Validation Testing of CADe for Pulmonary Nodules
Principal Investigator
Name
Yusuke Nakamura
Degrees
Master of Engineering
Institution
Plusman LLC
Position Title
Managing Partner
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-667
Initial CDAS Request Approval
Apr 30, 2020
Title
Validation Testing of CADe for Pulmonary Nodules
Summary
Plusman LLC developed a computer assisted detection (CADe) device which detects and characterizes pulmonary nodules on chest CT images. To get 510(k) clearance of the CADe, validation testing is mandatory. NLST data and data in two or more US based sites will be used for the validation testing as follows.
Reference standard: US certified radiologists make reference standard to indicate for patient data indicates whether or not the pulmonary nodule is present and may include such attributes as the extent or location of the pulmonary nodule. CADe device development and evaluation often relies on databases of radiology images or radiology device data with a reference standard addressing whether or not the pulmonary nodule is present within an individual patient and if so, its location and extent.
Scoring: CADe device development and evaluation often rely on determining whether the spatial location and extent of the CADe mark correspond to the location and extent of the pulmonary nodule. The scoring is processed to determine the correspondence between the CADe output and the reference standard.
Standalone performance assessment: Standalone performance assessment indicates the performance of the CADe. We compare the performance of the CADe and the predicate device. The data is processed with the CADe and the predicate device. The result will be evaluated by Free-Response Receiver Operating Characteristic (FROC) curve and the number of false-positive per case.
Reference standard: US certified radiologists make reference standard to indicate for patient data indicates whether or not the pulmonary nodule is present and may include such attributes as the extent or location of the pulmonary nodule. CADe device development and evaluation often relies on databases of radiology images or radiology device data with a reference standard addressing whether or not the pulmonary nodule is present within an individual patient and if so, its location and extent.
Scoring: CADe device development and evaluation often rely on determining whether the spatial location and extent of the CADe mark correspond to the location and extent of the pulmonary nodule. The scoring is processed to determine the correspondence between the CADe output and the reference standard.
Standalone performance assessment: Standalone performance assessment indicates the performance of the CADe. We compare the performance of the CADe and the predicate device. The data is processed with the CADe and the predicate device. The result will be evaluated by Free-Response Receiver Operating Characteristic (FROC) curve and the number of false-positive per case.
Aims
To collect evidences of our computer-assisted detection (CAD) device for 510(k) clearance
Collaborators
American College of Radiology and some medical institutions in the US but not be fixed yet