| 講演抄録/キーワード |
| 講演名 |
2022-07-08 14:40
[ショートペーパー]Automatic Cell Tracking in 4D Arabidopsis Root Microscopic Images Using Genetic Algorithm ○Yu Song・Yutaro Iwamoto(Ritsumeikan Univ.)・Tatsuaki Goh(NAIST)・Yohei Kondo(ExCELLS)・Keiji Nakajima(NAIST)・Yen-wei Chen(Ritsumeikan Univ.) MI2022-34 |
| 抄録 |
(和) |
(まだ登録されていません) |
| (英) |
Single-particle tracking (SPT) is a crucial progress to discover unknown principles of plants. Through accurate tracking of each particle on time-lapse images of biological samples taken by microscopes, many hypotheses in biological field can thus be verified. A commonly used method for tracking is called linear assignment problem (LAP) tracker. LAP tracker is well suited for particles undergoing Brownian motion. However, in our task, we are going to track the cortex cells of Arabidopsis root on 3D time-lapse (4D) images, where cells of Arabidopsis root are densely arranged. This makes LAP tracker ineffective when considering the distances costs of links. To realize accurate SPT on this task, we propose an automatic tracking method using genetic algorithm (GA). Based on the shape information we know about Arabidopsis roots, a 3D image at a frame can be viewed as 8 lines of cells. Therefore, we turn the SPT into the tracking on each of 8 lines in the time axis. Due to the elongation of cells, each line is not a straight line but more like a curve. For each frame, we use a GA to select an optimal plane to project cells. The projected cells are well separated to 8 regions, where each region represents a line of cells. After clustering the projected cells into 8 lines, we do the tracking based on 8 lines of cells. Finally, we operate SPT of each cortex cell. Through experiments, the proposed method realizes stable and accurate tracking of Arabidopsis root cortex cells. |
| キーワード |
(和) |
/ / / / / / / |
| (英) |
Arabidopsis Root / Cell Tracking / Genetic Algorithm / Single-Particle Tracking / / / / |
| 文献情報 |
信学技報, vol. 122, no. 98, MI2022-34, pp. 10-15, 2022年7月. |
| 資料番号 |
MI2022-34 |
| 発行日 |
2022-07-01 (MI) |
| ISSN |
Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
| PDFダウンロード |
MI2022-34 |
|