Politecnico di Torino (logo)

A skeletonization approach for the evaluation of vascular complexity using in-vitro phantoms and 3D LED-based photoacoustic images

Roberta Bruschetta

A skeletonization approach for the evaluation of vascular complexity using in-vitro phantoms and 3D LED-based photoacoustic images.

Rel. Filippo Molinari, Kristen Mariko Meiburger. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Biomedica, 2019

PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (10MB) | Preview

Vasculature analysis is a fundamental aspect in diagnosis, treatment, outcomes evaluation and follow-up in several medical fields. Since vessels branching, maturation or quiescence depend on microenvironment and angiogenic signals, vascular attributes are deeply affected by most of the diseases, from a simple cold to the tumours. Moreover, it was demonstrated that efficient therapies administration leads to attributes normalization. Therefore, vascular network characterization can be a powerful means for earlier pathologies revealing and for their monitoring. For this reason, the development of a non-invasive and quantitative method for the evaluation of blood vessels complexity is a very important issue. Many imaging techniques can be used for visualizing blood vessels, but their application is limited by problems such as high costs, need of contrast agents and use of ionizing radiations. Photoacoustic imaging (PAI) is an emerging hybrid modality, which makes use of optical excitation and ultrasound detection. This non-invasive and non-ionizing technique combines the qualities of both the methods: the contrast and the spectral specificity of optical imaging and the high penetration depth and the spatial resolution of acoustic imaging. The optical and radio-frequency waves used in this technique, enable the possibility to use endogenous agents such as haemoglobin thanks to its absorption properties. Moreover, the most recent LED-based systems are economic, safe and portable. Starting from biomedical images, many techniques have been developed to extract and represent vessels structure. They differ among them for the pre-processing steps, the computational time, the accuracy and the visual quality in results showing. Skeletonization is a very efficient method to evaluate vasculature architecture and has been used with 3D contrast enhanced ultrasound images to characterize thyroid nodular vessels and tumour vasculature in vivo studies. The aim of this thesis is the design and manufacturing of in-vitro vessel phantoms to evaluate the feasibility of using 3D LED-based photoacoustic images for vascular complexity quantification using a skeletonization algorithm. In particular, three in-vitro agar phantoms with different vessels complexity were designed using the CAD Software Solidworks and realised with the use of a 3D printer. 3D images acquisition was performed with a LED-based photoacoustic imaging system of Prexion Corporation. Then, the images obtained were segmented and the three-dimensional volumes were reconstructed. Automatic segmentation performances were evaluated through a comparison with a manual segmentation. From the reconstructed volumes, a 3D skeleton, based on medial axis extraction, was calculated and vascular parameters were measured. In particular, three tortuosity and three morphological parameters were computed: number of vascular trees (NT), vascular density (VD), number of branches (NB), distance metric (DM), inflection count metric (ICM) and sum of angles metric(SOAM). In addition, mean radius (MR) was calculated. Outcomes were compared with the ground truth obtained converting the original models from Solidworks to MATLAB. Analysing results, it's possible to conclude the validity of the skeletonization from 3D LED-based photoacoustic images as approach for vascular complexity quantification.

Relators: Filippo Molinari, Kristen Mariko Meiburger
Academic year: 2018/19
Publication type: Electronic
Number of Pages: 105
Corso di laurea: Corso di laurea magistrale in Ingegneria Biomedica
Classe di laurea: New organization > Master science > LM-21 - BIOMEDICAL ENGINEERING
Aziende collaboratrici: UNSPECIFIED
URI: http://webthesis.biblio.polito.it/id/eprint/11363
Modify record (reserved for operators) Modify record (reserved for operators)