(275 days)
AmCAD-UT is a Windows-based computer-aided detection (CADe) device intended to assist the medical professionals in analyzing thyroid ultrasound images, acquired from FDA-cleared ultrasound systems. The region of interest (ROI) of a user-selected thyroid nodule is defined by users or suggested by an AI contouring algorithm. After the initial review of the ultrasound images by the physicians, the device further provides detailed information with quantification and visualization of sonographic characteristics of thyroid nodules. The device is intended for use on ultrasound images of discrete thyroid nodules larger than 1cm, for which a biopsy recommendation is required.
AmCAD-UT is a Windows-based computer-assisted detection (CADe) software application device designed to assist medical professionals in analyzing thyroid ultrasound images with the region of interest (ROI) of a selected nodule defined by users or suggested by an Al algorithm after the user specifies the location of the nodule.
After the initial review of thyroid ultrasound images by the physician, he/she can use AmCAD-UT to analyze thyroid images for further interpretation. Once the ROI is confirmed, the physician may process the image for detection and quantification of sonographic characteristics (i.e., hyperechoic foci, echogenicity, texture, margin, orientation and anechoic areas) by AmCAD-UT. The device provides more detailed information with quantification and visualization of the sonographic characteristics of thyroid nodule that may assist physician in his/her complete interpretation.
The software application automatically generates reports given the user preference inputs (e.g., the nodule size, nodule location and shape, and the presence or absence of the sonographic characteristics) annotated during the image analysis process. A report form has been designed by AmCad to be consistent with the conventional clinical thyroid report form. An output of the report may be viewed and sent to paper printers or saved on the standalone PC or review station as PDF file.
The provided text describes a 510(k) premarket notification for AmCAD-UT, a computer-aided detection (CADe) device for analyzing thyroid ultrasound images.
Here's an analysis of the acceptance criteria and study information:
1. A table of acceptance criteria and the reported device performance
The document does not explicitly state acceptance criteria in a quantitative format (e.g., minimum sensitivity/specificity thresholds). Instead, it states that "the performance data demonstrates that it performs effectively and the device is as safe and effective as the predicate device."
However, it does mention that "the device was effective in determining the contour of thyroid nodules." This refers to the performance of the AI-suggested ROI contouring, which is a new feature compared to the predicate.
Given the information, a table would look something like this, acknowledging the lack of specific quantitative acceptance criteria:
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Demonstrates effective performance in AI-suggested ROI contouring of thyroid nodules. | The device was effective in determining the contour of thyroid nodules. Performance data demonstrates it performs effectively and is as safe and effective as the predicate device. |
2. Sample size used for the test set and the data provenance
The document does not specify the exact sample size (number of images or nodules) used for the test set in the standalone performance studies.
Regarding data provenance: While not explicitly stated, it's mentioned that the images are "acquired from FDA-cleared ultrasound systems." The manufacturer is AmCad BioMed Corporation, located in Taiwan, R.O.C. It is likely the data originated from (or was collected by) clinical sites associated with the manufacturer or collaborating institutions, possibly in Taiwan or internationally, but this is not explicitly stated. The study type is referred to as "standalone performance studies."
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
The document states: "The ground truth to be established for performance studies of the device is the ROI labeled by a panel of specialists."
- Number of experts: "a panel of specialists" (the exact number is not specified).
- Qualifications of experts: "specialists" (specific qualifications, e.g., "radiologist with 10 years of experience", are not provided).
4. Adjudication method for the test set
The document does not explicitly describe an adjudication method (such as 2+1 or 3+1) for the ground truth establishment by the "panel of specialists." It simply says the ground truth is the "ROI labeled by a panel of specialists," implying consensus or a collective determination, but without detailing the process.
5. 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
- Was an MRMC study done? No, an MRMC comparative effectiveness study was not done for this particular submission. The "Performance Testing Data to Support SE Determination" table explicitly contrasts the current device's data ("Results from standalone performance testing of the AI suggested ROI's of user-selected nodules") with the predicate device's data, which included "Results from standalone performance testing and clinical performance testing (MRMC study)." This indicates that the MRMC study was performed for the predicate device (K180006), not for the new K203555 submission.
- Effect size of human reader improvement: Since an MRMC study was not performed for this device, no effect size of human readers improving with AI vs. without AI assistance is reported for K203555.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
Yes, a standalone performance study was done. The document states:
- "AmCad BioMed Corporation has conducted standalone performance studies to validate and assess the performance of the AmCAD-UT for its added function of AI-suggested ROI contouring."
- "The standalone studies evaluated the performance of the contours suggested by the AI algorithm of user-selected nodules..."
7. The type of ground truth used
The document states:
- "The ground truth to be established for performance studies of the device is the ROI labeled by a panel of specialists."
- This indicates the ground truth for the ROI contouring was established by expert consensus/labeling.
It's worth noting that the ground truth for the predicate device (AmCAD-UT® Detection 2.2) included "the ROI, the presence of each sonographic characteristic, and the surgical pathology examination result," suggesting a more comprehensive ground truth for the predicate, potentially including pathology as a definitive outcome. However, for the current device's specific new function (AI-suggested ROI), the ground truth focus is on expert-labeled ROI.
8. The sample size for the training set
The document does not specify the sample size used for the training set.
9. How the ground truth for the training set was established
The document does not explicitly state how the ground truth for the training set was established. It only discusses the ground truth for "performance studies," which typically refers to test/validation sets. However, it's reasonable to infer that the training data would be labeled by similar expert methods, but this is not confirmed in the provided text.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).