Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhang, Haoran, Yang, Jianlong, Zhang, Jingqian, Zhao, Shiqing, Zhang, Aili
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2306.04512
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909062056116224
author Zhang, Haoran
Yang, Jianlong
Zhang, Jingqian
Zhao, Shiqing
Zhang, Aili
author_facet Zhang, Haoran
Yang, Jianlong
Zhang, Jingqian
Zhao, Shiqing
Zhang, Aili
contents Nonuniform rotational distortion (NURD) correction is vital for endoscopic optical coherence tomography (OCT) imaging and its functional extensions, such as angiography and elastography. Current NURD correction methods require time-consuming feature tracking or cross-correlation calculations and thus sacrifice temporal resolution. Here we propose a cross-attention learning method for the NURD correction in OCT. Our method is inspired by the recent success of the self-attention mechanism in natural language processing and computer vision. By leveraging its ability to model long-range dependencies, we can directly obtain the correlation between OCT A-lines at any distance, thus accelerating the NURD correction. We develop an end-to-end stacked cross-attention network and design three types of optimization constraints. We compare our method with two traditional feature-based methods and a CNN-based method, on two publicly-available endoscopic OCT datasets and a private dataset collected on our home-built endoscopic OCT system. Our method achieved a $\sim3\times$ speedup to real time ($26\pm 3$ fps), and superior correction performance.
format Preprint
id arxiv_https___arxiv_org_abs_2306_04512
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Cross-attention learning enables real-time nonuniform rotational distortion correction in OCT
Zhang, Haoran
Yang, Jianlong
Zhang, Jingqian
Zhao, Shiqing
Zhang, Aili
Image and Video Processing
Computer Vision and Pattern Recognition
Medical Physics
Nonuniform rotational distortion (NURD) correction is vital for endoscopic optical coherence tomography (OCT) imaging and its functional extensions, such as angiography and elastography. Current NURD correction methods require time-consuming feature tracking or cross-correlation calculations and thus sacrifice temporal resolution. Here we propose a cross-attention learning method for the NURD correction in OCT. Our method is inspired by the recent success of the self-attention mechanism in natural language processing and computer vision. By leveraging its ability to model long-range dependencies, we can directly obtain the correlation between OCT A-lines at any distance, thus accelerating the NURD correction. We develop an end-to-end stacked cross-attention network and design three types of optimization constraints. We compare our method with two traditional feature-based methods and a CNN-based method, on two publicly-available endoscopic OCT datasets and a private dataset collected on our home-built endoscopic OCT system. Our method achieved a $\sim3\times$ speedup to real time ($26\pm 3$ fps), and superior correction performance.
title Cross-attention learning enables real-time nonuniform rotational distortion correction in OCT
topic Image and Video Processing
Computer Vision and Pattern Recognition
Medical Physics
url https://arxiv.org/abs/2306.04512