five

Lineages

收藏
DataCite Commons2025-05-01 更新2024-08-18 收录
下载链接:
https://figshare.com/articles/dataset/Lineages/21756212/2
下载链接
链接失效反馈
官方服务:
资源简介:
Each of the .zip files in this dataset correspond to one of the microcolonies in the cib.zip file of the data repository of van Vliet et al (2017): https://zenodo.org/record/268921. Files generally correspond to the standard FAST formats. However, to facilitate the benchmarking process, several copies of each type of file exist containing datasets processed in different ways. In general, these can be split into <em>ground truth datasets</em> (fully manually corrected), <em>Corrected vs. automated analyses </em>(differ in whether inputs to the modules were manually corrected or not) and <em>all features vs. centroids only </em>(differ in whether all features or just the centroids were used by the tracking module. Specifically, each module is associated with the following files/directories: <br> Segmentations Original_Segmentations: Segmentations as output directly by the FAST segmentation module. Corrected_Segmentations: Segmentations corrected by hand. Feature extraction CellFeatures_AutoSeg.mat: Features extracted based on Original_Segmentations. CellFeatures_ManualSeg.mat: Features extracted based on Corrected_Segmentations. Tracking Tracks_AutoSeg_Auto.mat: Automatically generated tracks output by the Tracking module with length, width, fluorescence and position as features and Original_Segmentations as the raw segmentations. Tracks_AutoSeg_Auto_CentroidsOnly.mat: Automatically generated tracks output by the Tracking module with only positional coordinates as features and Original_Segmentations as the raw segmentations. Tracks_AutoSeg_Corrected.mat: Manually corrected ground truth for the tracks associated with Original_Segmentations. Generated with the track correction GUI. Tracks_ManualSeg_Auto.mat: Automatically generated tracks output by the Tracking module with length, width, fluorescence and position as features and Corrected_Segmentations as the raw segmentations. Tracks_ManualSeg_Corrected.mat: Manually corrected ground truth for the tracks associated with Corrected_Segmentations. Generated with the track correction GUI. Divisions Tracks_GoldStandardDivisions.mat: Manually corrected ground truth for division assignment, based on Tracks_ManualSeg_Corrected.mat. Tracks_AutoStreamDivisions.mat: Automatically generated output of the division detection module using Tracks_AutoSeg_Auto.mat as input. Tracks_AutoStreamDivisions_CentroidsOnly.mat: Automatically generated output of the division detection module using Tracks_AutoSeg_Auto_CentroidsOnly.mat as input. Tracks_ManualStreamDivisions.mat: Automatically generated output of the division detection module using Tracks_ManualSeg_Corrected.mat as input. In general, to use any one of these as an input to a later module, just rename the corresponding file to the FAST standard (e.g. rename Tracks_AutoSeg_Auto.mat to Tracks.mat to view the associated tracks in the overlays module). In some cases, you will also need to make sure that the Segmentation directory is filled with the corresponding segmentations (either Original or Corrected, depending on whether you are looking at the auto or the manual streams).

本数据集内的每个.zip文件均对应van Vliet等人(2017)的数据仓库文件cib.zip中的一个微菌落,数据仓库链接为:https://zenodo.org/record/268921。文件大多遵循标准FAST格式。为便于基准测试,每种类型的文件均提供多份副本,以对应不同的预处理方式。总体而言,这些文件可分为三类:**真值数据集(ground truth datasets)**(已完成全手动校正)、**校正版与自动化分析版**(二者的区别在于模块输入是否经过手动校正),以及**全特征版与仅质心版**(二者的区别在于跟踪模块是否使用全部特征,还是仅使用质心数据)。 具体而言,每个模块关联以下文件或目录: 1. 分割模块(Segmentations) - Original_Segmentations:FAST分割模块直接输出的原始分割结果 - Corrected_Segmentations:经人工手动校正的分割结果 2. 特征提取 - CellFeatures_AutoSeg.mat:基于Original_Segmentations提取的细胞特征文件 - CellFeatures_ManualSeg.mat:基于Corrected_Segmentations提取的细胞特征文件 3. 跟踪模块(Tracking) - Tracks_AutoSeg_Auto.mat:跟踪模块自动生成的轨迹文件,以轨迹长度、宽度、荧光强度与位置为特征,以Original_Segmentations作为原始分割输入 - Tracks_AutoSeg_Auto_CentroidsOnly.mat:跟踪模块自动生成的轨迹文件,仅以位置坐标为特征,以Original_Segmentations作为原始分割输入 - Tracks_AutoSeg_Corrected.mat:关联原始分割结果的轨迹手动校正真值文件,通过轨迹校正GUI生成 - Tracks_ManualSeg_Auto.mat:跟踪模块自动生成的轨迹文件,以轨迹长度、宽度、荧光强度与位置为特征,以Corrected_Segmentations作为原始分割输入 - Tracks_ManualSeg_Corrected.mat:关联校正后分割结果的轨迹手动校正真值文件,通过轨迹校正GUI生成 4. 细胞分裂检测模块(Divisions) - Tracks_GoldStandardDivisions.mat:基于Tracks_ManualSeg_Corrected.mat生成的手动校正细胞分裂分配真值文件 - Tracks_AutoStreamDivisions.mat:以Tracks_AutoSeg_Auto.mat为输入,由分裂检测模块自动生成的输出结果文件 - Tracks_AutoStreamDivisions_CentroidsOnly.mat:以Tracks_AutoSeg_Auto_CentroidsOnly.mat为输入,由分裂检测模块自动生成的输出结果文件 - Tracks_ManualStreamDivisions.mat:以Tracks_ManualSeg_Corrected.mat为输入,由分裂检测模块自动生成的输出结果文件 一般而言,若要将任意一个文件作为后续模块的输入,只需将对应文件重命名为FAST标准文件名即可(例如将Tracks_AutoSeg_Auto.mat重命名为Tracks.mat,即可在叠加模块中查看相关轨迹)。部分场景下,你还需确保分割目录中填入了对应的分割结果(根据自动化流程或手动流程,选择Original_Segmentations或Corrected_Segmentations)。
提供机构:
figshare
创建时间:
2023-07-03
搜集汇总
数据集介绍
main_image_url
以上内容由遇见数据集搜集并总结生成
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作