Search Results

Formus Hip is a preoperative surgical planning software. It is intended to assist qualified medical professionals in the preoperative planning of orthopedic surgical procedures.

Formus Hip is a semi-automated Software as a Medical Device (SaMD) that allows pre-operative planning of primary total hip arthroplasty in real time using the Zimmer Biomet Taperloc G7 system. Using a series of algorithms, the software creates a 3D model and relevant measurements derived from the patient's pre-dimensioned CT scan. Formus Hip generates a 3D model without any user input. Additional algorithms fit the femoral stem and acetabular cup based on the patient anatomy. The software allows the user to adjust the plan interactively to achieve the desired clinical targets.

Here's a summary of the acceptance criteria and study details for the Formus Hip device, as extracted from the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

2. Sample Sizes and Data Provenance

Image Processing Accuracy (Test Set):

• Sample Size: 60 images (from 60 patients)
• Data Provenance: Acquired in the US (retrospective, implied by "images acquired... to validate accuracy")

Implant Sizing (Test Set):

• Sample Size: 133 images (from 133 patients)
• Data Provenance: Implied to be retrospective, acquired via a variety of common clinical CT scanners. Country of origin for this specific dataset is not explicitly stated but is implicitly within the context of demonstrating generalizability for the US patient population.

3. Number and Qualifications of Experts for Ground Truth (Test Set)

Image Processing Accuracy:

• Number of Experts: 3 radiologists (2 segmenters, 1 senior reviewer)
• Qualifications: US-board registered radiologists experienced in 3D image segmentation.

Implant Sizing:

• Number of Experts: Orthopaedic surgeons (specific number not given, but implied to be multiple).
• Qualifications: Orthopaedic surgeons.

4. Adjudication Method for the Test Set

Image Processing Accuracy:

• Method: 2+1 adjudication. Each bone surface was manually segmented by two radiologists. A third senior radiologist reviewed each pair of segmentations and selected the most accurate segmentation as the final manually segmented mesh (ground truth).

Implant Sizing:

• Method: Not explicitly described as an adjudication process between multiple experts. Ground truth was obtained by orthopaedic surgeons using traditional templating methods. This implies a single expert's determination per case or a consensus if multiple were involved, but the mechanism for resolving discrepancies isn't detailed.

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

• No, a MRMC comparative effectiveness study was not reported. The studies described focus on the standalone algorithm's performance against expert-derived ground truth.

6. Standalone Performance Study

• Yes, a standalone (algorithm only without human-in-the-loop performance) was performed.
  • The "Image Processing Accuracy" study compared automatically generated 3D models from Formus Hip to manually segmented 3D models (ground truth).
  • The "Implant Sizing" study compared implant sizes recommended by Formus Hip to "ground truth" implant sizes determined by orthopaedic surgeons using traditional templating methods.

7. Type of Ground Truth Used

Image Processing Accuracy:

• Type: Expert consensus (adjudicated manual segmentation by radiologists).

Implant Sizing:

• Type: Expert determination (implant sizes determined by orthopaedic surgeons using traditional templating methods).

8. Sample Size for the Training Set

• Image Segmentation Algorithm: CT scans of male and female subjects with typical and atypical bony anatomy between the ages of 21 and 94. The exact number is not provided, only the demographics.
• Statistical Shape Models: Segmented 3D models of male and female subjects with typical and atypical bony anatomy between the ages of 18 and 89. The exact number is not provided, only the demographics.

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

• The document states that the AI-based automatic image segmentation algorithm was "trained on CT scans" and statistical shape models were "trained on segmented 3D models". However, the specific process for establishing the ground truth (e.g., manual segmentation by experts, pathology confirmation, etc.) for these training datasets is not explicitly detailed in the provided text. It only mentions that training datasets are independent from testing and validation datasets and tracked in a single record file.

Ask a specific question about this device