University of West Attica
3-διάστατη απεικόνιση ιστολογικών δεδομένων στην οπτική μικροσκοπία διερχομένου φωτός
Abstract
dc:descriptionOptical microscopy, also known as light microscopy, is a versatile and widely used technique for imaging and studying objects at the microscale and nanoscale level using visible light and optical lenses. It is a fundamental tool in various scientific fields such as biology, materials science, chemistry and medicine. In particular, bright-field microscopy, a fundamental technique in the field of optical microscopy, is one of the simplest and most common methods for observing biological and other transparent samples. However, there is deficiency of comprehensive 3D histological data acquisition and analysis techniques using conventional optical microscopy of transmitted light. Existing methodologies predominantly offer 2D snapshots of tissue sections, overlooking the intricate 3D microarchitecture that plays a vital role in understanding tissue function, development, and disease progression. Three-dimensional imaging is possible using specιαlized microscopes which are usually quite expensive. The question arises as to whether 3D imaging of a tissue is feasible using a conventional optical microscope of transmitted light. In this thesis, we first describe the basic principles of microscopy and common laboratory techniques in histopathology. Then, two techniques for three-dimensional tissue reconstruction from bright-field microscopy images are presented. The first technique involves automating a series of steps to create a digital 3D volume of a tissue from conventional bright-field microscope images. Specifically, digital images of region of interest (defined by an expert) are collected from successive tissue sections. Image registration and deep learning-based interpolation techniques are then used to generate a 3D volume of the region of interest. An expert histopathologist evaluated the information provided by the 3D volume and the results showed that the 3D volume may reveal details that are not clearly visible or even undetectable in the images of a conventional microscope. Unlike other 3D tissue imaging technologies, this technique has the following advantages. (b) It does not require a special device; it can be applied with any conventional bright-field microscope. (c) It can be used for any given routine application, not only for certain specialized clinical studies. The second technique involves the tomographic imaging of a tissue section with optical sectioning (as opposed to the previous technique which involves physical sectioning of the tissue under examination). This technique is based on the idea that each tissue section has a finite thickness and therefore it is possible to create images in different planes (optical slicing) within the tissue section, revealing details that would probably not be seen otherwise. Optical slicing was made possible by the development of a special device (using a stepper motor and Arduino) and software, which allow rotation of the slide containing the tissue section under consideration around the horizontal axis and image acquisition at each position. This rotation converts the optical microscope into a CT geometry with limited number of angles. In this way, tomographic reconstruction algorithms from projections can be used to visually slice the tissue section. Two classical reconstruction algorithms (the filtered back-projection and the algebraic reconstruction technique) were tested. The resulting images were of satisfactory quality, but exhibited some artefacts. In conclusion, the use of classical tomographic reconstruction algorithms with a limited number of projections allows to explore the sample in any desired optical plane, revealing information that would be difficult to detect when focusing only on conventional two-dimensional microscope images.
Degree
thesis:*- Grantor dc:publisher
- University of West Attica
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Koudounas, Panteleimon
- Κουδουνάς, Παντελεήμων
Subjects
dc:subject × 29- Οπτική μικροσκοπία
- Μικροσκοπία φωτεινού πεδίου
- Τομές ιστού
- Τρισδιάστατη ανακατασκευή
- Τομογραφική απεικόνιση
- Ευθυγράμμιση εικόνων
- Φιλτραρισμένη οπισθοπροβολή
- Αλγεβρική μέθοδος ανακατασκευής
- Βαθιά μάθηση
- Optical microscopy
- Βright field microscopy
- Τissue sections
- 3d reconstruction
- Τomographic imaging
- Image registration
- Filtered backprojection
- Algebraic reconstruction method
- Deep learning algorithms
- Άλλες Επιστήμες Μηχανικού και Τεχνολογίες
- Επιστήμες Μηχανικού και Τεχνολογία
- Μηχανική, γενικά
- Βιοϊατρική Μηχανική
- Τεχνολογία ιατρικών εργαστηρίων
- Other Engineering and Technologies
- Engineering and Technology
- Engineering, general
- Medical Engineering
- Biomedical Engineering
- Medical Laboratory Technology
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/55201