Image Quality: Ultra-Low Dose Scanner Versus Standard Dose Conventional Scanner for Thoraco-abdominopelvic Scans (UL2DLR)

Image Quality: Ultra-Low Dose Scanner Versus Standard Dose Conventional Scanner for Thoraco-abdominopelvic Scans

Concordance of Image Quality Between an Ultra-Low Dose (ULD) Scanner and Standard Dose Conventional Scanner for Thoraco-abdominopelvic Scans

The new image reconstruction algorithm (Precise Image, Philips Healthcare) has a strong potential to maintain sufficient image quality suitable for diagnosis with ultra-low dose (ULD) chest and abdomen-pelvis scans. The hypothesis is that the images obtained with the Precise Image algorithm for ULD acquisitions are of sufficient and suitable quality for the diagnosis of certain lung, abdominal-pelvic and bone lesions.

Recently a new image reconstruction algorithm based on Deep-learning has been developed (Precise Image, Philips Healthcare). Initial studies on phantoms have shown that this algorithm improves image quality and reduces patient dose compared to the iDose4 iterative reconstruction algorithm. Feasibility studies have validated the image quality for low-dose levels (LD). However, this algorithm has a strong potential to maintain sufficient image quality suitable for diagnosis with ultra-low dose (ULD) chest and abdomen-pelvis scans. The hypothesis is that the images obtained with the Precise Image algorithm for ULD acquisitions are of sufficient and suitable quality for the diagnosis of certain lung, abdominal-pelvic and bone lesions. The purpose of this study is to evaluate the concordance of the global quality of thoraco-abdominopelvic images of a ULD scan acquisition compared to a standard dose CT acquisition and measure the global agreement of the global quality of the images with a 4-point Likert scale. The ULD acquisition will allow a significant reduction in the X-ray dose delivered to patients compared to a standard dose conventional scanner. This reduction is estimated at 70%.

The reading of the images from the ULD scanner acquisition will be performed blinded to the patient's clinical record, the diagnosis made by the radiologist in charge of the patient on the basis of the standard dose conventional scanner, the result of the reading of the standard dose conventional scanner study and the reading of the other evaluator.

All adult patients (except pregnant women) presenting to the imaging department of the Institut de Cancérologie du Gard for a thoraco-abdominopelvic scan as part of an oncology follow-up will under a standard-dose can as well as an ultra-low-dose scan.

In addition to the usual management (i.e. standard dose Computed Tomography), an Ultra-Low-Dose thoraco-abdominopelvic scan will be routinely performed.

Concordance of the global quality of thoraco-abdomino-pelvic images from an Ultra-Low-Dose scanner acquisition compared to a standard dose scanner acquisition

A 4-point Likert scale will be used to evaluate the overall image quality to see whether it is good enough to visualize the structures of interest and possible lesions. Overall image quality is rated as follows: Completely interpretable without noise, smoothing, and/or artifact, Completely interpretable with low artifacts and/or low level of noise or/and smoothing, Interpretable with moderate noise and/or smoothing and moderate artifacts, Uninterpretable

Distribution of image quality for each of the two examinations of interest (ULD scanner and standard dose scanner) and to estimate the frequency of good quality images (level 1-2) for each of the two examinations. - Fully interpretable without noise, smoothing and/or artifact, - Fully interpretable with low artifacts and/or low noise and/or smoothing, - Interpretable despite moderate noise and/or smoothing and moderate artifacts, - Not interpretable

Distribution of image quality for each of the two examinations of interest (ULD scanner and standard dose scanner) and to estimate the frequency of good quality images (level 1-2) for each of the two examinations. Grouping of the variable into 2 classes for the estimation of the frequency of good quality (1-2) images:- 1-2 vs. 3-4

Distribution of image quality for each of the two examinations of interest (ULD scanner and standard dose scanner) and to estimate the frequency of good quality images (level 1-2) for each of the two examinations. - Fully interpretable without noise, smoothing and/or artifact, - Fully interpretable with low artifacts and/or low noise and/or smoothing, - Interpretable despite moderate noise and/or smoothing and moderate artifacts, - Not interpretable

Distribution of image quality for each of the two examinations of interest (ULD scanner and standard dose scanner) and to estimate the frequency of good quality images (level 1-2) for each of the two examinations. Grouping of the variable into 2 classes for the estimation of the frequency of good quality (1-2) images:- 1-2 vs. 3-4

To compare the confidence level of each radiologist's diagnosis for each of the two examinations of interest (Ultra-Low-Dose scan and standard dose scan) a subjective assessment of the each radiologist's (Junior/Senior) diagnostic confidence level will be made.

To compare the confidence level of each radiologist's diagnosis for each of the two examinations of interest (Ultra-Low-Dose scan and standard dose scan) a subjective assessment of the each radiologist's (Junior/Senior) diagnostic confidence level will be made.

Likert scales for global image quality and the level of confidence of the diagnosis will be used to assess inter-rater agreement (senior radiologist vs. junior radiologist) for each CT acquisition.

Likert scales for global image quality and the level of confidence of the diagnosis will be used to assess inter-rater agreement (senior radiologist vs. junior radiologist) for each CT acquisition.

Overall concordance of the diagnosis of thoracic, abdominal and/or pelvic lesions made from a ULD scan or from a conventional standard dose scan in these patients. Presence of at least one thoracic, (Yes/No)

Overall concordance of the diagnosis of thoracic, abdominal and/or pelvic lesions made from a ULD scan or from a conventional standard dose scan in these patients. Presence of at least one thoracic lesion (Yes/No)

Overall concordance of the diagnosis of thoracic, abdominal and/or pelvic lesions made from a ULD scan or from a conventional standard dose scan in these patients. Presence of at least one abdominal lesion (Yes/No)

Overall concordance of the diagnosis of thoracic, abdominal and/or pelvic lesions made from a ULD scan or from a conventional standard dose scan in these patients. Presence of at least one abdominal lesion (Yes/No)

Overall concordance of the diagnosis of thoracic, abdominal and/or pelvic lesions made from a ULD scan or from a conventional standard dose scan in these patients. Presence of at least one pelvic lesion (Yes/No)

Overall concordance of the diagnosis of thoracic, abdominal and/or pelvic lesions made from a ULD scan or from a conventional standard dose scan in these patients. Presence of at least one pelvic lesion (Yes/No)

Overall concordance of the diagnosis of bone lesions of the spine and pelvis made from a ULD scan or from a standard dose CT scan in these patients. Presence of at least one bone lesion within the spine (Yes/No).

Overall concordance of the diagnosis of bone lesions of the spine and pelvis made from a ULD scan or from a standard dose CT scan in these patients. Presence of at least one bone lesion within the spine (Yes/No).

Overall concordance of the diagnosis of bone lesions of the spine and pelvis made from a ULD scan or from a standard dose CT scan in these patients. Presence of at least one bone lesion on the pelvis (Yes/No).

Overall concordance of the diagnosis of bone lesions of the spine and pelvis made from a ULD scan or from a standard dose CT scan in these patients. Presence of at least one bone lesion on the pelvis (Yes/No).

Inclusion Criteria: Patient having given free and informed consent. Patient who has signed the consent form. Patient affiliated to or beneficiary of a health insurance plan. Patient with a BMI < 35 kg/m2 Exclusion Criteria: Patient participating in research involving human subjects defined as Category 1. Patient in an exclusion period as determined by another study. Patient under court protection, guardianship or trusteeship. Patient unable to give consent. Patient for whom it is impossible to give informed information. Pregnant patient. Patient with a BMI ≥ 35 kg/m2