Search Results

VEA Align:
This cloud-based software is intended for orthopedic applications in both pediatric and adult populations.

2D X-ray images acquired in EOS imaging's imaging systems is the foundation and resource to display the interactive landmarks overlayed on the frontal and lateral images. These landmarks are available for users to assess patient-specific global alignment.

For additional assessment, alignment parameters compared to published normative values may be available.

This product serves as a tool to aid in the analysis of spinal deformities and degenerative diseases, and lower limb alignment disorders and deformities through precise angle and length measurements. It is suitable for use with adult and pediatric patients aged 7 years and older.

Clinical judgment and experience are required to properly use the software.

spineEOS:
spineEOS is indicated for assisting healthcare professionals with preoperative planning of spine surgeries. The product provides access to EOS images with associated 3D datasets and measurements. spineEOS includes surgical planning tools that enable users to define a patient specific surgical strategy.

VEA Align:
VEA Align is a software indicated for assisting healthcare professionals with global alignment assessment through clinical parameters computation.

The product uses biplanar 2D X-ray images, exclusively generated by EOS imaging's EOS (K152788) and EOSedge (K202394) systems and generates an initial placement of the patient anatomic landmarks on the images using a machine learning-based algorithm. The user may adjust the landmarks to align with the patient's anatomy. Landmark locations require user validation. The clinical parameters communicated to the user are inferred from the landmarks and are recalculated as the user adjusts the landmarks. 3D datasets may be exported for use in spineEOS for surgical planning.

The product is hosted on a cloud infrastructure and relies on EOS Insight for support capabilities, such as user access control and data access. 2D X-ray image transmissions from healthcare institutions to the cloud are managed by EOS Insight. EOS Insight is classified as non-device Clinical Decision Support (CDS) software.

spineEOS:
spineEOS is indicated for assisting healthcare professionals with preoperative planning of spine surgeries. spineEOS provides access to EOS images with associated 3D datasets and measurements. spineEOS includes surgical planning tools that enable users to define a patient specific surgical strategy.

Here's a breakdown of the acceptance criteria and the study proving the device meets them, based on the provided FDA 510(k) clearance letter for VEA Align.

1. Table of Acceptance Criteria and Reported Device Performance

Note: The document states that "All the studies performed indicate acceptable performances of the algorithm for its intended population," implying that both acceptance criteria (Median error ≤ 3mm and 3rd Quartile ≤ 5mm) were met. Actual reported numerical values for the performance metrics are not explicitly provided in this document, only that the criteria were met.

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

• Test Set Sample Size: 361 patients.
• Data Provenance: Images were collected from EOS (K152788) and EOSedge (K202394) systems at a variety of sites from 2007-2023. The subgroups analysis includes "Data site location - Different US states," indicating that at least some of the data originates from the US. The document does not explicitly state whether the data was retrospective or prospective, but given the collection period (2007-2023), it is likely retrospective.

3. Number of Experts Used to Establish Ground Truth and Qualifications

• The document states that the ground truth was established by "EOS 3DServices reconstruction (model) from sterEOS (K172346)." This implies that the ground truth is derived from a previously cleared and validated 3D reconstruction system.
• The number and qualifications of experts involved in creating these "EOS 3DServices reconstruction" models are not specified in this document.

4. Adjudication Method for the Test Set

• The document does not describe an explicit adjudication method for the test set involving multiple human reviewers. The ground truth for the test set is established by the "EOS 3DServices reconstruction (model) from sterEOS."

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

• A formal MRMC comparative effectiveness study comparing human readers with and without AI assistance is not explicitly mentioned in this document. The performance evaluation is focused on the standalone AI algorithm compared to ground truth.

6. Standalone (Algorithm Only) Performance Study

• Yes, a standalone (algorithm only) performance study was done.
  • Description: "To assess the standalone performance of the AI algorithm of the VEA Align, the test was performed with:
    • A dedicated test data set containing different data from the training data set...
    • For each patient of this data set, a ground truth EOS 3DServices reconstruction (model) from sterEOS (K172346) that was available for comparison with VEA Align reconstruction generated by the AI algorithm."
  • This confirms that the study focused on the AI algorithm's performance without direct human intervention in the loop for the performance metrics measured.

7. Type of Ground Truth Used

• The ground truth used for the test set was based on "EOS 3DServices reconstruction (model) from sterEOS (K172346)." This refers to 3D anatomical models and landmark placements generated by a previously cleared medical imaging and reconstruction system (sterEOS). This can be categorized as a type of expert-system-derived ground truth, as sterEOS itself relies on validated methodologies and presumably expert input/validation in its operation.

