Search Results

Ask a specific question about this device

The CADDIE computer-assisted detection device is intended to assist the gastroenterologist in detecting suspected colorectal polyps only. The gastroenterologist is responsible for reviewing CADDIE suspected polyp areas and confirming the presence or absence of a polyp based on their own medical judgment.

CADDIE is not intended to replace a full patient evaluation, nor is it intended to be relied upon to make a primary interpretation of endoscopic procedures, medical diagnosis, or recommendations of action for patients. The CADDIE computer-assisted detection device is limited for use with standard white-light endoscopy imaging only.

CADDIE is cloud based artificial intelligence medical device software. CADDIE interfaces with the video feed generated by an endoscopic video processor during a colonoscopy procedure

The software is intended to be used by trained and qualified healthcare professionals as an accompaniment to video endoscopy for the purpose of drawing attention to regions with visual characteristics consistent with colonic mucosal lesions (such as polyps and adenomas).

CADDIE analyses the data from the endoscopic video processor in real-time and provides information to aid the endoscopist in detecting suspected colorectal polyps, if they are in the field of view of the endoscope.

The areas highlighted by CADDIE are not to be interpreted as definite polyps or adenomas. The responsibility to make a decision as to whether or not a highlighted region contains a polyp or is an adenoma lies with the user. The endoscopist is responsible for reviewing CADDIE suspected polyp areas and confirming the presence or absence of a polyp and its classification based on their own medical judgement.

Here's a breakdown of the acceptance criteria and study details for the CADDIE device, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

Polyp Detection (Standalone Bench-testing Data):

*Primary Endpoints; **Secondary Endpoints

Cecum AI (Standalone Bench-testing Data):

Conclusion: The device met all stated acceptance criteria in the standalone testing and demonstrated superiority in APC with non-inferiority in PPA in the clinical study.

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

Polyp Detection Standalone Bench-testing Dataset:

• Sample Size (Subjects): 389 subjects.
• Data Provenance: Not explicitly stated, but the demographics include African American, American Indian, Asian, Caucasian, and Hispanic races/ethnicities, suggesting a diverse multi-region dataset, potentially US-based given the specific racial categories listed. This was a retrospective analysis as it used recorded colonoscopy videos and compared results to historical control (known polyp status per frame).

Cecum AI Standalone Bench-testing Dataset:

• Sample Size (Frames): 5092 total frames (2833 positive frames, 2259 negative frames).
• Data Provenance: Not explicitly stated, but it uses recorded colonoscopy frames, implying a retrospective analysis.

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

Polyp Detection Standalone Bench-testing Dataset:

• Experts: A team of trained clinical annotators initially labeled polyp structures, followed by an additional layer of review by a separate team of experts.
• Qualifications: The "separate team of experts" had "over 2000 endoscopic procedures experience."

Cecum AI Standalone Bench-testing Dataset:

• Experts: A team of trained clinical annotators labeled cecal structures.
• Qualifications: Not explicitly stated beyond "trained clinical annotators."

4. Adjudication Method for the Test Set

Polyp Detection Standalone Bench-testing Dataset:

• Annotation was performed on a per-frame basis. A "team of trained clinical annotators" labeled polyp structures, followed by an "additional layer of review by a separate team of experts." This indicates a multi-reader review process, likely with a consensus or hierarchical adjudication, though the exact method (e.g., 2+1, 3+1) is not specified.

Cecum AI Standalone Bench-testing Dataset:

• Annotation was performed on a per-frame basis by a "team of trained clinical annotators." No additional layer of review or specific adjudication method (like 2+1) is mentioned.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size of Human Improvement with AI vs. Without AI Assistance

• Yes, a prospective, multi-center, MRMC, randomized controlled, parallel group trial was done.
• Effect Size of Human Improvement with AI vs. Without AI Assistance:
  • Adenomas Per Colonoscopy (APC): CADDIE (with AI) resulted in an APC of 0.82 ± 1.40, while Standard of Care (without AI) had an APC of 0.62 ± 1.19. The ratio of CADe to SoC was 1.33 (95% CI: 1.06, 1.67), meaning 33% more adenomas per colonoscopy were detected with AI assistance.
  • Adenoma Detection Rate (ADR): CADe group had an ADR of 42.9%, SoC had 35.9%. The difference was 7.1% (95% CI: 0.5%, 13.7%), meaning AI assistance led to a 7.1% absolute increase in the proportion of examinations with at least one adenoma detected.
  • AI assistance also led to significant increases in detection of diminutive (≤5 mm) adenomas/adenocarcinomas (29% more) and large (≥10 mm) adenomas/adenocarcinomas (93% more).

