Search Results

Acorn 3D Software is a modular image processing software intended for use as an interface for visualization of medical images, segmentation, treatment planning, and production of an output file.

The Acorn 3D Segmentation module 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 3D Segmentation is also intended for measuring and treatment planning. The Acorn 3D Segmentation output can also be used for the fabrication of physical replicas 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.

The Acorn 3D Alignment and Measurement module contains registration capabilities and measurement functionality based on anatomical reference geometry. It is intended to allow the user to align anatomical structures between datasets, perform spinopelvic measurements on 3D models of anatomy, and plan surgical procedures in pediatric and adult patients.

Acorn 3D Software and 3DP Models should be used in conjunction with expert clinical judgment.

Acorn 3D Software 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 3D Software for measuring, treatment planning and producing an output file to be used for additive manufacturing (3D printing). Acorn 3D Software 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 fully 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
• Image registration

The Acorn 3D Segmentation module 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) and the pelvis (sacrum). Semi-automatic and manual segmentation tools are intended to be used for all musculoskeletal anatomy.

Acorn 3DP Model is an additively manufactured physical replica of the virtual 3D model generated in Acorn 3D Segmentation. The output file from Acorn 3D Segmentation is used to additively manufacture the Acorn 3DP Model.

The Acorn 3D Alignment and Measurement module contains registration capabilities and spinopelvic measurement functionality. It is intended to align spinopelvic anatomical structures between datasets. The module allows the user to perform spinopelvic measurements on 3D models of anatomy, and to plan surgical procedures.

Here's a breakdown of the acceptance criteria and study details for the Acorn 3D Software, based on the provided FDA 510(k) clearance letter:

Acceptance Criteria and Reported Device Performance

Study Details

1. Sample Sizes and Data Provenance

Test Set:

• Vertebral Model (T1-T12, L1-L5):
  • In-House (Mighty Oak Medical): 35 cases (450 segments - 139 lumbar / 311 thoracic).
  • VERSE '20: 36 cases (401 segments - 144 lumbar / 257 thoracic).
  • Total Test Set for Vertebral Model: 71 cases.
• Sacral Model:
  • In-House (Mighty Oak Medical): 40 cases.
  • Total Test Set for Sacral Model: 40 cases.

Data Provenance (Test Set):

• In-House (Mighty Oak Medical): Retrospective (implied, as it's an internal database). Country of origin not explicitly stated.
• VERSE '20: Publicly available dataset. Inclusion criteria: Patients > 30 years of age, no history of bone metastases. Specific country of origin not explicitly stated, but typically these are multi-institutional datasets from various countries. Retrospective.
• CTPelvic1K: Publicly available dataset used only for training and tuning the Sacral model. Country of origin not explicitly stated.

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

• Mighty Oak Medical (In-House): "Highly-trained experts" and "operators who have undergone extensive classroom training and tests for efficacy, with an established track record of creating accurate segmentations." Specific number not mentioned. Qualifications are described as having an "established track record of creating accurate segmentations in support of the FIREFLY Pedicle Screw Navigation Guide."
• VERSE '20: "Widely accepted as ground truth, with three levels of annotation and expert review (by medical and/or graduate students, radiology fellows, and finally a senior consultant)." Specific number not mentioned. The qualifications range from medical/graduate students to radiology fellows and senior consultants.

3. Adjudication Method for the Test Set

• Mighty Oak Medical (In-House): "Verified by feedback loops of expert review and manual editing and correction."
• VERSE '20: "Three levels of annotation and expert review (by medical and/or graduate students, radiology fellows, and finally a senior consultant)." This implies a multi-level consensus approach, but a specific method like "2+1" or "3+1" is not explicitly stated.

4. MRMC Comparative Effectiveness Study

No Multi-Reader Multi-Case (MRMC) comparative effectiveness study was mentioned where human readers' improvement with AI vs. without AI assistance was evaluated. The performance data focuses on the standalone algorithm's accuracy, and the device is intended to be used "in conjunction with expert clinical judgment."

5. Standalone Performance Study

Yes, standalone (algorithm only) performance was done. The "Machine-Learning Model Performance Testing" section directly compares the model's output (segmentations) with ground truth segmentations using the DICE Similarity Coefficient (DSC). The "Alignment Accuracy" evaluation also assesses the algorithm's performance directly against ground truth.

6. Type of Ground Truth Used

• Expert Consensus: For both In-House (Mighty Oak Medical) and VERSE '20 datasets, the ground truth was established through expert review and refinement of segmentations.
• Pathology/Outcomes Data: Not explicitly mentioned as being used for ground truth for the segmentation or alignment accuracy.

7. Sample Size for the Training Set

• Vertebral Model (T1-T12, L1-L5): 147 cases (from Mighty Oak Medical In-House dataset).
• Sacral Model: 204 cases (104 from Mighty Oak Medical In-House dataset, 100 from CTPelvic1K dataset).

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

• Mighty Oak Medical (In-House): Similar to the test set, segmentations are "produced on clinical cases by operators who have undergone extensive classroom training and tests for efficacy, with an established track record of creating accurate segmentations." These were "verified by feedback loops of expert review and manual editing and correction."
• CTPelvic1K: "Widely accepted as ground truth, with three levels of annotation and expert review (by medical and/or graduate students, radiology fellows, and finally a senior consultant), and being used in support of public machine-learning segmentation research."

Ask a specific question about this device