data model . . . . . . . . . . . . . . . . . . . . . . . . 199
12.7.1.2 Linearity of the response . . . . . . . . . . . . . . . . . 200
12.7.1.3 Estimation of electron-count-to-DU conversion factor . 201
12.7.1.4 Estimation of readout noise mean and variance . . . . 202
12.7.1.5 Estimation of mean of dark current . . . . . . . . . . . 202
12.7.2 Determination of EMCCD detector specifications . . . . . . . . 202
12.7.2.1 Data model . . . . . . . . . . . . . . . . . . . . . . . . 202
12.7.2.2 Estimation of electron-count-to-DU conversion factor . 203
12.7.2.3 Estimation of readout noise mean and variance . . . . 204

13 Geometrical Optics 205

13.1 Reflection and refraction . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
13.1.1 Reflection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
13.1.2 Refractive index . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
13.1.3 Snell’s law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
13.1.4 Total internal reflection . . . . . . . . . . . . . . . . . . . . . . . 207
13.1.5 Extreme rays in microscopy optics . . . . . . . . . . . . . . . . . 207
13.2 Lenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
13.2.1 Focal points and focal planes . . . . . . . . . . . . . . . . . . . . 210
13.2.2 Image formation . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
13.2.3 Lensmaker’s formula and lens formula . . . . . . . . . . . . . . . 211
13.3 Magnification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
13.3.1 Lateral magnification . . . . . . . . . . . . . . . . . . . . . . . . 213
13.3.2 Axial magnification . . . . . . . . . . . . . . . . . . . . . . . . . 216
Contents xi

13.3.3 Dependence of lateral magnification on axial position . . . . . . 218

13.4 Applications to microscopy . . . . . . . . . . . . . . . . . . . . . . . . . 218

14 Diffraction 223
14.1 Wave description of light . . . . . . . . . . . . . . . . . . . . . . . . . . 223
14.1.1 Plane waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
14.1.1.1 Planes of identical phase . . . . . . . . . . . . . . . . . 224
14.1.1.2 Speed of wave propagation . . . . . . . . . . . . . . . . 225
14.1.1.3 Wave number and wavelength . . . . . . . . . . . . . . 226
14.1.1.4 Propagation in different media . . . . . . . . . . . . . . 226
14.1.1.5 Optical path length . . . . . . . . . . . . . . . . . . . . 227
14.1.2 Spherical waves . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
14.1.2.1 Converging and diverging spherical waves . . . . . . . . 228
14.1.3 Spatial part of a wave . . . . . . . . . . . . . . . . . . . . . . . . 228
14.2 What does a camera detect? . . . . . . . . . . . . . . . . . . . . . . . . 229
14.3 Effect of a thin lens on waves . . . . . . . . . . . . . . . . . . . . . . . . 230
14.4 Huygens-Fresnel principle and Fresnel integral . . . . . . . . . . . . . . . 234
14.4.1 Huygens-Fresnel principle . . . . . . . . . . . . . . . . . . . . . . 235
14.5 Imaging through a thin lens . . . . . . . . . . . . . . . . . . . . . . . . . 237
14.5.1 Amplitude point spread function . . . . . . . . . . . . . . . . . . 239
14.5.2 Convolution description . . . . . . . . . . . . . . . . . . . . . . . 240
14.5.3 Relationship to geometrical optics . . . . . . . . . . . . . . . . . 241
14.5.4 Point spread function and Fourier transformation . . . . . . . . 241
14.5.4.1 In-focus point spread function . . . . . . . . . . . . . . 243
14.5.5 Imaging with defocus and the 3D point spread function . . . . . 245
14.5.5.1 3D point spread function evaluated on the optical axis 247
14.5.5.2 Depth of field and depth of focus . . . . . . . . . . . . 249
14.5.5.3 Heuristic 3D resolution criterion . . . . . . . . . . . . . 251
14.6 Convolution for intensity profiles . . . . . . . . . . . . . . . . . . . . . . 251

