Hyperspectral imaging is a hybrid modality that combines both spectroscopy and digital imaging [1]. Unlike other techniques, such as RGB, this method allows the continuous acquisition of a spectrum for each individual pixel, not just in the visible range [1]. The final data from this approach are presented in the form of a "hyperspectral cube" identified by two spatial and one spectral dimension [1] . Over time, this modality has been widely used in a number of sectors, such as agriculture, forensics, industry, medicine and disease diagnosis, due to its ability to discriminate and identify the spectral content of chemical compounds that make up a substance [1]. The main objective of this study is to verify how a hyperspectral system can be used in different biomedical fields. The first field examined is ophthalmology, in particular the combination of the Senop HSC-2 hyperspectral camera mounted on a Haag-Streit slit lamp eyepiece. Initially, a calibration procedure was performed using the Spectralon, appropriately inserted into the cavity of a mask simulating the human face. This first step made it possible to obtain the white reference target for all the subsequent acquisitions. Subsequently, discussing with a doctor specialized in ophthalmology, it was possible to understand whether this instrument could be valid as a medical device for the diagnosis and identification of ocular pathologies. The second field analyzed, using both the Senop-HSC-2 and HinaLea hyperspectral cameras, is dermatology. The spectral analysis of the variation in melanin content between light and dark colored skin and between a mole and healthy skin is conducted. Finally, the HinaLea hyperspectral chamber is used for the analysis of a wound caused by a vascular ulcer. The results of this study are manifold; first of all, a correct functioning of the hyperspectral chamber integrated to the slit lamp is observed for the analysis of the anterior segment of the eye. Some issues arise during the studying of the posterior segment due to the non-optimal compatibility between specialist’s lenses and slit lamp setup. For the dermatological analysis, there is a good ability of the hyperspectral system to discriminate not only light skin from dark skin, but also how a healthy and small mole in the 700 nm to 750 nm range shows different spectral characteristics to the surrounding skin. In the same way, during the analysis of wound a very good discriminatory ability of this system emerged respect to the healed one and healthy skin. Limitations and disadvantages were found for each of these areas. During the acquisitions performed in the three different fields, the primary problem revealed is the difficulty in obtaining a clean and sharp image for all wavelengths. Consequently, for an acquisition as fast as possible and not disturbed by motion artifacts, caused by the patient for the long duration of the examination, it is necessary to choose a limited spectral range characterized by specific wavelengths

Biomedical Applications of Hyperspectral Imaging in Medicine

DISCEPOLO, SILVIA
2020/2021

Abstract

Hyperspectral imaging is a hybrid modality that combines both spectroscopy and digital imaging [1]. Unlike other techniques, such as RGB, this method allows the continuous acquisition of a spectrum for each individual pixel, not just in the visible range [1]. The final data from this approach are presented in the form of a "hyperspectral cube" identified by two spatial and one spectral dimension [1] . Over time, this modality has been widely used in a number of sectors, such as agriculture, forensics, industry, medicine and disease diagnosis, due to its ability to discriminate and identify the spectral content of chemical compounds that make up a substance [1]. The main objective of this study is to verify how a hyperspectral system can be used in different biomedical fields. The first field examined is ophthalmology, in particular the combination of the Senop HSC-2 hyperspectral camera mounted on a Haag-Streit slit lamp eyepiece. Initially, a calibration procedure was performed using the Spectralon, appropriately inserted into the cavity of a mask simulating the human face. This first step made it possible to obtain the white reference target for all the subsequent acquisitions. Subsequently, discussing with a doctor specialized in ophthalmology, it was possible to understand whether this instrument could be valid as a medical device for the diagnosis and identification of ocular pathologies. The second field analyzed, using both the Senop-HSC-2 and HinaLea hyperspectral cameras, is dermatology. The spectral analysis of the variation in melanin content between light and dark colored skin and between a mole and healthy skin is conducted. Finally, the HinaLea hyperspectral chamber is used for the analysis of a wound caused by a vascular ulcer. The results of this study are manifold; first of all, a correct functioning of the hyperspectral chamber integrated to the slit lamp is observed for the analysis of the anterior segment of the eye. Some issues arise during the studying of the posterior segment due to the non-optimal compatibility between specialist’s lenses and slit lamp setup. For the dermatological analysis, there is a good ability of the hyperspectral system to discriminate not only light skin from dark skin, but also how a healthy and small mole in the 700 nm to 750 nm range shows different spectral characteristics to the surrounding skin. In the same way, during the analysis of wound a very good discriminatory ability of this system emerged respect to the healed one and healthy skin. Limitations and disadvantages were found for each of these areas. During the acquisitions performed in the three different fields, the primary problem revealed is the difficulty in obtaining a clean and sharp image for all wavelengths. Consequently, for an acquisition as fast as possible and not disturbed by motion artifacts, caused by the patient for the long duration of the examination, it is necessary to choose a limited spectral range characterized by specific wavelengths
2020
2021-07-19
Biomedical Applications of Hyperspectral Imaging in Medicine
File in questo prodotto:
File Dimensione Formato  
Silvia D. Tesi - 1.pdf

Open Access dal 20/07/2023

Descrizione: Tesi con frontespizio non firmato
Dimensione 10.46 MB
Formato Adobe PDF
10.46 MB Adobe PDF Visualizza/Apri

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12075/379