6. If a Standalone (Algorithm Only Without Human-in-the-Loop Performance) Was Done

• Yes, standalone performance testing was performed for both the Polyp Detection component and the Cecum AI component.
  • For Polyp Detection, a set of recorded colonoscopy videos was analyzed by CADDIE, and the results were compared to historical controls.
  • For Cecum AI, a set of recorded colonoscopy frames were analyzed by Cecum AI, and the results were compared to historical controls.

7. The Type of Ground Truth Used

Polyp Detection Standalone Bench-testing:

• Histology: Each polyp was "histologically confirmed." The ground truth for polyp annotations was based on these confirmed histology reports.

Cecum AI Standalone Bench-testing:

• Expert Annotation: Ground truth reference standards were "annotations performed on a per-frame basis, where a team of trained clinical annotators labelled cecal structures with a bounding box."

Clinical Study (MRMC):

• Histology/Pathology: The primary and secondary endpoints (APC, PPA, ADR, etc.) were based on "histologically confirmed" findings of polyps, adenomas, adenocarcinomas, and sessile serrated lesions.

8. The Sample Size for the Training Set

Polyp Detection Development Datasets:

• Number of Polyps: 1711 polyps.
• Number of Patients: 906 patients.
• Number of Frames: 318,603 frames (162,207 polyp frames and 156,396 non-polyp frames).
  • This dataset was used for training, tuning, and testing (development data, separate from bench-testing data).

Cecum AI Development Datasets:

• Number of Patients: 1467 patients.
• Number of Images: 17,844 images.
  • This dataset was used for training, tuning, and testing (development data, separate from bench-testing data).

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

Polyp Detection Development Datasets:

• Ground truth was based on a combination of histology (for 1296 polyps from 714 patients) and optical confirmation by additional endoscopists (for 415 polyps confirmed through resection or photo-documentation, but not histopathology).

Cecum AI Development Datasets:

• The ground truth for the Cecum AI development dataset was established by using "informative static photo-documentation images, as well as images extracted from videos of cecal landmarks including appendiceal orifice (AO), ileocecal valve (ICV)." While not explicitly stated as "expert annotation" for the training set, this description implies that the landmarks were identified and labeled. The standalone test set confirmed ground truth by "a team of trained clinical annotators," suggesting a similar method for development.

Ask a specific question about this device

Not Found

The Cath Caddie was developed by the grandmother of a pediatric cancer patient who received chemotherapy via a Hickman Catheter Constant removal of taped dressings placed in his chest. resulted in skin irritations to the youth. The inventor created the Cath Caddie for her grandchild to eliminate the need for tape. Subsequently, she provided Cath Caddies for several additional patients of Dr. David Rosen. She estimates that Dr.Rosen has issued as many as 1,500 of the covers through his practice and through the Children's Miracle Network. Additional prototypes of the Cath Caddie were provided to Dr. Jerry H. Feagan who in turn passed them on to the Endoscopy and Surgery Center of Topeka. This practice utilized the Cath Caddie to cover and secure PEG tubes of enteral feeding tubes in the abdomens of some of their patients.

This 510(k) summary for the Cath Caddie does not contain the information requested regarding acceptance criteria and a study demonstrating the device meets those criteria.

The document describes the origin and anecdotal use of the Cath Caddie, a device designed to cover and secure catheters (Hickman and PEG tubes). It highlights perceived benefits such as preventing skin irritation from tape, improving patient comfort, and potentially reducing healthcare costs.

However, the summary lacks any quantitative data, specific acceptance criteria, or details of a formal study that would be necessary to fulfill the request.

Therefore, I cannot provide:

1. A table of acceptance criteria and the reported device performance: No specific performance metrics or acceptance criteria are mentioned.
2. Sample size used for the test set and the data provenance: The document refers to "1,500 of the covers" issued by Dr. Rosen and use at the Endoscopy and Surgery Center, but these are not described as part of a formal test set with rigorous data collection. The data is observational and anecdotal, not from a controlled test.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts: There is no mention of ground truth establishment or expert review in a study context.
4. Adjudication method for the test set: No adjudication method is described.
5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done: No such study is mentioned.
6. If a standalone performance (i.e. algorithm only without human-in-the-loop performance) was done: This is a physical device, not an algorithm, so this question is not applicable.
7. The type of ground truth used: No ground truth in an evaluative sense is described.
8. The sample size for the training set: No training set is applicable or mentioned.
9. How the ground truth for the training set was established: No training set or ground truth for it is applicable or mentioned.

The document is purely descriptive and testimonial in nature, focusing on the perceived benefits and the origin story of the device.

Ask a specific question about this device