|Title:||Automatic classification of histological hyperspectral images: algorithms and instrumentation||Authors:||Ortega Sarmiento, Samuel||Director:||Marrero Callicó, Gustavo Iván
|UNESCO Clasification:||3314 Tecnología médica||Issue Date:||2021||Abstract:||Hyperspectral imaging (HSI), also known as imaging spectroscopy, is a technology capable of sampling hundreds of narrow spectral bands across the electromagnetic spectrum through the use of optical elements that disperse the incoming radiation into discrete wavelengths. This technology combines the main features of two existing technologies: imaging and spectroscopy, making it possible to exploit both the morphological features and the chemical composition of objects captured by a camera. The interaction between electromagnetic radiation and matter is distinctive for each material, therefore by using this technology it is possible to discriminate among different materials. Although historically HSI has been applied to remote sensing, in recent years this technology has become a trending topic in different research fields such as food quality analysis, military and security applications or precision agriculture, among many others. HSI is also an emerging imaging modality in the medical field. The study of light propagation through biological tissues is useful to identify several diseases. These properties of the interaction between light and biological tissue motivate the use of technologies that exploit the information of light propagation through tissues to develop tools for diagnosis support. As an alternative diagnostic tool, one of the strengths offered by HSI is being completely non-invasive and label-free. Traditional computational pathology, also known as digital pathology, is an emerging technology that promises quantitative diagnosis of pathological samples, reduction of inter-observer variability among pathologists, and saving time in the manual examination of histological samples. Traditional computational pathology relies on RGB digitized histology images. Within computational pathology, several research groups have begun to explore whether hyperspectral/multispectral (HS/MS) imaging are technologies able to provide further advantages to this end. In this Ph.D. thesis, we evaluate the potential of HSI as a diagnostic tool for the analysis of histological samples. First, we perform a literature systematic review, where we analyze the use of both HSI and MSI for pathological diagnosis, digital staining and other similar applications. Such systematic review adheres to the guidelines of the preferred reporting items for systematic reviews and meta-analyses (PRISMA). Additionally, we survey the most common processing methods which are used to extract useful information for disease detection and diagnosis using HSI. Second, we characterize the instrumentation used to capture microscopic HS images, and we propose a methodology where we propose some recommendations to correctly set up the instrumentation in order to acquire high quality microscopic HS images. Next, different databases composed by histological HS data from both brain and breast tumors have been generated. The specimens consist of pathological slides where the Pathologists have indicated the areas corresponding to a concrete diagnosis, i.e., tumor or not tumor areas. The pathological slides used in this thesis were processed and analyzed by the Pathological Anatomy Department of the University Hospital Doctor Negrín at Las Palmas of Gran Canaria (Las Palmas de Gran Canaria, Spain), and by the Department of Pathology from the Tortosa Verge de la Cinta Hospital (Tortosa, Spain) Due to the nature of the problem, HS images are processed using both Machine Learning and Deep Learning algorithms in order to evaluate the performance of automatic diagnosis using HS images. In this dissertation, we demonstrate that the combination of hyperspectral microscopic imaging and image processing techniques is a promising tool for future computational pathologies.||Description:||Programa de Doctorado en Tecnologías de Telecomunicación e Ingeniería Computacional por la Universidad de Las Palmas de Gran Canaria||URI:||http://hdl.handle.net/10553/107311|
|Appears in Collections:||Tesis doctoral|
checked on Oct 29, 2022
checked on Oct 29, 2022
Items in accedaCRIS are protected by copyright, with all rights reserved, unless otherwise indicated.