8. Sample Size for the Training Set

• Training Set Sample Size: 10,376 X-ray images, with a total of 5,188 corresponding 3D reconstructions.

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

• The document states, "The AI algorithm was trained using 10,376 X-ray images and a total of 5,188 corresponding 3D reconstructions." It also notes that the images were collected from EOS and EOSedge systems. While it doesn't explicitly detail the method for establishing the ground truth for each training image, the context implies that these 3D reconstructions served as the ground truth. Similar to the test set, it's highly probable these "corresponding 3D reconstructions" were also derived from the established and validated EOS 3DServices/sterEOS pipeline.

Ask a specific question about this device

VEA Align:
This cloud-based software is intended for orthopedic applications in both pediatric and adult populations.
2D X-ray images acquired in EOS imaging's imaging systems is the foundation and resource to display the interactive landmarks overlayed on the frontal and lateral mages. These landmarks are available for users to assess patient-specific global alignment.
For additional assessment, alignment parameters compared to published normative values may be available.
This product serves as a tool to aid in the analysis of spinal deformities and degenerative diseases, and lower limb alignment disorders and deformities through precise and length measurements. It is suitable for use with adult and pediatric patients aged 7 years and older.
Clinical judgment and experience are required to properly use the software.

spineEOS:
spineEOS is indicated for assisting healthcare professionals with preoperative planning of spine surgeries. The product provides access to EOS images with associated 3D datasets and measurements. spineEOS includes surgical planning tools that enable users to define a patient specific surgical strategy.

VEA Align is a software indicated for assisting healthcare professionals with global alignment assessment through clinical parameters computation. The product uses biplanar 2D X-ray images, exclusively generated by EOS imaging's EOS (K152788) and EOSedge (K202394) systems and generates an initial placement of the patient anatomic landmarks on the images using a machine learning-based algorithm. The user may adjust the landmarks to align with the patient's anatomy. Landmark locations require user validation. The clinical parameters communicated to the user are inferred from the landmarks and are recalculated as the user adjusts the landmarks. 3D datasets may be exported for use in spineEOS for surgical planning. The product is hosted on a cloud infrastructure and relies on VEA Portal for support capabilities, such as user access control and data access. 2D X-ray image transmissions from healthcare institutions to the cloud are managed by VEA Portal is a Class | 510(k)-exempt device (LMD).

spineEOS is a software indicated for assisting healthcare professionals with preoperative planning of spine surgeries. EOS images (generated from EOS imaging's acquisition system) and associated 3D datasets are used as inputs of the software. The product manages clinical measurements and allows user to access surgical planning tools to define a patient specific surgical strategy. The product is indicated for adolescent and adult patients.

The provided text describes the performance data for the VEA Align device, focusing on the standalone performance of its AI algorithm.

Here's the breakdown of the acceptance criteria and the study proving the device meets them:

1. A table of acceptance criteria and the reported device performance

2. Sample size used for the test set and the data provenance

• Test set sample size: 538 patients.
• Data provenance: Not explicitly stated as country of origin, but the images were collected from EOS (K152788) and EOSedge (K202394) systems at a variety of sites. The subgroup analysis includes "US vs. OUS" (Outside US), implying international data. The data collection period was from 2007-2023. The study seems to be retrospective as it uses previously collected images.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

The document states that the ground truth for the test set was an "EOS 3DServices reconstruction (model) from sterEOS Workstation (K172346)". It does not explicitly state the number or qualifications of experts used to establish this ground truth for the test set. However, the nature of the sterEOS Workstation suggests that these 3D reconstructions are typically performed or validated by trained specialists.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set

The document does not specify an adjudication method for the test set ground truth. It relies on the "ground truth EOS 3DServices reconstruction (model) from sterEOS Workstation (K172346)."

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

No MRMC comparative effectiveness study was described where human readers' improvement with AI vs. without AI assistance was evaluated. The performance testing focused on the standalone performance of the AI algorithm. The VEA Align device involves a machine learning-based algorithm for initial landmark placement, but then explicitly states, "The user may adjust the landmarks to align with the patient's anatomy. Landmark locations require user validation." This implies a human-in-the-loop system, but the described performance study is primarily on the algorithm's initial accuracy, not human improvement.

6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done

Yes, a standalone performance test of the AI algorithm was done. The document explicitly states: "To assess the standalone performance of the Al algorithm of the VEA Align, the test was performed with..."

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

The ground truth used for the standalone algorithm performance was "a ground truth EOS 3DServices reconstruction (model) from sterEOS Workstation (K172346)". This suggests a reconstructed anatomical model derived from clinically used software, likely validated by trained operators or experts who generated that model previously.

