Cell Tracking in C. elegans with Cell Position Heatmap-Based Alignment and Pairwise Detection

TL;DR

This paper introduces cell position heatmap-based non-rigid alignment with test-time fine-tuning, which can warp the detected points to near the positions at the next frame and proposes a pairwise detection method, which uses the information of detection results at the previous frame for detecting cells at the current frame.

Abstract

3D cell tracking in a living organism has a crucial role in live cell image analysis. Cell tracking in C. elegans has two difficulties. First, cell migration in a consecutive frame is large since they move their head during scanning. Second, cell detection is often inconsistent in consecutive frames due to touching cells and low-contrast images, and these inconsistent detections affect the tracking performance worse. In this paper, we propose a cell tracking method to address these issues, which has two main contributions. First, we introduce cell position heatmap-based non-rigid alignment with test-time fine-tuning, which can warp the detected points to near the positions at the next frame. Second, we propose a pairwise detection method, which uses the information of detection results at the previous frame for detecting cells at the current frame. The experimental results demonstrate the effectiveness of each module, and the proposed method achieved the best performance in comparison.

Figures (5)

• Figure 1: Example of large movement issue. Left: detection results at $t$ (blue '+'). Right: those at $t+1$ (orange '+').
• Figure 2: Overview of our method. First, given two consecutive frames of time-lapse images as inputs, a detection module estimates cell position at $t+1$ using cell position information at $t$. Second, the alignment module warps cell positions at $t$ to $t+1$. Finally, we associate the warped cell positions at $t$ and those at $t+1$.
• Figure 3: Example of tracking results of (a) Ground truth, (b)w/o registration, and (c) our method. The same cell trajectory has the same color and ID. '$\diamond$' indicates a switching error.
• Figure 4: Examples of warped positions by cell position heatmap-based alignment. Left: original displacement. Right: displacement after warping. blue is the position at $t$, green is the warped position from $t$ to $t+1$, red is that at $t+1$, and the line is the displacement.
• Figure 5: Examples of effectiveness of pairwise detection. (a)Original image, (b)Ground truth, (c)Single detection, and (d)Pairwise detection.