Search Results
Found 1 results
510(k) Data Aggregation
(206 days)
Acorn Segmentation is intended for use as a software interface and image segmentation system for the transfer of CT or CTA medical images to an output file. Acorn Segmentation is also intended for measuring and treatment planning. The Acorn Segmentation output can also be used for the fabrication of the output file using additive manufacturing methods, Acorn 3DP Models. The physical replica can be used for diagnostic purposes in the field of musculoskeletal and craniomaxillofacial applications.
Acorn Segmentation and 3DP Models should be used in conjunction with expert clinical judgment.
Acorn Segmentation is an image processing software that allows the user to import, visualize and segment medical images, check and correct the segmentations, and create digital 3D models. The models can be used in Acorn Segmentation for measuring, treatment planning and producing an output file to be used for additive manufacturing (3D printing). Acorn Segmentation is structured as a modular package. This includes the following functionality:
- Importing medical images in DICOM format
- Viewing images and DICOM data
- Selecting a region of interest using generic segmentation tools
- Segmenting specific anatomy using dedicated semi-automatic tools or automatic algorithms
- Verifying and editing a region of interest
- Calculating a digital 3D model and editing the model
- Measuring on images and 3D models
- Exporting 3D models to third-party packages
Acorn Segmentation contains both machine learning based auto-segmentation as well as semi-automatic and manual segmentation tools. The auto-segmentation tool is only intended to be used for thoracic and lumbar regions of the spine (T1-T12 and L1-L5). Semi-automatic and manual segmentation tools are intended to be used for all musculoskeletal and craniomaxillofacial anatomy.
Acorn 3DP Model is an additively manufactured physical replica of the digital 3D model generated in Acorn Segmentation. The output file from Acorn Segmentation is used to additively manufacture the Acorn 3DP Model.
The provided text describes the acceptance criteria and the study that proves the device meets those criteria for the Acorn 3D Software (AC-SEG-4009) and Acorn 3DP Model (AC-101-XX).
Here's the breakdown:
1. A table of acceptance criteria and the reported device performance
| Acceptance Criteria (Metric) | Target Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Geometric accuracy of digital models (Dice-Sorensen Coefficient) - Automatic Segmentation (Thoracic and Lumbar spine) | Average Dice-Sorensen Coefficient ≥ 0.93 | Average Dice-Sorensen Coefficient > 0.93 |
| Geometric accuracy of digital models (Dice-Sorensen Coefficient) - Semi-automatic and Manual Segmentation (Musculoskeletal & Craniomaxillofacial bone) | Average Dice-Sorensen Coefficient ≥ 0.93 | Average Dice-Sorensen Coefficient > 0.93 |
| Geometric accuracy of physical replicas (Mean Deviation) | Mean Deviation < 1mm | Mean Deviation < 1mm |
2. Sample size(s) used for the test set and the data provenance
The document does not specify the exact sample size for the test set nor the data provenance (e.g., country of origin, retrospective or prospective). It only states that software verification and validation included bench testing for geometric accuracy.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
The document does not explicitly state the number of experts or their qualifications used to establish the ground truth for the test set.
4. Adjudication method for the test set
The document does not mention any specific adjudication method (e.g., 2+1, 3+1) for the test set.
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
A Multi-Reader, Multi-Case (MRMC) comparative effectiveness study was not explicitly mentioned or described. The study primarily focused on the algorithmic performance against predicate devices and physical models, rather than human reader improvement with AI assistance.
6. If a standalone (i.e. algorithm only without human-in-the loop performance) was done
Yes, a standalone (algorithm only) performance study was conducted. The "Accuracy of automatic segmentation" was evaluated, which by definition means the algorithm's performance without direct human control. The bench testing for geometric accuracy directly assesses the algorithm's output against a reference.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)
The ground truth for the geometric accuracy of digital models was established by comparing the Acorn Segmentation output against "predicate device segmentations." This implies that the segmentations generated by previously cleared and accepted devices served as the reference standard. For physical replicas, the ground truth was the "digital models" themselves.
8. The sample size for the training set
The document states that the "Acorn Segmentation contains both machine learning based auto-segmentation... Using a collection of images and masks as a training dataset for machine-learning segmentation algorithm". However, the specific sample size for the training set is not provided in this document.
9. How the ground truth for the training set was established
The document states that the ground truth for the training set was established by "using a collection of images and masks as a training dataset for machine-learning segmentation algorithm." This implies that pre-existing segmented images (masks) were used as the ground truth for training the machine learning model. The method by which these initial "masks" were generated (e.g., manual segmentation by experts, semi-automatic segmentation, etc.) is not detailed in this document.
Ask a specific question about this device
Page 1 of 1