Javascript must be enabled for the correct page display

Finding Balance: A Series of U-Net Models for Image Segmentation of Overlapping Organoids

Radu, Stefania (2022) Finding Balance: A Series of U-Net Models for Image Segmentation of Overlapping Organoids. Bachelor's Thesis, Artificial Intelligence.

[img] Text
toestemming.pdf
Restricted to Registered users only

Download (148kB)
[img]
Preview
Text
Thesis__Finding_balance.pdf

Download (4MB) | Preview

Abstract

The interest in automatically analyzing biomedical images increased in the past years, as an accurate localization and segmentation of organoids can help with the early detection of malignancies and predict diseases, such as cancer. The morphometric appearances of these images and the high level of overlapping in the organoids make the segmentation task challenging. This paper studied a simple U-Net and also proposes a Dual U-Net model with a shared encoder and two decoders, one for binary segmentation of the mask and one for the multi-class segmentation of overlaps. A significant addition to the U-net are the residual-atrous skip connections which reduce the semantic gap between the encoder and the decoder. The issue of high imbalance between the classes is addressed using a combination between the Focal Loss and the Focal Tversky Loss, which significantly improved the performance of the model. Ten networks were trained on more than 17,000 images with overlapping and non-overlapping organoids and obtained promising results. When tested on 88 new images, the final models achieved an F1 score of 0.83 for the mask channel and 0.43 for the overlapping channel. The Jaccard Index was 0.72 for the mask and 0.34 for the overlap.

Item Type: Thesis (Bachelor's Thesis)
Supervisor name: Schomaker, L.R.B. and Haja, A.
Degree programme: Artificial Intelligence
Thesis type: Bachelor's Thesis
Language: English
Date Deposited: 20 Jul 2022 11:35
Last Modified: 20 Sep 2023 09:42
URI: https://fse.studenttheses.ub.rug.nl/id/eprint/27995

Actions (login required)

View Item View Item