Automatic Segmentation of the Inferior Vena Cava from M-mode Ultrasound Images

Systemic venous congestion (SVC) is a critical factor in patients with heart and kidney failure, a condition known as cardiorenal syndrome (CRS).The Venous Excess Ultrasound grading System (VExUS) protocol represents a significant advance in the assessment of SVC using pointof-care ultrasound.The first step of this protocol consists of determining the maximum diameter of the inferior vena cava (IVC) in M-mode ultrasound images, a task currently performed manually.This work presents a method for the automatic segmentation and characterization of the IVC.The method was evaluated on a set of 20 images belonging to 13 patients with CRS.The maximum diameter determined automatically was compared with that measured during clinical practice.Results showed a mean error of -0.015 0.318 cm, with a Spearman correlation of s = 0.850 and a Lin coefficient of c = 0.864.Therefore, the proposed method achieves a reasonably accurate assessment of the maximum diameter of the IVC within the context of the VExUS protocol.