Validation-data Generation for Brightfield Microscopy Cell Tracking using Fluorescence Samples
This work focuses on the use of fluorescent cancer cell images as data to validate the results obtained in segmenting brightfield cancer cell images, as the latter’s current validation consists of manual annotation of cells in the original images. The procedure uses pattern recognition and starts wi...
| Autores Principales: | Quinde-Cobos, Patricia, Quirós, Steve, Siles-Canales, Francisco |
|---|---|
| Formato: | Artículo |
| Idioma: | Inglés |
| Publicado: |
Editorial Tecnológica de Costa Rica (entidad editora)
2020
|
| Materias: | |
| Acceso en línea: |
https://revistas.tec.ac.cr/index.php/tec_marcha/article/view/5083 https://hdl.handle.net/2238/12076 |
| Sumario: |
This work focuses on the use of fluorescent cancer cell images as data to validate the results obtained in segmenting brightfield cancer cell images, as the latter’s current validation consists of manual annotation of cells in the original images. The procedure uses pattern recognition and starts with preprocessing the fluorescent samples to ensure cell detection, focused on area and intensity value. As the fluorescent images are segmented, each cell’s nucleus is detected and counted, with a high success rate as each nucleus’s contour was detected with its original shape. As each image’s density is calculated, they can be clustered according to their density value and used for cell detection in brightfield samples. |
|---|