8. The sample size for the training set

The AI algorithm was trained using 10,376 X-ray images and a total of 5,188 corresponding 3D reconstructions.

9. How the ground truth for the training set was established

The document states that the training data included "corresponding 3D reconstructions" presumably generated by sterEOS Workstation (K172346), similar to the test set ground truth. These 3D reconstructions would have been based on images from EOS systems and likely performed by trained personnel using the sterEOS Workstation. It's implied that these served as the ground truth for training the AI algorithm to generate its initial placements.

Ask a specific question about this device

spineEOS software is indicated for assisting healthcare professionals with preoperative planning of spine surgeries. spineEOS provides access to EOS images with associated 3D datasets and measurements. spineEOS includes surgical planning tools that enable users to define a patient specific surgical strategy.

spineEOS is a software indicated for assisting healthcare professionals with preoperative planning of spine surgeries. EOS images (generated from EOS imaging's acquisition system) and associated 3D datasets are used as inputs of the software. The product manages clinical measurements and allows user to access surgical planning tools to define a patient specific surgical strategy. The product is indicated for adolescent and adult patients.

The provided document is a 510(k) premarket notification for the device spineEOS. It primarily focuses on demonstrating substantial equivalence to a predicate device (K160407 spineEOS) rather than detailing a clinical study with acceptance criteria and performance data for a novel AI/ML-based device.

Therefore, the document does not contain the detailed information required to fully answer the request, specifically relating to a study proving the device meets acceptance criteria for new or modified AI/ML features. The information regarding performance data (Section 7) states that "Nonclinical performance testing performed on the subject device, spineEOS, supports substantial equivalence to the predicate device" and explicitly, "Determination of substantial equivalence is not based on an assessment of clinical performance data."

However, based on the provided text, I can extract information relevant to the comparison for substantial equivalence as presented, and highlight what is missing for a comprehensive answer on acceptance criteria and a study proving performance for a new AI/ML feature.

Here's an attempt to answer the questions based on the available information, noting where data is absent:

1. A table of acceptance criteria and the reported device performance

The document does not provide a table of acceptance criteria for specific performance metrics of a new AI/ML model. Instead, it presents a comparison for substantial equivalence between the modified spineEOS and its predicate (K160407 spineEOS). This comparison focuses on demonstrating that technological differences do not affect safety or effectiveness, not on proving a specific quantitative performance metric against a set acceptance criterion for a novel feature.

Here's a re-interpretation of "acceptance criteria" based on the "Substantially Equivalent?" column in Table 5-1, which represents the justification for equivalence rather than quantitative performance.

2. Sample size used for the test set and the data provenance

The document states "Nonclinical performance testing performed on the subject device, spineEOS, supports substantial equivalence to the predicate device." It also mentions "Validation activities" including "Validation of the multifunctional requirements in terms of design" and "Usability testing".

• Sample Size for Test Set: This information is not provided. The document makes no mention of a specific dataset size for evaluating the performance of the claimed features, especially for any new AI/ML components. Given the "nonclinical performance" and "not based on clinical performance data" statements, it's highly likely that a traditional "test set" in the context of an AI/ML study (i.e., a set of patient cases to evaluate algorithm performance) was not part of this 510(k) submission.
• Data Provenance: This information is not provided. (e.g., country of origin, retrospective or prospective).

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

This information is not provided. Since the submission states "Determination of substantial equivalence is not based on an assessment of clinical performance data," it's unlikely that such an expert-adjudicated ground truth dataset, typical for AI/ML device evaluations, was used. The focus was on demonstrating that the modified features (like the new Spondylolisthesis tool) function as intended and do not introduce new safety concerns compared to the predicate, rather than on proving diagnostic or measurement accuracy against a gold standard.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

This information is not provided.

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

No, an MRMC comparative effectiveness study was not done. The document explicitly states: "Determination of substantial equivalence is not based on an assessment of clinical performance data." The core of this 510(k) is demonstrating substantial equivalence based on technological comparisons and nonclinical performance (e.g., software verification and validation activities), not an MRMC study.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

This information is not explicitly provided. The document mentions "Validation of the multifunctional requirements in terms of design" and "Usability testing." The product, spineEOS, is presented as a tool "assisting healthcare professionals," implying a human-in-the-loop model for its primary use ("clinical judgment and experience are required to properly use the software"). There's no separate mention of a "standalone" algorithm-only performance evaluation, especially one related to diagnostic accuracy without human input.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)

Since the submission states it's not based on clinical performance data and focuses on substantial equivalence via nonclinical V&V, the concept of "ground truth" in the context of clinical accuracy (e.g., expert consensus, pathology) for performance claims is not applicable or provided. The "ground truth" for the V&V activities would be the design specifications and requirements.

