Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Decoding the epithelial-stromal interactome in allergic rhinitis through single-cell multi-omics integration.

The Journal of allergy and clinical immunology·2026
Same author

Corrigendum to "Aspirin mitigates sepsis-associated encephalopathy by modulating ICAM-1 and MMP-9 signaling: Evidence from clinical cohorts and mechanistic experiments" [Experimental Neurology 403 (2026) 115813].

Experimental neurology·2026
Same author

Clinical research progress and challenges of vaccine for human papillomavirus-associated cancers.

Human vaccines & immunotherapeutics·2026
Same author

Outdoor high-precision 3D dense mapping system based on stereo visual SLAM.

Scientific reports·2026
Same author

Aspirin mitigates sepsis-associated encephalopathy by modulating ICAM-1 and MMP-9 signaling: Evidence from clinical cohorts and mechanistic experiments.

Experimental neurology·2026
Same author

Envisioning population-scale immune multi-omics atlas projects.

Clinical and translational medicine·2026

Related Experiment Video

Updated: Aug 14, 2025

Lesion Explorer: A Video-guided, Standardized Protocol for Accurate and Reliable MRI-derived Volumetrics in Alzheimer's Disease and Normal Elderly
12:50

Lesion Explorer: A Video-guided, Standardized Protocol for Accurate and Reliable MRI-derived Volumetrics in Alzheimer's Disease and Normal Elderly

Published on: April 14, 2014

40.3K

Local Axial Scale-Attention for Universal Lesion Detection.

Chuanyu Liu1, Yonghong Hou2, Pengyu Zhao1

  • 1School of Electrical and Information Engineering, Tianjin University, Tianjin, 300072, China.

Journal of Digital Imaging
|January 17, 2023
PubMed
Summary

This study introduces a novel scale-attention module to improve universal lesion detection (ULD) in CT scans, significantly reducing false positives by mimicking radiologist techniques. The new method enhances accuracy in identifying small, similar lesions, outperforming existing approaches.

Keywords:
Axial pixelsScale-attentionUniversal lesion detection

More Related Videos

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
10:25

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

Published on: September 25, 2019

48.2K
In vivo Structural Assessments of Ocular Disease in Rodent Models using Optical Coherence Tomography
07:44

In vivo Structural Assessments of Ocular Disease in Rodent Models using Optical Coherence Tomography

Published on: July 24, 2020

3.0K

Related Experiment Videos

Last Updated: Aug 14, 2025

Lesion Explorer: A Video-guided, Standardized Protocol for Accurate and Reliable MRI-derived Volumetrics in Alzheimer's Disease and Normal Elderly
12:50

Lesion Explorer: A Video-guided, Standardized Protocol for Accurate and Reliable MRI-derived Volumetrics in Alzheimer's Disease and Normal Elderly

Published on: April 14, 2014

40.3K
Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
10:25

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

Published on: September 25, 2019

48.2K
In vivo Structural Assessments of Ocular Disease in Rodent Models using Optical Coherence Tomography
07:44

In vivo Structural Assessments of Ocular Disease in Rodent Models using Optical Coherence Tomography

Published on: July 24, 2020

3.0K

Area of Science:

  • Medical Imaging
  • Computer-Aided Diagnosis
  • Artificial Intelligence in Radiology

Background:

  • Universal lesion detection (ULD) in computed tomography (CT) is crucial for computer-aided diagnosis (CAD) but challenged by small lesion sizes and high similarity to non-lesions, leading to frequent false positives.
  • Non-lesion structures like bowel and vasculature can obscure small lesions, making differentiation difficult and decreasing detection accuracy.
  • Existing methods struggle to effectively distinguish subtle lesions from complex anatomical backgrounds.

Purpose of the Study:

  • To develop a novel scale-attention module to enhance feature discrimination between lesion and non-lesion regions in CT images.
  • To reduce false positives in universal lesion detection by leveraging radiologist domain knowledge.
  • To improve the accuracy and reliability of computer-aided diagnosis systems for detecting abnormalities in CT scans.

Main Methods:

  • A local axial scale-attention (LASA) module was proposed, inspired by radiologists dividing images into smaller areas for detection.
  • The LASA module adaptively re-weights pixels by aggregating local features from multiple scales, incorporating axial pixel combinations and position embedding.
  • The method was integrated into Convolutional Neural Networks (CNNs) and evaluated on the DeepLesion dataset.

Main Results:

  • The proposed method achieved high sensitivities across various false positive (FP) rates: 78.30% at 0.5 FP/image, 84.96% at 1 FP/image, 89.86% at 2 FP/image, 93.14% at 4 FP/image, 95.36% at 8 FP/image, and 95.54% at 16 FP/image.
  • The average sensitivity at [0.5, 1, 2, 4] FPs per image was 86.56%.
  • Performance significantly outperformed previous methods, demonstrating enhanced feature discrimination between lesions and non-lesions.

Conclusions:

  • The novel LASA module effectively enhances feature discrimination, significantly reducing false positives in universal lesion detection.
  • Exploiting radiologist domain knowledge through the LASA module shows significant potential for improving CAD systems.
  • The proposed method offers a flexible and effective approach for improving lesion detection accuracy in CT imaging.