Diagnosis and Follow-up of Oncologic and Vascular Pathology of Surgical Interest: The Role of Diagnostic Imaging

Diagnostic imaging is pivotal in the clinical and surgical management of patients, notably it is paramount in the decision of surgical option and characterization of vessels. The development of semi-automatic software for the analysis of volumetric radiological images is currently highly investigated because it increases the sensitivity and the diagnostic accuracy of the radiological test. Moreover, these software reduce the inter-observer variability. Notably, the semi-automatic analysis of images shows high potentiality in the diagnosis and staging of tumor from lung and head-neck compartment, as well as for characterization of vessels Advanced lung cancer can be treated by neoadjuvant chemotherapy with the aim of reducing the neoplastic mass and, thereafter, consider therapeutic surgical option. Diagnostic imaging is useful to determine the response to neoadjuvant chemotherapy by means of objective parameters such as quantification of volume decrease even after the first administration. Furthermore, the early diagnosis of lung cancer is fostered by screening by low-dose computed tomography (LDCT). LDCT has high sensitivity (e.g. >50% prevalence of lung nodule) but low sensitivity. To increase specificity, size thresholds were defined (based on diameter and/or volume) as well as parameters to quantify the evolution (e.g. volume doubling time, DVT, and mass doubling time, MDT), with the aim of categorizing nodules with higher likelihood of malignancy. In lung cancer screening by LDCT, semi-automatic software are paramount for optimization of the management algorithm. Oral cancer are diagnosed by clinical assessment and are verified by biopsy, which is relatively easy because of the superficial location of these tumors. Nevertheless, the pre-surgical assessment needs specific characterization of deep infiltration and lymph node involvement. Diagnostic imaging plays a key role in this task. Notably, bone infiltration is paramount for surgical planning. The combination of magnetic resonance imaging (MRI) and computed tomography (CT) provides the optimal diagnostic accuracy that allows maximum reduction of surgical invasiveness and radical treatment. The features of large and small vessels can be described by diagnostic imaging with non-invasive technique, both for panoramic and detailed description. Furthermore, diagnostic imaging is also accurate to assess temporal evolution of vascular features. Subjects with hepatic failure and portal hypertension may show more signs of disease severity, such as early abnormalities of pulmonary vascularization before the onset of portopulmonary hypertension and hepatopulmonary syndrome. The characteristics of small pulmonary arteries before and after liver transplant might represent a prognostic indicator. Notably, the numerousness of small pulmonary arteries can be quantified non-invasively by semi-automatic software by means of geometrical parameters that reflect the anatomical characteristics of arteries. In this thesis, experimental scientific studies will be reported, notably concerning the aforementioned field of interest, with particular discussion of the utility of software for semi-automatic analysis of imaging datasets.