8. The sample size for the training set

This information is not provided. The document does not describe the specific AI/ML algorithms used or their training. The changes highlighted (like the new Spondylolisthesis tool) are additions to existing functionality, and the filing focuses on their technological equivalence and safety, not on a new AI/ML model requiring a training dataset.

9. How the ground truth for the training set was established

This information is not provided, as there is no mention of a training set for an AI/ML model for this submission.

Summary of what the document does provide regarding performance and validation:

• Nonclinical performance testing: This includes design input review, unit testing, software integration, system integration, validation of multifunctional requirements in terms of design, and usability testing.
• Safety & Effectiveness Argument: The core argument is that the modified device, including new tools like "Spondylolisthesis" and "Rod Curvature Management," operates on principles similar to existing cleared functionalities (e.g., segmental alignment tool), and therefore does not raise new questions of safety or effectiveness.
• No Clinical Performance Data: The submission explicitly states that "Determination of substantial equivalence is not based on an assessment of clinical performance data." This significantly limits the type of "acceptance criteria" and "study" information available, as it implies a focus on V&V of software functionality and comparison to predicate rather than a clinical trial proving new performance claims.

Ask a specific question about this device

Using 3D data and models obtained with sterEOS workstation, spineEOS software is indicated for assisting healthcare professionals in viewing, measuring images as well as in preoperative planning of spine surgeries. The device includes tools for measuring spine anatomical components for placement of surgical implants. Clinical judgment and experience are required to properly use the software online.

spineEOS 1.0 allows surgeons to perform preoperative surgical planning of spine surgeries in case of Adolescent Idiopathic Scoliosis (AIS) or deformative spine. The software provides surgical tools for the correction of the curvature, for the placement of cages and for the achievement of osteotomies. The images displayed are x-rays from EOS System (K152788) and 3D model of the spine from sterEOS Workstation (K141137). spineEOS also displays preoperative parameters compared with reference values and updated values of parameters after planning. spineEOS is accessible on any computer via ONEFIT Management System (Class I device - Product code LMD - 510(k) Exempt) that provides a secure interface and storage through authentication mechanisms.

The FDA 510(k) summary for spineEOS provides some information regarding its performance data, but it does not contain a detailed study with acceptance criteria, specific reported device performance metrics, sample sizes, or information about experts and ground truth as requested.

The document primarily focuses on establishing substantial equivalence to a predicate device (Surgimap 2.0) by comparing intended use, indications, and technological characteristics.

Here's an analysis of what is available and what is missing from the provided text, structured according to your request:

1. A table of acceptance criteria and the reported device performance

• Missing from the document. The document states: "Software verification and validation testing were conducted and documentation was provided as recommended by FDA's Guidance for Industry and FDA Staff, 'Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices'." However, specific acceptance criteria or detailed results of these tests (e.g., accuracy of measurements, success rate of planning tools) are not provided in this 510(k) summary.

2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

• Missing from the document. The summary mentions "Software verification and validation testing," but does not specify the sample size of any test set or the provenance of the data used for such testing.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)

• Missing from the document. There is no mention of experts, ground truth establishment, or their qualifications for any validation testing.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

• Missing from the document. No information about adjudication methods for a test set is provided.

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

• Missing from the document. The document makes no mention of a multi-reader multi-case (MRMC) comparative effectiveness study. The focus is on demonstrating equivalence to the predicate device's existing functionality rather than quantifying human performance improvements.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

• Implied, but not detailed. The "Software verification and validation testing" would typically involve standalone performance testing of the algorithms and software functionalities. However, the specifics of these tests and their results are not detailed. The spineEOS is described as "assisting healthcare professionals," implying it's a human-in-the-loop device, but standalone testing of its components would be part of standard V&V. Again, no specific results are provided.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

• Missing from the document. As no specific performance study is detailed, the type of ground truth used is not mentioned.

8. The sample size for the training set

• N/A (or not explicitly stated as a "training set"). The spineEOS is a software for viewing, measuring, and planning based on existing 3D data and models (from sterEOS workstation). It's not described as a machine learning device that requires a distinct "training set" in the sense of a deep learning model. Its validation would focus on the accuracy of its measurements and the functionality of its planning tools against known standards or expert opinion, not on learning from a dataset.

9. How the ground truth for the training set was established

• N/A. Since a classical machine learning "training set" is not explicitly mentioned or implied for this type of device, the method for establishing its ground truth is not applicable in that context.

Ask a specific question about this device