| Title |
Algoritmas gyvų ląstelių klasifikavimui ir stebėsenai giliųjų neuroninių tinklų pagrindu / |
| Translation of Title |
Algorithm for live cell classification and monitoring based on deep neural networks. |
| Authors |
Dirvanauskas, Darius |
| Full Text |
|
| Pages |
46 |
| Keywords [eng] |
embryo image analysis ; deep learning ; image classification ; spatial model reconstruction |
| Abstract [eng] |
Every sixth couple in the world has difficulty having children. There are about 50 thousand such families in Lithuania and their number is growing every year. Some of these families are treated with an in vitro method for growing an embryo. Time-lapse microscopy has provided new tools for embryo image inspection. These machines are used for continuous monitoring of embryos in different layers and capturing images of embryo evolution, measuring embryo’s development stage duration and inspecting embryo cell shape without removing embryos from their growth environment. The aim of dissertation is to develop a computerized method, which would predict the embryo growth stage and reconstruct its spatial model from the layer images. We propose a hybrid embryo classification model which is applicable for constant embryo monitoring and stage prediction. Our offered deep learning method can be used for embryo spatial model reconstruction from different layer images. The created tool for embryo spatial image visual evaluation gives embryologists new possibilities for embryo analysis. |
| Dissertation Institution |
Kauno technologijos universitetas. |
| Type |
Summaries of doctoral thesis |
| Language |
Lithuanian |
| Publication date |
2021 |