Notes 255

Exercises 259

IV Data Analysis 263

Overview 265

15 From Photons to Image: Data Models 267

15.1 Accounting for each photon: fundamental data model . . . . . . . . . . 267
15.1.1 Temporal component of photon detection — Poisson process . . 268
15.1.1.1 Mean number of detected photons . . . . . . . . . . . . 270
15.1.2 Spatial component of photon detection — spatial density function 272
15.1.2.1 Translational invariance and image function . . . . . . 273
15.1.3 Background component . . . . . . . . . . . . . . . . . . . . . . . 276
15.1.4 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
15.2 Practical data models . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
15.2.1 Poisson data model . . . . . . . . . . . . . . . . . . . . . . . . . 279
15.2.2 CCD/CMOS data model . . . . . . . . . . . . . . . . . . . . . . 284
15.2.3 Deterministic data model . . . . . . . . . . . . . . . . . . . . . . 285
15.2.3.1 Gaussian approximation for the CCD/CMOS data model 286
15.2.4 EMCCD data model . . . . . . . . . . . . . . . . . . . . . . . . . 286
xii Contents

15.2.4.1 High gain approximation for the EMCCD data model . 288
15.2.4.2 Gaussian approximation for the EMCCD data model . 289

16 Parameter Estimation 293

16.1 Maximum likelihood estimation . . . . . . . . . . . . . . . . . . . . . . . 293
16.1.1 Example 1: mean of a Poisson random variable . . . . . . . . . . 294
16.1.2 Example 2: mean of a Gaussian random variable . . . . . . . . . 295
16.2 Log-likelihood functions for the image data models . . . . . . . . . . . . 297
16.2.1 Log-likelihood function for the fundamental data model . . . . . 297
16.2.1.1 Example 3: Localization of an object with a 2D Gaussian
image profile . . . . . . . . . . . . . . . . . . . . . . . . 299
16.2.2 Log-likelihood functions for the practical data models . . . . . . 300
16.3 Obtaining the maximum likelihood estimate . . . . . . . . . . . . . . . . 302
16.4 Maximum likelihood estimation and least squares estimation . . . . . . 303
16.5 Unbiased estimator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
16.5.1 Example 1: sample mean . . . . . . . . . . . . . . . . . . . . . . 304
16.5.2 Example 2: sample variance . . . . . . . . . . . . . . . . . . . . . 304
16.5.3 Example 3: center of mass as an object location estimator under
the fundamental data model . . . . . . . . . . . . . . . . . . . . 306
16.6 Variance of an estimator . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
16.6.1 Example 1: mean of a Poisson and a Gaussian random variable . 308
16.6.2 Example 2: center of mass as an object location estimator under
the fundamental data model . . . . . . . . . . . . . . . . . . . . 309
16.6.3 Example 3: center of mass as an object location estimator under a
fixed photon count data model . . . . . . . . . . . . . . . . . . . 311

17 Fisher Information and Cramér-Rao Lower Bound 313

17.1 Cramér-Rao inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
17.1.1 Sketch of derivation of Cramér-Rao lower bound . . . . . . . . . 314
17.1.2 Multivariate Cramér-Rao lower bound . . . . . . . . . . . . . . . 316
17.1.3 Example 1: mean of a Poisson random variable . . . . . . . . . . 317
17.1.4 Example 2: mean of a Gaussian random variable . . . . . . . . . 318
17.2 Fisher information for the fundamental data model . . . . . . . . . . . . 318
17.2.1 Example: known photon detection rate . . . . . . . . . . . . . . 320
17.2.2 Example: known photon distribution profile . . . . . . . . . . . . 321
17.3 Fisher information for the practical data models . . . . . . . . . . . . . 322
17.3.1 Noise coefficient and the Fisher information . . . . . . . . . . . . 323
17.4 Noise coefficient analysis of the pixel signal level . . . . . . . . . . . . . 327
17.4.1 Noise coefficient — an in-depth look . . . . . . . . . . . . . . . . 327
17.4.2 Noise coefficient for CCD/CMOS detectors . . . . . . . . . . . . 327
17.4.3 EMCCD detectors as low-light detectors . . . . . . . . . . . . . 328
17.4.4 Comparison of CCD/CMOS and EMCCD detectors . . . . . . . 329
17.5 Fisher information for multi-image data . . . . . . . . . . . . . . . . . . 330

