RESEARCH PROGRESS OF DEEP LEARNING IN LOW-DOSE CT IMAGE DENOISING

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Radiology

Background:

• Low-dose computed tomography (CT) reduces radiation exposure but increases image noise and artifacts.
• Image noise and artifacts in low-dose CT hinder accurate radiological diagnosis.
• Effective denoising is crucial for maintaining diagnostic quality in low-dose CT imaging.

Purpose of the Study:

• To explore and evaluate deep learning-based methods for low-dose CT image denoising.
• To compare the performance of deep learning techniques against traditional denoising approaches.
• To demonstrate the superiority of deep learning in improving image quality for low-dose CT.

Main Methods:

• Investigated three state-of-the-art deep learning algorithms for image denoising.
• Included four traditional image denoising methods as a control group for comparison.
• Conducted comprehensive experiments to assess denoising effectiveness.

Main Results:

• Deep learning methods achieved superior denoising results compared to traditional methods.
• Improvements were observed in both subjective visual quality and objective quantitative metrics.
• Deep learning effectively reduced noise and artifacts in low-dose CT images.

Conclusions:

• Deep learning-based methods are highly effective for low-dose CT image denoising.
• These advanced techniques offer significant advantages over traditional approaches for medical imaging.
• Deep learning enhances diagnostic confidence and accuracy in low-dose CT examinations.