Enhanced DeepLab Based Nerve Segmentation with Optimized Tuning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Thomas, Akhil John, Boerkamp, Christiaan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913948343730176
author Thomas, Akhil John
Boerkamp, Christiaan
author_facet Thomas, Akhil John
Boerkamp, Christiaan
contents Nerve segmentation is crucial in medical imaging for precise identification of nerve structures. This study presents an optimized DeepLabV3-based segmentation pipeline that incorporates automated threshold fine-tuning to improve segmentation accuracy. By refining preprocessing steps and implementing parameter optimization, we achieved a Dice Score of 0.78, an IoU of 0.70, and a Pixel Accuracy of 0.95 on ultrasound nerve imaging. The results demonstrate significant improvements over baseline models and highlight the importance of tailored parameter selection in automated nerve detection.
format Preprint
id arxiv_https___arxiv_org_abs_2507_13394
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Enhanced DeepLab Based Nerve Segmentation with Optimized Tuning
Thomas, Akhil John
Boerkamp, Christiaan
Image and Video Processing
Computer Vision and Pattern Recognition
Nerve segmentation is crucial in medical imaging for precise identification of nerve structures. This study presents an optimized DeepLabV3-based segmentation pipeline that incorporates automated threshold fine-tuning to improve segmentation accuracy. By refining preprocessing steps and implementing parameter optimization, we achieved a Dice Score of 0.78, an IoU of 0.70, and a Pixel Accuracy of 0.95 on ultrasound nerve imaging. The results demonstrate significant improvements over baseline models and highlight the importance of tailored parameter selection in automated nerve detection.
title Enhanced DeepLab Based Nerve Segmentation with Optimized Tuning
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2507.13394