Lesion Detection Assessment in the Liver: Standard vs Low Radiation Dose Using Varied Post-Processing Techniques
Study Details
Study Description
Brief Summary
To compare 2 different image creation/processing techniques during a standard CT scan in order to "see" problems in the liver and learn which method provides better image quality. The techniques use new artificial intelligence software to decrease image noise, which helps the radiologist to evaluate.
Condition or Disease | Intervention/Treatment | Phase |
---|---|---|
|
N/A |
Detailed Description
Primary Objective:
To evaluate whether post-processing software Adaptive Statistical Iterative Reconstruction (ASIR), ASIR-V, Veo 3.0 (GE version of Model-based Iterative Reconstruction (MBIR), and Deep Learning Image Reconstruction (DLIR) is able to preserve lesion detection in the liver and other measures of image quality at reduced radiation doses for computed tomography (CT).
Secondary Objectives:
Assessment of whether post-processing software enhances lesion detection in the liver and other measures of image quality at standard and reduced radiation doses.
Assessment of whether DLIR and GSI DLIR reconstructions perform differently, both in terms of accuracy and image quality metrics such as noise reduction.
Study Design
Arms and Interventions
Arm | Intervention/Treatment |
---|---|
Experimental: Computed Tomography Scan - 50% Dose Reduction Participants undergo routine standard of care CT examination for colon carcinoma restaging, then have an additional scan of the liver at 50% dose reduction. |
Diagnostic Test: Computed Tomography Scan - 50% Dose Reduction
Participants undergo routine standard of care CT examination for colon carcinoma restaging, then have an additional scan of the liver at 50% dose reduction.
Other Names:
|
Experimental: Computed tomography Scan - 70% Dose Reduction Participants undergo routine standard of care CT examination for colon carcinoma restaging, then have an additional scan of the liver at 70% dose reduction. |
Diagnostic Test: Computed Tomography Scan - 70% Dose Reduction
Participants undergo routine standard of care CT examination for colon carcinoma restaging, then have an additional scan of the liver at 70% dose reduction.
Other Names:
|
Experimental: Deep Learning Image Reconstruction (DLIR) DLIR is available in both single (SE) and dual/multi energy (DE) CT scanning modes. DLIR SECT and DLIR DECT reconstructions have yet to be compared. |
Diagnostic Test: Deep Learning Image Reconstruction (DLIR)
Participants to receive standard-of-care imaging without the artificial intelligence software and imaging technique.
|
Outcome Measures
Primary Outcome Measures
- Metastasis Detection Accuracy [1 day]
Primary endpoint is metastasis detection accuracy status of each patient, where the standard of care scan reviewed by ''truth readers'' (independent to the blinded radiologists) serve as the gold standard. If any lesion of a patient is diagnosed as metastasis by "truth readers" or blinded readers' consensus, that patient will be considered true positive and diagnosis positive, respectively. The expected accuracy of standard CT is 95%, and a low dose CT detection be considered non-inferior if its accuracy is 85% or higher.
Eligibility Criteria
Criteria
Inclusion Criteria:
-
Patient must be >/= 18 years of age and </=90 years of age
-
Men and non-pregnant women
-
Pathology proven diagnosis of colon or colorectal carcinoma
-
Liver metastases on most recent CT examination
-
Standard of care CT abdomen examination planned WITH IV contrast
Exclusion Criteria:
-
Patients cannot give informed consent
-
Patients cannot undergo CT examination
Contacts and Locations
Locations
Site | City | State | Country | Postal Code | |
---|---|---|---|---|---|
1 | University of Texas MD Anderson Cancer Center | Houston | Texas | United States | 77030 |
Sponsors and Collaborators
- M.D. Anderson Cancer Center
Investigators
- Principal Investigator: Corey T. Jensen, MD, M.D. Anderson Cancer Center
Study Documents (Full-Text)
None provided.More Information
Additional Information:
Publications
None provided.- 2016-1135
- NCI-2018-01272