New framework removes artifacts from Photoacoustic CT images

Researchers have developed a novel self-supervised framework to remove artifacts from Photoacoustic Computed Tomography (PACT) images. This method utilizes a Siamese neural network and a composite loss function that considers cross-domain fidelity and uncertainty-weighted consistency. The framework effectively separates dual-domain features to filter out artifacts, demonstrating significant improvements in image quality across simulations, phantoms, and both rat and human experimental data. Additionally, the approach offers computational efficiency through accelerated inverse operators. AI

IMPACT Improves image quality in medical imaging applications, potentially aiding in pre-operative planning and diagnosis.

RANK_REASON Academic paper detailing a new technical method. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• arXiv
• brown rat
• Hugging Face
• human
• Siamese neural network

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →