An eye-opening discussion of 3D optical sensing, imaging, analysis, and pattern recognition
Artificial intelligence (AI) has made great progress in recent years. Digital holographic imaging has recently emerged as a powerful new technique well suited to explore cell structure and dynamics with a nanometric axial sensitivity and the ability to identify new cellular biomarkers. By combining digital holography with AI technology, including recent deep learning approaches, this system can achieve a record-high accuracy in non-invasive, label-free cellular phenotypic screening. It opens up a new path to data-driven diagnosis.
Artificial Intelligence in Digital Holographic Imaging introduces key concepts and algorithms of AI to show how to build intelligent holographic imaging systems drawing on techniques from artificial neural networks, convolutional neural networks, and generative adversarial network. Readers will be able to gain an understanding of the basics for implementing AI in holographic imaging system designs and connecting practical biomedical questions that arise from the use of digital holography with various AI algorithms in intelligence models.
What’s Inside - Introductory background on digital holography - Key concepts of digital holographic imaging - Deep-learning techniques for holographic imaging - AI techniques in holographic image analysis - Holographic image-classification models - Automated phenotypic analysis of live cells
For readers with various backgrounds, this book provides a detailed discussion of the use of intelligent holographic imaging system in biomedical fields with great potential for biomedical application.
Table of Contents
Part I. Digital Holographic Microscopy (DHM)
1. Introduction
References
2. Coherent optical imaging
2.1 Monochromatic fields and irradiance
2.2 Analytic expression for Fresnel diffraction
2.3 Transmittance function of lens
2.4 Geometrical imaging concepts
2.5 Coherent imaging theory
References
3. Lateral and depth resolutions
3.1 Lateral resolution
3.2 Depth (or axial) resolution
References
4. Phase unwrapping
4.1 Branch cuts
4.2 Quality-guided path-following algorithms
References
5. Off-axis digital holographic microscopy
5.1 Off-axisdigital holographic microscopy designs
5.2 Digital hologram reconstruction
References
6. Gabor digital holographic microscopy
6.1 Introduction
6.2 Methodology
References
Part II. Deep Learning in DHM Systems
7. Introduction
References
8. No-search focus prediction in DHM with deep learning
8.1 Introduction
8.2 Materials and methods
8.3 Experimental results
8.4 Conclusions
References
9. Automated phase unwrapping in DHM with deep learning
9.1 Introduction
9.2 Deep learning model
9.3 Unwrapping with deep learning model
9.4 Conclusions
References
10. Noise-free phase imaging in Gabor DHM with deep learning
10.1 Introduction
10.2 A deep learning model for Gabor DHM
10.3 Experimental results
10.4 Discussion
10.5 Conclusions
References
Part III. Intelligent DHM for Biomedical Applications
11. Introduction
References
12. Red blood cells phase image segmentation
12.1 Introduction
12.2 Marker-controlled watershed algorithm
12.3 Segmentation based on marker-controlled watershed algorithm
12.4 Experimental results
12.5 Performance evaluation
12.6 Conclusions
References
13. Red blood cells phase image segmentation with deep learning
13.1 Introduction
13.2 Fully convolutional neural networks
13.3 Red blood cells phase image segmentation via deep learning
13.4 Experimental results
13.5 Conclusions
References
14. Automated phenotypic classification of red blood cells
14.1 Introduction
14.2 Feature extraction
14.3 Pattern recognition neural network
14.4 Experimental results and discussion
14.5 Conclusions
References
15. Automated analysis of red blood cell storage lesions
15.1 Introduction
15.2 Quantitative analysis of red blood cell 3D morphological changes
15.3 Experimental results and discussion
15.4 Conclusions
References
16. Automated red blood cells classification with deep learning
16.1 Introduction
16.2 Proposed deep learning model
16.3 Experimental results
16.4 Conclusions
References
17. High-throughput label-free cell counting with deep neural networks
17.1 Introduction
17.2 Materials and methods
17.3 Experimental results
17.4 Conclusions
References
18. Automated tracking of temporal displacements of red blood cells
18.1 Introduction
18.2 Mean-shift tracking algorithm
18.3 Kalman filter
18.4 Procedure for single RBC tracking
18.5 Experimental results
18.6 Conclusions
References
19. Automated quantitative analysis of red blood cells dynamics
19.1 Introduction
19.2 Red blood cell parameters
19.3 Quantitative analysis of red blood cell fluctuations
19.4 Conclusions
References
20. Quantitative analysis of red blood cells during temperature elevation
20.1 Introduction
20.2 Red blood cell sample preparations
20.3 Experimental results
20.4 Conclusions
References
21. Automated measurement of cardiomyocytes dynamics with DHM
21.1 Introduction
21.2 Cell culture and imaging
21.3 Automated analysis of cardiomyocytes dynamics
21.4 Conclusions
References
22. Automated analysis of cardiomyocytes with deep learning
22.1 Introduction
22.2 Region of interest identification with dynamic beating activity analysis
22.3 Deep neural network for cardiomyocytes image segmentation
22.4 Experimental results
22.5 Conclusions
References
23. Automatic quantification of drug-treated cardiomyocytes with DHM
23.1 Introduction
23.2 Materials and methods
23.3 Experimental results and discussion
23.4 Conclusions
References
24. Analysis of cardiomyocytes with holographic image-based tracking
24.1 Introduction
24.2 Materials and methods
24.3 Experimental results and discussion
24.4 Conclusions
References
25. Conclusion and future work
