logo SBA

ETD

Archivio digitale delle tesi discusse presso l’Università di Pisa

Tesi etd-11202019-001834


Tipo di tesi
Tesi di laurea magistrale LM6
Autore
GEGA, EVISJOLA
URN
etd-11202019-001834
Titolo
MRI Radiomics of salivary glands tumors
Dipartimento
RICERCA TRASLAZIONALE E DELLE NUOVE TECNOLOGIE IN MEDICINA E CHIRURGIA
Corso di studi
MEDICINA E CHIRURGIA
Relatori
relatore Prof. Neri, Emanuele
Parole chiave
  • Tumors
  • Salivary
  • Radiomics
Data inizio appello
10/12/2019
Consultabilità
Non consultabile
Data di rilascio
10/12/2089
Riassunto
Salivary gland tumors are presented as painless enlarging masses. The majority of them is located in the parotid gland and 80% of the tumors are benign. The principal problem in their management lies in the difficulty to distinguish benign from malignant tumors, or different types of them for e.g. pleomorphic adenoma from Warthin’s tumor. Investigations such as Fine Needle Aspiration Cytology (FNAC), and Magnetic Resonance Imaging (MRI) scans offer valuable information, but most cases will require surgical excision as a means of coming to a definitive diagnosis. Our aim was to investigate textural features of major salivary gland tumors as a potential additional pre-surgical tool of lesion discrimination between the two most common benign neoplasms histotypes: pleomorphic adenoma and Warthin’s tumor. This study is based on retrospective evaluation of pre-surgical MRI scans of 20 patients with major salivary glands tumors (MSGT). There were 20 benign tumors: 10 pleomorphic adenomas (50%) and 10 Warthin’s tumors (50%). Histology results were available for all tumors. A region of interest (ROI) was delimited manually where tumor was noticeable in axial T2 weighted images; 29 radiomic features have been extracted from the ROI by QUIBIM SL, a dedicated company that is qualified in studying quantitative imaging biomarkers in medicine. Quantitative data were described by mean and standard deviation with dichotomous variables [(malignancy: no, yes); (Warthin’s tumor: no, yes); (Pleomorphic Adenoma: no, yes)]. T-test (two-tailed) was used and subsequently the features statistically more significant (p < 0,05) were: AutoCorrelation (p<0,006); Cluster Prominence (p=0,028); Sum Average (p=0,002); Skewness (p=0.012).
File