Image-domain deep learning denoising for low-dose chest CT on a single 128-slice CT platform: a retrospective image-quality assessment - Scorecard - MDSpire

Image-domain deep learning denoising for low-dose chest CT on a single 128-slice CT platform: a retrospective image-quality assessment

  • By

  • Kaiqing Yao

  • Xue Jiang

  • Liang Lv

  • Yang Li

  • Guangpeng Zhang

  • Zhiyuan Zhang

  • Zhiwei Zhang

  • Xinyou Li

  • Fajin Lv

  • July 14, 2026

Share

Clinical Scorecard: Deep Learning Denoising in the Image Domain for Low-Dose Chest CT on a 128-Slice Scanner: A Retrospective Evaluation of Image Quality

At a Glance

CategoryDetail
ConditionLow-Dose Chest CT
Key MechanismsImage-domain deep learning denoising algorithm (AiR Denoising) applied to low-dose chest CT to improve image quality.
Target PopulationPatients undergoing unenhanced chest CT.
Care SettingRadiology departments utilizing low-dose CT protocols.

Key Highlights

  • LD-AiR significantly reduced image noise compared to LD-SAFIRE.
  • Increased signal-to-noise ratio and contrast-to-noise ratio with LD-AiR.
  • Subjective image quality scores were higher for lung parenchyma and mediastinal soft tissue with LD-AiR.
  • LDCT protocol associated with approximately 76% lower effective dose than SDCT.
  • Comparison of LD-AiR and SD-SAFIRE metrics should not be interpreted as evidence of equivalence.

Guideline-Based Recommendations

Diagnosis

  • Evaluate image quality metrics in low-dose chest CT using deep learning denoising.

Management

  • Consider vendor-independent image-domain denoising for improving image quality in low-dose CT.

Monitoring & Follow-up

  • Assess objective and subjective image quality metrics post-processing.

Risks

  • Increased image noise and artefacts in low-dose CT may obscure fine anatomical details.

Patient & Prescribing Data

198 patients who underwent unenhanced chest CT.

Utilization of low-dose CT protocols with deep learning denoising may enhance image quality.

Clinical Best Practices

  • Implement image-domain deep learning denoising in routine low-dose chest CT.
  • Monitor image quality metrics to ensure diagnostic confidence.

Related Resources & Content

Original Source(s)

Image-domain deep learning denoising for low-dose chest CT on a single 128-slice CT platform: a retrospective image-quality assessment

  • by Kaiqing Yao, Xue Jiang, Liang Lv, Yang Li, Guangpeng Zhang, Zhiyuan Zhang, Zhiwei Zhang, Xinyou Li, Fajin Lv

  • July 14, 2026

Related Content

Does Pediatric Readiness Reduce ED Imaging?

Hospitals with stronger pediatric emergency capability used less imaging overall, but differences by insurance status and race and ethnicity remained.

Advances in diagnosis of lung fibrosis: focus on present and future approaches

Quantitative multi-slice spiral CT perfusion parameters as predictors of collateral status and 90-day functional outcome in acute ischemic stroke