18 Localizing Objects and Single Molecules in Two Dimensions 331

18.1 Object localization as a parameter estimation problem . . . . . . . . . . 331
18.2 Example: estimating the location of a single molecule . . . . . . . . . . 333
18.3 How well can the location of an object be estimated? . . . . . . . . . . . 335
18.3.1 Bias of location estimation . . . . . . . . . . . . . . . . . . . . . 335
18.3.1.1 Bias of the center of mass as a location estimator under
the practical data models . . . . . . . . . . . . . . . . . 338
Contents xiii

18.3.2 Variance of location estimation . . . . . . . . . . . . . . . . . . . 338

18.4 Estimation of other parameters . . . . . . . . . . . . . . . . . . . . . . . 340
18.5 Cramér-Rao lower bound for location estimation — fundamental data
model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
18.5.1 Cramér-Rao lower bound for the Airy image function . . . . . . 344
18.5.2 Cramér-Rao lower bound for the 2D Gaussian image function . 346
18.5.3 Extensions to further experimental situations . . . . . . . . . . . 347
18.6 Cramér-Rao lower bound for location estimation — practical data model 348
18.6.1 Poisson data model — effects of pixelation, finite image size, and
background noise . . . . . . . . . . . . . . . . . . . . . . . . . . 349
18.6.2 Localizing objects from CCD/CMOS and EMCCD images . . . 351
18.6.3 Object location makes a difference . . . . . . . . . . . . . . . . . 354
18.7 Efficiency of estimators: how well is the behavior of estimators described by
the Cramér-Rao lower bound? . . . . . . . . . . . . . . . . . . . . . . . . 356
18.7.1 Fundamental data model . . . . . . . . . . . . . . . . . . . . . . 356
18.7.2 Practical data models . . . . . . . . . . . . . . . . . . . . . . . . 357
18.8 Approximations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
18.8.1 Gaussian approximations for the CCD/CMOS and EMCCD data
models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
18.8.2 Inverse square root approximation of the dependence on the photon
count . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
18.9 Lower bound as a tool for the design of data analysis . . . . . . . . . . . 360
18.9.1 Choosing the region of interest . . . . . . . . . . . . . . . . . . . 361
18.9.2 Improving estimation performance by adding images . . . . . . . 362
18.10 Example: single molecule localization from experimentally acquired images 363
18.10.1 Choice of data model based on the detector used . . . . . . . . . 365
18.10.2 Modeling the image of the molecule and the background component 365
18.10.3 Determining the “known” parameters . . . . . . . . . . . . . . . 366
18.10.4 Location estimates . . . . . . . . . . . . . . . . . . . . . . . . . . 367
18.10.5 Initial values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
18.10.6 Assessing the standard deviation of the localization . . . . . . . 368

19 Localizing Objects and Single Molecules in Three Dimensions 371

19.1 Parameter estimation for object localization in three dimensions . . . . 371
19.2 Cramér-Rao lower bound for 3D location estimation — fundamental data 372
19.2.1 3D localization of a point source . . . . . . . . . . . . . . . . . . 374
19.3 Cramér-Rao lower bound for 3D location estimation — practical data . 378
19.4 Depth discrimination problem . . . . . . . . . . . . . . . . . . . . . . . . 379
19.5 Dependence of lateral location estimation on the axial position . . . . . 381
19.6 Multifocal plane microscopy . . . . . . . . . . . . . . . . . . . . . . . . . 382
19.6.1 Estimating the axial location from MUM data . . . . . . . . . . 384
19.6.2 Experimental example . . . . . . . . . . . . . . . . . . . . . . . . 386
19.6.3 Maximum likelihood localization with simulated data . . . . . . 386
19.6.4 Overcoming the depth discrimination problem . . . . . . . . . . 389
19.6.5 Zero Fisher information and the depth discrimination problem . 391
19.6.6 Experimental design: finding appropriate focal plane spacings . . 392
19.6.7 Further approaches to address the depth discrimination problem 394
xiv Contents

