Advanced Spectral Analysis for Automatic Echographic Monitoring of an Evolving Tumour Mass

In the last decades, minimally invasive technologies have experienced a significant diffusion in various clinical specialties, finding their application in diagnosis and therapy of acute and chronic diseases. In particular early and effective cancer diagnosis is amongst the goals of recent studies focused on the improvement of innovative medical instruments. Moreover, several research projects are currently operating on the optimization of novel diagnostic paradigms, seeking to minimal exposure of the patient to ionizing radiations, considered highly health threatening. Therefore, the aim of this study was to assess the accuracy of a prototypal software algorithm for advanced spectral analysis on echographic images (RULES, ELEN SpA, Florence, Italy) in the automatic segmentation of a simulated tumour mass with variable physical conditions (position, shape, pressure exerted by the surrounding tissues). Different phantoms were used to mimic specific pathological conditions: an early stage cancer and a hard tumour mass. Specificity and sensitivity of the procedure were calculated throughout the condition variation cycle for each model and compared. Results demonstrated the possibility of selecting an appropriate configuration of the algorithm to perform an automatic echographic monitoring of a tissue mass of given mechanical properties while it experiences variations of position, shape and pressure exerted by the surrounding tissues. Clinical implications of the reported findings could be crucial for management of cancer patients and for disease assessment in absence of contrast agent injection.