Test of a new ultrasound system for monitoring the edges of the inferior vena cava

The size and pulsatility of the inferior vena cava (IVC) allow the extraction of parameters that are difficult to estimate but essential for a correct diagnosis, such as the right atrial pressure or the blood volume status of patients. No standardized method exists for the analysis of this vessel. One of the most widely used techniques is M-mode ultrasound. However, this only allows the IVC scan to be viewed along a fixed direction and therefore cannot refer to the same section of the vessel as the vein moves during respiratory and cardiac activity. To overcome these limitations, a new software has been developed by Vein Image Processing for Edge Rendering (VIPER). This software allows the identification of vein edges and displacements, both along the transverse and longitudinal planes, by means of a semi-automatic analysis of B-mode ultrasound videos. The aim of this thesis is to evaluate the ability of this software to correctly segment IVC in different working conditions. The analysis was carried out by means of a comparison with manual segmentation masks. In particular, the research assessed the software's dependence on user input, the reliability and the resilience to occasional events. Finally, the algorithm was tested on ultrasound videos acquired from doctors. To objectively analyze the results obtained, the variability of manual segmentations was also evaluated. Along the transverse plane, the software is independent of user input, while greater sensitivity to input was found in the longitudinal plane. The software can correctly segment the vessel both in static conditions and during vein movements. The main limitations are the sensitivity to noise, the inability to recognize the disappearance of the vessel and the difficulty in segmenting small veins. Nevertheless, the good performance achieved shows that this software is an excellent tool to overcome the limits currently existing in the clinical field for the analysis of IVC pulsatility.