20 Resolution 395
20.1 Resolution as a parameter estimation problem . . . . . . . . . . . . . . 395
20.2 Cramér-Rao lower bound for distance estimation — fundamental data . 395
20.3 Two in-focus objects: an information-theoretic Rayleigh’s criterion . . . 397
20.4 Two objects in 3D space . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
20.5 Cramér-Rao lower bound for distance estimation — practical data . . . 405

21 Deconvolution 411
21.1 The deconvolution problem . . . . . . . . . . . . . . . . . . . . . . . . . 411
21.2 Discretization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
21.2.1 Linear algebra formulation . . . . . . . . . . . . . . . . . . . . . 414
21.3 Linear least squares algorithm . . . . . . . . . . . . . . . . . . . . . . . . 414
21.3.1 Condition number of a matrix . . . . . . . . . . . . . . . . . . . 415
21.3.1.1 Example of an ill-conditioned least squares problem . . 416
21.3.2 Regularization of the least squares problem . . . . . . . . . . . . 416
21.3.2.1 Example continued: regularization of the ill-conditioned
least squares problem . . . . . . . . . . . . . . . . . . . 417
21.3.3 A Fourier transform approach . . . . . . . . . . . . . . . . . . . 418
21.4 Maximum likelihood formulation . . . . . . . . . . . . . . . . . . . . . . 419
21.4.1 Expectation maximization algorithm . . . . . . . . . . . . . . . . 419
21.5 Positron emission tomography . . . . . . . . . . . . . . . . . . . . . . . . 421
21.5.1 Deconvolution for the Poisson data model . . . . . . . . . . . . . 424
21.5.2 An illustrative example . . . . . . . . . . . . . . . . . . . . . . . 425

22 Spatial Statistics 427

22.1 Formal definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
22.1.1 Spatial Poisson processes . . . . . . . . . . . . . . . . . . . . . . 428
22.2 Intensity functions of spatial processes . . . . . . . . . . . . . . . . . . . 429
22.2.1 Computing the intensity functions . . . . . . . . . . . . . . . . . 430
22.2.2 Stationary point processes . . . . . . . . . . . . . . . . . . . . . 431
22.3 K function and L function . . . . . . . . . . . . . . . . . . . . . . . . . 432
22.3.1 An example of an inhibition process . . . . . . . . . . . . . . . . 433
22.3.2 Estimating the K function . . . . . . . . . . . . . . . . . . . . . 435

Notes 439

Exercises 445

Online Appendix — Mathematical Background

http://www.routledge.com/9781138598980 A-1
A Probability and Statistics A-1
A.1 Tutorial: Poisson and Gaussian random variables . . . . . . . . . . . . . A-1
A.1.1 Poisson random variables . . . . . . . . . . . . . . . . . . . . . . A-3
A.1.1.1 Additivity of Poisson random variables . . . . . . . . . A-4
A.1.2 Gaussian random variables . . . . . . . . . . . . . . . . . . . . . A-4
A.2 Expectation of a Poisson random variable given a sum of Poisson random
variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-5
A.3 Additivity of Fisher information matrices . . . . . . . . . . . . . . . . . A-7
Contents xv

B Analysis A-9
B.1 Delta function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-9
B.2 Taylor series approximation . . . . . . . . . . . . . . . . . . . . . . . . . A-9
B.3 Change of variables theorem . . . . . . . . . . . . . . . . . . . . . . . . . A-9
B.3.1 Change of Cartesian coordinates to polar coordinates . . . . . . A-10

C Fourier Transform A-11

C.1 Fourier transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-11
C.1.1 Fourier transform of circularly symmetric functions . . . . . . . A-12
C.2 Discrete Fourier transform . . . . . . . . . . . . . . . . . . . . . . . . . . A-12
C.3 Multidimensional discrete Fourier transform . . . . . . . . . . . . . . . . A-14

D Least Squares Minimization A-17

D.1 Least squares minimization problem . . . . . . . . . . . . . . . . . . . . A-17
D.2 Linear least squares in the Fourier domain . . . . . . . . . . . . . . . . . A-19

