A Study on Segmentation Architectures for Thyroid Scintigram Delineation
- 1Universitat zu Lubeck, Ratzeburger Allee 160, 23562Lubeck, Germany
Abstract
Automated thyroid segmentation in scintigraphic imaging is critical for functional assessment but remains challenging due to low resolution, photon noise, diffuse boundaries, and class imbalance. This study compares four deep learning architectures - U-Net, U-Net++, TransUNet, and Swin-UNet - trained and evaluated on a public thyroid scan dataset (189 training, 23 validation, 23 evaluation studies) under identical conditions. TransUNet achieved the highest accuracy, slightly outperforming U-Net while matching its boundary precision. U-Net followed closely with superior efficiency, lowest latency and highest throughput, making it ideal for realtime deployment. Swin-UNet showed competitive but lower performance on fine structures, while U-Net++ exhibited substantially reduced accuracy and high variability. These results indicate, that increased architectural complexity does not necessarily improve performance. While transformer-based models offer modest gains through global context modeling, convolutional architectures remain robust and efficient. U-Net provides the best accuracy-efficiency trade-off;
Related
- Research on precise segmentation model of thyroid nodules based on deep learning
- Adversarial robustness of deep learning-based thyroid nodule segmentation in ultrasound
- Nanoprobe‐Enabled Precision Imaging and Panoramic Boundary Delineation in Retinoblastoma
- Ferumoxytol-Enhanced Cardiac Magnetic Resonance for Delineation of Hyperacute Intramyocardial Hemorrhage with Ex Vivo Validation
- Utilization of 68Ga-FAPI PET/CT improves the gross tumor volume delineation in nasopharyngeal carcinoma within a safe radiation dose to organs at risk
BibTeX
@article{moritz2026study,
title = {{A Study on Segmentation Architectures for Thyroid Scintigram Delineation}},
author = {Mau Moritz A.},
journal = {Current Directions in Biomedical Engineering},
year = {2026},
month = oct,
doi = {10.1515/cdbme-2026-0185},
url = {https://doi.org/10.1515/cdbme-2026-0185},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - A Study on Segmentation Architectures for Thyroid Scintigram Delineation
AU - Mau Moritz A.
PY - 2026
DA - 2026/10/26
JO - Current Directions in Biomedical Engineering
DO - 10.1515/cdbme-2026-0185
UR - https://doi.org/10.1515/cdbme-2026-0185
AB - Automated thyroid segmentation in scintigraphic imaging is critical for functional assessment but remains challenging due to low resolution, photon noise, diffuse boundaries, and class imbalance. This study compares four deep learning architectures - U-Net, U-Net++, TransUNet, and Swin-UNet - trained and evaluated on a public thyroid scan dataset (189 training, 23 validation, 23 evaluation studies) under identical conditions. TransUNet achieved the highest accuracy, slightly outperforming U-Net while matching its boundary precision. U-Net followed closely with superior efficiency, lowest latency and highest throughput, making it ideal for realtime deployment. Swin-UNet showed competitive but lower performance on fine structures, while U-Net++ exhibited substantially reduced accuracy and high variability. These results indicate, that increased architectural complexity does not necessarily improve performance. While transformer-based models offer modest gains through global context modeling, convolutional architectures remain robust and efficient. U-Net provides the best accuracy-efficiency trade-off;
SN - 2364-5504
ER -