Search Results

iMAR is Siemens' iterative reconstruction software designed to reduce metal artifacts in CT images. It is a user-selected option, to be used with conventional reconstruction methods (WFBP or iterative reconstruction) on Siemens CT systems.

The amount of metal artifact reduction and corresponding improvement in image quality depends on a number of factors including: composition and size of the metal object, patient size, anatomical location and clinical practice.

It is recommended to perform reconstructions with iMAR enabled in addition to conventional reconstruction without iMAR.

iMAR (Metal Artifact Reconstruction) SW is a new software option, for Siemens Computed Tomography X-ray systems with the Somaris platform operating system. The iMAR software is designed to reduce metal artifacts caused by large and/or dense metal objects in computed tomography images. This submission includes the following device modifications:

• 1. New optional iMAR Software
• 2. A new specific Indication for Use Statement
• 3. Inclusion of a product Claims List

The iMAR software is based on a modification to the predicate device, previously cleared SOMATOM Definition AS Open using "MARIS" (Metal Implant Reduction in Image Space) VA46 (K130901) computed tomography system cleared under Premarket Notification K130901 on 01/02/2014.

The provided text does not contain detailed acceptance criteria or a specific study proving the device meets those criteria with the level of detail requested in the prompt. The document is a 510(k) summary for the iMAR software, which focuses on demonstrating substantial equivalence to a predicate device rather than presenting a full performance study with acceptance criteria.

However, based on the available information, I can extract and infer some points:

1. Table of Acceptance Criteria and Reported Device Performance

Acceptance Criteria (Inferred from general statements):

2. Sample Size for Test Set and Data Provenance

• Sample Size: Not specified. The document mentions "Evaluation of clinical images was conducted" but does not provide a number of cases or patients.
• Data Provenance: Not specified, but likely from Siemens' internal testing or clinical sites where their CT systems are used. It's a "non-clinical test" and "evaluation of clinical images," which might suggest retrospective data or internal prospective data collection.

3. Number of Experts and Qualifications for Ground Truth

• Number of Experts: "a certified radiologist" (singular).
• Qualifications: "certified radiologist." No years of experience are specified.

4. Adjudication Method

• Adjudication Method: Not specified. With only "a certified radiologist" mentioned for evaluation, formal adjudication with multiple experts seems unlikely.

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

• MRMC Study: No, an MRMC comparative effectiveness study is not mentioned. The evaluation was a qualitative judgment by a single radiologist to deem clinical images "acceptable." No comparison of human readers with vs. without AI assistance is detailed.

6. Standalone Performance Study

• Standalone Study: Yes, performance testing including "software verification/validation" and "non-clinical tests (integration and functional), including bench testing" was conducted. It states, "The test results show that all of the software specifications have met the acceptance criteria." The evaluation of clinical images by a radiologist also demonstrates standalone performance in a clinical context.

7. Type of Ground Truth Used

• Ground Truth Type: Expert consensus (from a single certified radiologist) on the acceptability of clinical images and achievement of "software specifications." It's primarily image quality assessment in the context of metal artifact reduction.

8. Sample Size for Training Set

• Sample Size for Training Set: Not specified. The document describes the software as "iterative Metal Artifact Reduction" and mentions "iterative reconstruction software," but it does not explicitly state that it's a machine learning-based AI that requires a training set in the modern sense. It refers to a "modification to the predicate device" and focuses on the "technological characteristics" of the algorithm (e.g., correcting beam hardening, scatter, undersampling, photon starvation in both projection and image domains). If it does involve a training set (e.g., for learned iterative reconstruction), the details are not provided.

9. How Ground Truth for Training Set was Established

• Ground Truth for Training Set: Not specified, as the presence of a training set itself is not clearly identified or detailed.

Ask a specific question about this device