E Fisher Information for the Fundamental Data Model A-21

F Fisher Information for the Deterministic Data Models A-29

F.1 Simple form of the deterministic data model . . . . . . . . . . . . . . . . A-29
F.2 Gaussian approximation for the CCD/CMOS data model . . . . . . . . A-30
F.3 Gaussian approximation for the EMCCD data model . . . . . . . . . . . A-30

G Models of EMCCD Electron Multiplication A-31

G.1 Geometric multiplication-based EMCCD probability mass function . . . A-31
G.1.1 Branching processes . . . . . . . . . . . . . . . . . . . . . . . . . A-31
G.1.2 Electron multiplication as a branching process . . . . . . . . . . A-32
G.1.3 Geometric distribution of offspring electrons . . . . . . . . . . . A-32
G.1.4 Probability generating function of output electron count given one
initial electron . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-32
G.1.5 Probability generating function of output electron count given i
initial electrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-33
G.1.6 Probability mass function of output electron count given i initial
electrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-33
G.1.7 Probability mass function of output electron count . . . . . . . . A-36
G.2 Exponential multiplication-based EMCCD probability density function . A-37
G.3 Geometric multiplication-based EMCCD noise coefficient . . . . . . . . A-38
G.4 Exponential multiplication-based EMCCD noise coefficient . . . . . . . A-40

Notes A-43

Exercises A-45

Figure Credits 451

Bibliography 457

List of Symbols 465

Index of Names 469

Index 471
Preface

The investigation of cellular processes is undergoing revolutionary developments that are

propelled by modern technologies that allow the study of questions in ways that not long
ago were unimaginable. One of these technologies is imaging and, in particular, fluorescence
microscopy, which has the specificity to investigate the roles and interactions of different
molecular species in a cellular environment. This specificity has been made possible by new
fluorescent labeling techniques, foremost amongst them those based on fluorescent proteins,
which permit the specific labeling of receptors and other molecules in cells. Importantly,
imaging in general has become a significantly more powerful technology with the advent
and continued development of, for example, modern light sources and high-sensitivity image
detectors. Light sources such as lasers have enabled or facilitated various methods of sample
illumination that can improve the image quality. At the same time, modern detectors, com-
bined with computer control, have allowed the recording of vast amounts of image data at
high spatial and temporal resolution. Additionally, advances in data acquisition have been
made more valuable by the development of sophisticated data analysis algorithms that en-
able the extraction of biological quantities of interest from the acquired images. In the case
of fluorescence microscopy, the use of appropriate illumination techniques in conjunction
with suitable detectors and data analysis algorithms have enabled the visualization and
accurate quantitation of highly dynamic cellular processes at the level of individual, specif-
ically labeled molecules of interest. Indeed, imaging at the single molecule level has been
one of the more recent high points in the field of quantitative bioimaging.
The fascination of quantitative bioimaging is that it is a paradise for multi- and inter-
disciplinary research as important aspects of different disciplines are necessary for mastering
the technique. A detailed understanding of the biological processes that are investigated
is required to develop a well-thought-out experimental plan. Knowledge of chemistry is
needed as approaches from chemistry form the basis for the labeling of molecules of interest
with fluorescent tags. A good understanding of optics is crucial as it is the discipline that
informs us about the principles of image formation. Familiarity with statistical data analysis
is important as the theory and mathematics involved underlie the ever more sophisticated
image analysis algorithms that need to be employed to analyze the acquired image data.
Classically, textbooks on quantitative bioimaging and microscopy focus on one of the
underlying disciplines, for example the optical principles or image analysis. There are also
textbooks that concentrate on the practical aspects of using specific imaging techniques.
Here we are pursuing a different approach as we present, in a fair amount of depth, the
various constituent scientific components of our composite discipline. The discussion of two
important biological problems forms a thread throughout a significant part of the book and
allows us to introduce experimental considerations using concrete biological questions.
Quantitative bioimaging, and microscopy in particular, is a vast field, and it is not
possible to cover all important aspects in significant detail. As we believe strongly that it is
important to understand the fundamental principles behind cellular microscopy experiments
and their analysis, we have opted to present some key aspects within that context in detail
and depth. Invariably there are personal preferences in our choices of emphasis and other
authors might have chosen other topics to highlight. We believe that students who study

xvii