Search Results

The 1-650 Mammography System is intended to produce radiographic images of the breast. Its specific intended use is for screening and diagnostic mammography. Screening mammography involves the production of images for initial examination for breast cancer diagnosis. Diagnostic mammography includes the production of magnified images for more thorough examination of areas of the breast determined suspicious through screening mammography, special views, spot compression views, and the production of images used by a physician in preparation for biopsy.

The Lorad I-650 mammography system is based on Lorad's currently marketed D-550 (K934870) and Elite (K973631) Dedicated Mammographic Systems. The mechanical packaging is similar to the D-550, while the electronics are similar to the Elite. The Lorad I-650 is a standalone mammography device, mounted on a base to be permanently mounted in position. The unit is AC powered, and must be connected to a 220 VAC/50-60 Hertz (nominal) line to operate. The standalone unit has two major assemblies: The C-arm and the Gantry. The C-arm, which contains the x-ray tube, an image receptor support, compression device, a beam limiting device, switches that actuate the C-arm functions, and sensors for detecting installed accessories and making radiation measurements, is attached to the front (patient side) of the Gantry. The C-arm frame is constructed of two vertical steel rails, which are precision machined to accept the components which it supports. Molded plastic covers enclose the x-ray tube, which is purchased from OEM manufacturers, and the beam limiting device, while aluminum covers enclose the sides, back, top and bottom of the C-arm framework. The C-arm is attached by a locking pivot mechanism to the Gantry. The pivot allows the C-arm to be manually rotated about its axis. The pivot mechanism is attached to the Gantry via a vertically moving carriage. This carriage is capable of movement in the vertical direction, driven by an AC motor and a drive screw mounted to the base of the Gantry. The pivoting action of the C-arm is locked by an electromagnetic brake attached to the vertical carriage, which, in its unactuated condition, prevents rotation. The brake must be electrically actuated to release the C-arm for movement. The Gantry is constructed of a welded steel frame, to which the electrical assemblies and components of the unit are mounted. The operator control is mounted on the side panel of the Gantry, or may be mounted remotely. The input power and input circuit breaker are located on the lower rear panel of the Gantry.

This document describes the Lorad I-650 Mammography System. Here's an analysis of the provided text for acceptance criteria and the study proving it:

Key Takeaway: This document is a 510(k) summary for a mammography system. The "acceptance criteria" are primarily regulatory and performance specifications for the device itself, rather than acceptance criteria for an AI algorithm. The "study" described is a series of bench tests to verify these specifications, not a clinical trial involving human readers or AI performance.

Acceptance Criteria and Device Performance (Table)

The document outlines numerous specifications for the Lorad I-650 Mammography System, which serve as its acceptance criteria. The "reported device performance" are the results from the non-clinical testing.

Study Details (as related to the provided document)

This document describes a medical device submission (510(k)) for a mammography system. It focuses on the technical specifications and non-clinical performance validation of the X-ray machine itself, not an AI algorithm. Therefore, many of the requested points for AI studies are not applicable.

1. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective):

   • Focal Spot Measurement: Not specified in terms of "sample size" of X-ray images, but tests were done at 3 kV settings (25, 28, 30 kV). Data provenance is "bench tests" performed by the manufacturer (Lorad Division of Trex Medical Corporation, Danbury, Connecticut).
   • System Resolution: Not specified in terms of "sample size," but tests were done at 3 kV settings (25, 28, 30 kV) for both focal spots. Data provenance is "bench tests."
   • Object Phantom Tests: Two phantom films were generated. This refers to the number of physical phantom images evaluated. Data provenance is "bench tests."
   • AEC Tracking: Exposures were obtained using breast tissue equivalent materials ranging from 2 cm to 8 cm thickness across various kV settings (e.g., 25-30kV). No specific "sample size" of exposures is given for each condition, but the results table lists 6 conditions for Mo/Mo large focal spot, 3 for Rh large focal spot, and 6 for Mo small focal spot, with multiple thicknesses evaluated for each. Data provenance is "bench tests."
   • Stereotactic Localization: One simulated lesion was targeted and localized. Data provenance is "bench tests."

2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience):

   • Object Phantom Tests: "three readers" were used to score each of the two phantom films. Their qualifications are not specified beyond being "readers."
   • For other tests (Focal Spot, System Resolution, AEC Tracking, Stereotactic Localization), the "ground truth" was derived from physical measurements against known standards or specifications, not human expert interpretation of images.

3. Adjudication method (e.g. 2+1, 3+1, none) for the test set:

   • Object Phantom Tests: The results were "averaged" from the three readers. No formal adjudication (like 2+1 or 3+1 consensus) is described.
   • Not applicable to other tests as ground truth was not based on human consensus.

4. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

   • No. This document pertains to the primary performance of a mammography X-ray system itself, not an AI or CAD system used to assist human readers. Therefore, no MRMC study or effect size related to AI assistance is present.

5. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

   • No. This document is about a hardware medical device (mammography machine), not an AI algorithm. Its performance is evaluated inherently in a "standalone" fashion as a machine, through physical measurements and phantom imaging tests. The "algorithm" here would refer to the machine's internal controls (like AEC), which were tested for their inherent accuracy and reproducibility.

6. The type of ground truth used (expert consensus, pathology, outcomes data, etc):

   • Physical Measurements against Standards/Specifications:
     • Focal Spot Size: Measured physically against NEMA method criteria.
     • System Resolution: Measured from images of line pair bar patterns against MQSA minimums.
     • Object Phantom Tests: Scored visually by readers against ACR/MQSA minimums (e.g., number of discernible fibers, specks, masses). The "ground truth" is the objective presence of these elements in the phantom.
     • AEC Tracking: Measured optical density (OD) consistency against MQSA optical density variation limits using breast tissue equivalent materials.
     • Stereotactic Localization: Comparison of localized coordinates to a known, pre-set 3D position.
   • Referenced Regulatory Standards: Many "acceptance criteria" are derived directly from regulatory bodies like MQSA, ACR, NEMA, IEC, and 21 CFR.

7. The sample size for the training set:

   • Not applicable. This document describes a physical mammography system, not an AI model. There is no concept of a "training set" in this context. The device's internal parameters and controls (e.g., AEC calibration) are set during manufacturing and validated through the non-clinical tests described.

8. How the ground truth for the training set was established:

   • Not applicable. As there is no AI training set, this question does not apply.

Ask a specific question about this device