Tumor Motion Tracking in Liver Ultrasound Images Using Mean Shift and Active Contour

TL;DR

This paper introduces a novel two-step method combining mean shift and active contour models for accurate tumor motion tracking in liver ultrasound sequences, enhancing reliability and boundary detection.

Contribution

The paper presents a new combined approach using mean shift with multiple features and active contours for improved liver tumor tracking in ultrasound images.

Findings

Reliable tumor tracking demonstrated in experiments

Effective boundary detection of liver tumors

Enhanced accuracy over existing methods

Abstract

In this paper we present a new method for motion tracking of tumors in liver ultrasound image sequences. Our algorithm has two main steps. In the first step, we apply mean shift algorithm with multiple features to estimate the center of the target in each frame. Target in the first frame is defined using an ellipse. Edge, texture, and intensity features are extracted from the first frame, and then mean shift algorithm is applied to each feature separately to find the center of ellipse related to that feature in the next frame. The center of ellipse will be the weighted average of these centers. By using mean shift actually we estimate the target movement between two consecutive frames. Once the correct ellipsoid in each frame is known, in the second step we apply the Dynamic Directional Gradient Vector Flow (DDGVF) version of active contour models, in order to find the correct boundary of tumors. We sample a few points on the boundary of active contour then translate those points based on the translation of the center of ellipsoid in two consecutive frames to determine the target movement. We use these translated sample points as an initial guess for active contour in the next frame. Our experimental results show that, the suggested method provides a reliable performance for liver tumor tracking in ultrasound image sequences.