(134 days)
The Scenium display and analysis software has been developed to aid the clinician in the assessment and quantification of pathologies taken from PET and SPECT scans.
The software is deployed via medical imaging workplaces and is organized as a series of workflows which are specific to use with particular drug and disease combinations.
The software aids in the assessment of human brain scans enabling automated analysis through quantification of mean pixel values located within standard regions of interest. It facilitates comparison with existing databases of normal patients and normal parameters derived from these databases, derived from FDG-PET, amyloid-PET, and SPECT studies, calculation of uptake ratios between regions of interest, and subtraction between two functional scans.
Scenium VE10 display and analysis software enables visualization and appropriate rendering of multimodality data, providing a number of features which enable the user to process acquired image data.
Scenium VE10 consists of three workflows:
- Database Comparison
- Ratio Analysis
- Subtraction
These workflows are used to assist the clinician with the visual evaluation, assessment and quantification of pathologies, such as dementia (i.e., Alzheimer's), movement disorders (i.e., Parkinson's) and seizure analysis (i.e., Epilepsy).
The modifications made to the Scenium VE10 software (K162339) to create the Scenium VE20 software include:
- The ability to create and support normal databases in the Striatal Analysis workflow
- DaTSCAN-SPECT normals database
- Improvements related to the analysis screen for reporting in Striatal Analysis
In addition, workflow structures changed within the VE20 release. Previously, the three workflows (Database Comparison, Ratio Analysis, and Subtraction) encompassed the Scenium software. Ratio Analysis has since been split into two separate workflows. Now, the following four workflows exist within Scenium VE20:
- Database Comparison
- Striatal Analysis
- Cortical Analysis
- Subtraction
These changes are based on current commercially available software features and do not change the technological characteristics of the device.
Scenium VE20 analysis software is intended to be run on commercially available software platforms such as the Siemens syngo.MI Workflow software platform (K150843).
The provided text is a 510(k) summary for the Scenium VE20 device, which is an image processing software for PET and SPECT scans. It primarily focuses on demonstrating substantial equivalence to a predicate device (Scenium VE10) rather than presenting a detailed study with specific acceptance criteria and performance metrics for the VE20 device itself.
Therefore, the document does not contain the detailed information required to fill out all sections of your request about acceptance criteria and a specific study proving the device meets them. The text states that "All testing has met the predetermined acceptance values" but does not elaborate on what those values were or what performance metrics were used to determine them for the Scenium VE20 specifically.
Here's what can be extracted and inferred, along with the information that is explicitly missing:
Acceptance Criteria and Reported Device Performance
| Acceptance Criteria | Reported Device Performance | Comments |
|---|---|---|
| Specific quantitative acceptance criteria for Scenium VE20 performance (e.g., accuracy, precision, sensitivity, specificity for pathology detection/quantification) | Not explicitly stated in the document. The document broadly states "All testing has met the predetermined acceptance values." This suggests internal performance criteria were met, but the details are not provided for the public record. | The document implies that the modifications to VE20 (support for DaTSCAN-SPECT normals database and analysis screen improvements) did not change the fundamental technological characteristics or intended use. Therefore, any performance met by VE10 would implicitly be carried over, but no specific performance numbers for VE20 are given. |
| Qualitative functional requirements (e.g., proper execution of workflows, accurate display of multimodality data) | "Verification and Validation activities have been successfully performed on the software package, including assurance that functions work as designed, performance requirements and specifications have been met..." | This confirms the software's functional integrity. |
| Safety and Effectiveness (e.g., no new issues of safety and effectiveness compared to predicate) | "There have been no changes that raise any new issues of safety and effectiveness as compared to the predicate device." | This is a regulatory acceptance criterion for substantial equivalence. |
| Compliance with standards (e.g., ISO 14971, ISO 13485, IEC 62304) | "Risk Management has been ensured via risk analyses in compliance with ISO 14971:2012... Siemens Medical Solutions USA, Inc. adheres to recognized and established industry standards for development including ISO 13485 and IEC 62304." | This indicates compliance with recognized standards. |
Study Details Provided
The document refers to "Verification and Validation activities" and "All testing" but does not describe a specific clinical or technical study designed to prove the device meets explicit acceptance criteria in the way a clinical trial or performance study would. Instead, it relies on demonstrating equivalence to an already cleared device.
-
Sample size used for the test set and the data provenance:
- Not explicitly stated for Scenium VE20. The document does not describe a separate clinical test set or its sample size.
- Data provenance: Not mentioned.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Not applicable / Not stated. No specific external "test set" and associated ground truth establishment process is described in this document for the Scenium VE20.
-
Adjudication method (e.g., 2+1, 3+1, none) for the test set:
- Not applicable / Not stated. No specific external "test set" and adjudication process is described.
-
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 explicitly mentioned or described as part of this 510(k) submission. The device (Scenium VE20) is primarily an image processing software for quantification and comparison, aiding clinicians, but the submission doesn't present it as an AI assistant in the typical sense that would necessitate an MRMC study comparing human performance with and without its aid. The improvements are related to expanding database and analysis capabilities.
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- Not explicitly described. The document states "Verification and Validation activities have been successfully performed on the software package, including assurance that functions work as designed, performance requirements and specifications have been met..." This implies internal testing of the algorithm's functions, but details of a formal "standalone" performance study are not provided. The device "aids the clinician" and its output facilitates "comparison," implying human interpretation remains central.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- Not explicitly stated. Given the nature of the software (quantification and comparison with normal databases), the ground truth for the underlying databases (FDG-PET, amyloid-PET, and SPECT studies, DaTSCAN-SPECT normals database) would likely have been established through clinical diagnosis, expert consensus, or follow-up outcomes, but this is not detailed in the context of VE20's "testing."
-
The sample size for the training set:
- Not explicitly stated for the Scenium VE20's development. The document mentions the "ability to create and support normal databases in the Striatal Analysis workflow" including a "DaTSCAN-SPECT normals database." The size and composition of these "normal databases" are not specified as "training sets" for a deep learning model, as the document doesn't explicitly state the use of AI/deep learning in the typical sense for image interpretation. This sounds more like statistical comparison to pre-existing normal population data.
-
How the ground truth for the training set was established:
- Not explicitly stated for the Scenium VE20's development. For the normal databases, the ground truth would inherently be "normal patients" (as described in the text "existing databases of normal patients"). How the "normal" status was confirmed for these patients in these databases is not elaborated upon in this submission.
In summary, this 510(k) submission focuses on demonstrating substantial equivalence to a predicate device by highlighting that the modifications do not alter the fundamental technological characteristics or intended use, and that internal verification and validation activities confirmed the software functions as designed and met its requirements. It does not provide the kind of detailed performance study and acceptance criteria that might be found in submissions for novel AI-powered diagnostic devices.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).