Liver CT Dose Reduction With Deep Learning Based Reconstruction

Seoul National University Hospital300 enrolled

Overview

A deep learning-based de-noising (DLD) reconstruction algorithm (ClariCT.AI) has the potential to reduce image noise and improve image quality. This capability of the CliriCT.AI program might enable dose reduction for contrast-enhanced liver CT examination. In this prospective multicenter study, whether the ClariCT.AI program can reduce the noise level of low-dose contrast-enhanced liver CT (LDCT) data and therefore, can provide comparable image quality to the standard dose of contrast-enhanced liver CT (SDCT) images will be evaluated. The aim of this study is to compare image quality and diagnostic capability in detecting malignant tumors of LDCT with DLD to those of SDCT with MBIR using the predefined non-inferiority margin.

Study Type

OBSERVATIONAL

Enrollment

300

Conditions

Radiation Exposure Liver Cancer

Interventions

Eligibility

Sex: ALLMin age: 20 YearsMax age: 85 Years

Medical Language ↔ Plain English

Inclusion Criteria: * Age between 20-year-old and 85 years old * patients referred to the Radiology department to perform contrast-enhanced liver CT under the suspicion of focal liver lesions Exclusion Criteria: * patients with estimated glomerular filtration rate \< 60 mL/min/1.73m2 * previous history of severe adverse reaction to iodinated contrast media.

Locations (3)

Tubingen University Hospital

Tübingen, Germany

Seoul National University Hospital

Seoul, South Korea

Korea University Guro Hospital

Seoul, South Korea

Outcomes

Primary Outcomes

Measurement of standard deviation of CT attenuation values at the liver

Standard deviation of CT attenuation values at the liver parenchyma

Time frame: within 6 months from acquisition of liver CT scans

Secondary Outcomes

Sensitivity to detect malignant liver tumor

Sensitivity of liver CT scans to detect malignant liver tumor

Time frame: within 6 months from acquisition of liver CT scans