Introduction

Corinne FOURNIER1 and Olivier HAEBERLÉ2

1Laboratoire Hubert Curien, CNRS, IOGS, Télécom Saint-Étienne, Université Jean Monnet, Saint-Étienne, France

2IRIMAS, Université de Haute-Alsace, Mulhouse, France

I.1. Context

Optical imaging in biological systems has undergone spectacular developments in recent years, which have resulted in providing a quantity and a quality of information that, until about 20 years ago, was merely a dream for physicists, biologists or physicians.

The extraordinary progress that has been made can be explained through a combination of contributions in the field of the physics of image formation, in instrumentation with the arrival of new sensors with remarkable performance, and finally by the extraordinary computing power and memory capacity of current computers. To illustrate these remarks, we can refer to two spectacular examples (not addressed in this book).

In fluorescence microscopy, the diffraction barrier has been “broken” in a spectacular way thanks to the developments of STED (Stimulated Emission Depletion) microscopy first, and followed by the invention of pointillist microscopies. STED microscopy utilizes the temporal properties of the fluorescence phenomenon, which is not instantaneous, to cause a stimulated emission. The latter is controlled by the system, being spatially, temporally, spectrally and directionally dissociable from the spontaneous fluorescence emission. Proposed as early as 1994, this technique has only recently become widespread, with the emergence of inexpensive and high-performance lasers and detectors. Pointillist microscopies make use of the statistical properties of fluorescence emission, which exhibits a random nature. By processing large quantities of individual images, with few fluorescent spots, it is possible to detect and locate them individually, and thus to reconstruct an image with improved resolution. These techniques have now even made imaging of living specimens possible; this is mainly due to the quality of fluorescence labelings, the right photon yield of fluorophores, but especially the extreme sensitivity of modern cameras, and the speed of image processing allowed by current computers.

Another field that has undergone remarkable development is imaging through scattering media. A large body of techniques has been developed or improved in recent years. The most astounding is perhaps the one that makes use of scattering matrices. It is based on the fact that, in an imaging system, the Green’s matrix (set of Green’s functions linking each source emitting light to each detector pixel) contains the information accessible in linear scattering processes. Modern experimental means (sensors of large dimension and sensitivity, spatial light modulators with many degrees of freedom, and powerful computers) now contribute to experimentally characterize scattering media by accurately measuring the scattering matrix. Impressive operations until recently unimaginable can now be performed; these include the characterization of surface or volume scattering, separation of the contributions of single and multiple scattering and even focusing through a scattering medium, which paves the way to both transmission and reflection imaging.

This book is concerned with other areas of biological imaging that have also been driven, like these two examples, by advances in instrumentation and computer science, and that have led to making possible concepts that were until recently purely theoretical.

I.2. Unconventional imaging

Unconventional imaging, as opposed to conventional imaging, allows us to access physical quantities (such as the opacity, the refractive index, the property of wave polarization, the chemical composition of an object and so on) not directly accessible by optical systems, in which sensors can only measure information in terms of intensity.

The use of particular optical setups and computer processing for the captured images/signals has enabled the reconstruction of physical quantities. This is also referred to as computational imaging. Typical unconventional imaging modalities are: polarimetry, interferometry, holography and phase imaging, as well as hyperspectral imaging. Unconventional imaging can also be used for the purposes of miniaturizing, reducing the costs or improving the quantitativity of conventional imaging systems. This type of imaging requires the co-design of the optical system, sensors and signal and image processing algorithms. A large variety of information obtained through unconventional imaging allows for improved detection, quantitative characterization and classification of imaged objects. These systems are used in many fields of biological imaging, namely endoscopy, microscopy, skin or through-skin imaging and fast phenomena imaging.

I.3. Book contents

In this book, we focus on various unconventional imaging modalities developed for the biomedical field; these include wavefront analysis imaging, digital holography, optical nanoscopy, endoscopy and single-pixel imaging. The first six chapters are complementary and address phase-imaging.

Chapter 1 presents quantitative phase microscopy using a wavefront analyzer. Wavefront analyzers are particular devices that are capable of measuring a wave without making use of holography (more details in the following chapters). In this chapter, the notion of wave phase is first recalled, followed by an introduction to the notion of phase object in the field of biology. The particular modality of quadriwave lateral shearing interferometry is described in detail and applied to solve major problems in biological imaging, such as dry mass measurements, the study of fast biological phenomena and birefringence measurements.

Chapter 2 presents digital holography through a detailed presentation of the principle of interferometry used in holography, based on superposing a reference wave onto the wave diffracted by the object. Introduced by Gabor in electron microscopy, holography is in fact known in particular for its applications in optics. The most commonly used holographic configurations are described. Finally, the digital processing of the obtained data, in order to extract the characteristics of the measured wavefront, in amplitude and phase, is detailed. This chapter serves as an introduction to Chapter 4, which deals with digital holography, in the configuration known as in-line holography, as well as to Chapter 5, which presents tomographic diffractive microscopy.

Chapter 3 is concerned with a general methodology for numerical reconstruction using an approach based on “inverse problems”. This approach leads to estimating the parameters of interest of imaged objects from unconventional images. It is presented in a very general way and thus can be applied to different modalities of unconventional imaging. This chapter presents the simple case of likelihood maximization between data and linear imaging models, and then addresses the problem of the nonlinear phase reconstruction (phase retrieval) with or without taking into account the a priori on the reconstructed objects. This chapter contains examples of in-line digital holography in connection with the next chapter.

Chapter 4 is dedicated to in-line holographic microscopy. It presents the experimental configurations of holographic microscopy used in the biomedical field along with their specificities. The problem of numerical reconstruction is addressed by comparing different historical and state-of-the-art approaches with experimental images. As a follow-up to Chapter 3, approaches based on “inverse problems” are used and their potential for holographic microscopy is discussed.

Chapter 5 gives an extension of holography, with application to transmission microscopy. Tomographic diffractive microscopy is a technique coupling holography with scanning light illumination on the specimen or with a rotation thereof. This technique allows for much more information to be acquired, which results in a 3D reconstruction of the distribution of optical indices in the observed specimen. The refractive index (in refraction and absorption) has become a contrast allowing for many applications on specimens without preparation (no coloring, no fluorescent tagging), which now finds applications in many fields (such as the study of stem cells, yeasts, bacterial growth and so on).

Chapter 6 describes another phase microscopy technique, known as white light interferometric microscopy. Based on a particular configuration (Linnick interferometer), it differs from the techniques described in the previous chapters in that it is a reflection technique. The demodulation of the reflected signal leads to unequaled precision of measurement, namely in the nanometric range. However, the lateral resolution remains that of a conventional optical microscope. A recent extension involves combining microsphere-assisted microscopy with interferometry, paving the way for surface nanometrology. These techniques are now also used for biological imaging, in particular cellular or tissue imaging.

Chapter 7 addresses white light and endoscopic fluorescence imaging. Following a presentation of the principles of endoscopy, particular emphasis is given to the problem of the 3D reconstruction of hollow organs observed through this technique. Endoscopic images have specific characteristics, requiring dedicated approaches, namely the low...