Search Results

It is used by radiation oncology department to register multimodality images and segment (non-contrast) CT images, to generate needed information for treatment planning, treatment evaluation and treatment adaptation.

The proposed device, AccuContour™, is a standalone software which is used by radiation oncology department to register multimodality images and segment (non-contrast) CT images, to generate needed information for treatment planning, treatment evaluation and treatment adaptation.

The product has two image process functions:
(1) Deep learning contouring: it can automatically contour the organ-at-risk, including head and neck, thorax, abdomen and pelvis (for both male and female),
(2) Automatic Registration, and
(3) Manual Contour.

It also has the following general functions:
Receive, add/edit/delete, transmit, input/export, medical images and DICOM data;
A Patient management;
Review of processed images;
Open and Save of files.

Here's a breakdown of the acceptance criteria and study information for the AccuContour™ device, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state numerical acceptance criteria for DICE Similarity Coefficients (DSC) for segmentation or Normalized Mutual Information (NMI) for registration. Instead, it states the acceptance criterion is non-inferiority compared to the predicate device.

2. Sample Size Used for the Test Set and Data Provenance

• Segmentation Performance Test:

  • Test Set Description: Two separate tests were performed.
    • One test involved images generated in healthcare institutions in China using scanner models from GE, Siemens, and Philips.
    • The other test involved images generated in healthcare institutions in the US using scanner models from GE, Siemens, and Philips.
    • For each body part, all intended organs were included in images from both US and China datasets.
  • Sample Size: The exact number of images or cases in each test set is not specified.
  • Data Provenance: Retrospective, from healthcare institutions in China and the US.

• Registration Performance Test:

  • Test Set Description: Two separate tests were performed.
    • One test involved images generated in healthcare institutions in China using scanner models from GE, Siemens, and Philips, tested on multi-modality image sets from the same patients.
    • The other test involved most images generated in healthcare institutions in the US, with a small amount of moving images adopted from online databases (originally from non-US sources). This test was on multi-modality image sets from different patients.
    • Both tests covered various modalities (CT/CT, CT/MR, CT/PET).
  • Sample Size: The exact number of images or cases in each test set is not specified.
  • Data Provenance: Retrospective, from healthcare institutions in China and the US, with some online database images (non-US origin) for the US registration test.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

• Number of Experts: At least three licensed physicians.
• Qualifications of Experts: Licensed physicians. (Further sub-specialty or years of experience are not specified, but licensure implies a professional medical qualification.)

4. Adjudication Method for the Test Set

The ground truth was generated from the consensus of at least three licensed physicians. This implies an adjudication method where all experts agree, or a majority agreement based on the "consensus" phrasing, but the specific process (e.g., voting, discussion to reach full agreement) is not detailed.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

No, a multi-reader multi-case (MRMC) comparative effectiveness study was not performed or reported in this summary. The comparison was algorithm-to-algorithm (proposed device vs. predicate device), not involving human readers' performance with and without AI assistance.

6. Standalone Performance Study

Yes, a standalone (algorithm only without human-in-the-loop performance) study was performed. The segmentation and registration accuracies (DICE and NMI respectively) were calculated for the proposed device's algorithm and compared to the predicate device's algorithm.

7. Type of Ground Truth Used

The ground truth used was expert consensus. Specifically, for both segmentation and registration, ground truthing of each image was generated from the consensus of at least three licensed physicians.

8. Sample Size for the Training Set

The document does not specify the sample size used for the training set. It only describes the test sets.

9. How the Ground Truth for the Training Set was Established

The document does not provide information on how the ground truth for the training set was established. It only details the ground truth establishment for the test sets.

Ask a specific question about this device