(171 days)
Not Found
Yes
The device description explicitly states that the analysis module consists of "pre-trained convolutional neural networks (CNN)", which are a type of artificial intelligence/machine learning algorithm.
No.
The device is intended for automatic annotation, visualization, and quantification of segmentable brain structures from MRI scans to accelerate and improve the quantification of brain structures. It does not provide treatment or therapy.
No
The "Intended Use / Indications for Use" states that the device is for "automatic annotation, visualization, and quantification of segmentable brain structures" and to "accelerate and improve the quantification of brain structures". Additionally, the "Device Description" explicitly states: "The results from iQ-solutions™ are not intended to be used as a diagnostic tool".
Yes
The device is explicitly described as a "standalone software device" and its function is to process existing MRI scans, interacting with PACS systems. There is no mention of accompanying hardware or hardware components being part of the device itself.
Based on the provided information, this device is not an IVD (In Vitro Diagnostic).
Here's why:
- Intended Use: The intended use clearly states that the software is for "automatic annotation, visualization, and quantification of segmentable brain structures from a set of brain MRI scans." It is intended to "accelerate and improve the quantification of brain structures that would otherwise require a manual process." This describes a tool for image analysis and measurement, not a test performed in vitro (outside the living body) on biological specimens to provide diagnostic information.
- Device Description: The device processes "brain MRI scans," which are medical images, not biological samples. It interacts with PACS systems and radiology software, which are part of the medical imaging workflow.
- Lack of Biological Specimen Analysis: IVD devices are designed to analyze biological specimens such as blood, urine, tissue, etc., to provide information about a patient's health status. This device does not perform any such analysis.
- Results are not for Diagnosis: The description explicitly states, "The results from iQ-solutions™ are not intended to be used as a diagnostic tool, and the interpretation of measurements contained in the iQ-solutions™ report remains entirely the responsibility of the qualified radiologists or physicians managing the individual patient's care." This further reinforces that it is a tool to aid in the interpretation of images, not a diagnostic test itself.
In summary, iQ-solutions™ is a medical image analysis software device, not an in vitro diagnostic device.
No
The document states "Not Found" under the "Control Plan Authorized (PCCP) and relevant text" section, indicating that a PCCP was not authorized for this device.
Intended Use / Indications for Use
iQ-solutions™ is a software medical device intended for automatic annotation, visualization, and quantification of segmentable brain structures from a set of brain MRI scans. It is intended to accelerate and improve the quantification of brain structures that would otherwise require a manual process of identifying, and measuring regions of interest in brain MRI scans.
iQ-solutions TM consists of both cross-sectional and longitudinal analysis pipelines.
· The cross-sectional pipeline is intended to conduct structure segmentation and volume analysis on brain MRI scans at a single time point.
· The longitudinal pipeline is intended to conduct volume change analysis on a single patient's MRI scans acquired on the same scanner, with consistent image acquisition protocol and with contrast at two different time points. The results of the cross-sectional pipeline cannot be compared with the results of the longitudinal pipeline.
Product codes
QIH, LLZ
Device Description
iQ-solutions™ is a standalone software device that processes brain MRI scans to outline and quantify the brain structures described in the intended use. The iQ-solutions™ software interacts with the user's picture archiving and communication system (PACS) to receive scans and returns the results to the same destination or other user-specified radiology software system or database.
iQ-solutions™ analysis module consists of pre-trained convolutional neural networks (CNN) that have been verified and validated to segment the specific brain structures and create binary masks accordingly using the incoming head MRI scans. Each convolutional neural network is coupled with a pre-processing component that transforms the MRI scans to the standard position and contrast; and a post-processing component that prepares for the output annotations, statistical calculation, and the input of the next analysis component in either analysis pipeline.
iQ-solutions™ operates together with a medical image routing software or an integration platform that can connect with iQ-solutions™ and a PACS system that can fulfill the requirements of iQsolutions™.
MRI scans are sent to iQ-solutions™ by means of transmission functions within the user's PACS system. Upon completion of processing, iQ-solutions™ returns results to the user's PACS or other userspecified radiology software system or database.
The results from iQ-solutions™ are not intended to be used as a diagnostic tool, and the interpretation of measurements contained in the iQ-solutions™ report remains entirely the responsibility of the qualified radiologists or physicians managing the individual patient's care.
iQ-solutions™ takes as input brain MRI scans. During image processing, cross-sectional and longitudinal segmentation is performed, and measurements of the segmented brain structures are provided. The resulting outputs are generated and provided in the form of quantitative reports and the annotation masks of specific segmentations in DICOM format. The reports are in PDF format which are returned to the PACS, together with DICOM images that display colored overlays of specific brain substructures and, where relevant, white matter hyperintensities superimposed on duplicates of the original DICOM images (with watermarks to inform users that these images are not to be used for diagnostic purposes). iQ-solutions™ also provide dynamic reports in a webpage format in which the scan overview panel becomes scrollable, allowing the user to browse through different slices of the scan in the sagittal, coronal, and axial planes.
Mentions image processing
Yes
Mentions AI, DNN, or ML
iQ-solutions™ analysis module consists of pre-trained convolutional neural networks (CNN) that have been verified and validated to segment the specific brain structures and create binary masks accordingly using the incoming head MRI scans.
Segmentation by machine learning and deep learning algorithms (supervised segmentation with Convolutional Neural Networks)
Input Imaging Modality
MRI scans, Pre-contrast T1-weighted MRI scans, Fluid-attenuated inversion recovery (FLAIR) MRI scans, Post-contrast T1-weighted MRI scans
Anatomical Site
Brain, Head
Indicated Patient Age Range
Not Found
Intended User / Care Setting
trained professionals in hospitals, imaging centers or in imageprocessing labs
Description of the training set, sample size, data source, and annotation protocol
The algorithms were developed based on SNAC in-house datasets collected from more than 1500 patients over a period of ten years, commencing July 2012, acquired from a wide range of clinical scanners, including 1.5T and 3T scanners from Siemens (1750 scans in total), GE (2214 scans in total) and Philips (1593 scans in total). Additionally, publicly available MRI datasets from healthy people were included for both training and testing, including 2629 scans obtained from 2159 subjects. The in-house dataset contains patients diagnosed with multiple sclerosis, and the patients are also with brain atrophy and related cognitive dysfunctions. The scans of healthy people were also included in the training process to avoid bias.
Training data sample sizes for each module:
Sequence Classification: 4862 cases
Brain Extraction: 1843 cases
Scaling Factor Estimation: 2986 cases
White matter hyperintensity Segmentation: 1858 cases
Contrast-Enhancing Lesion Segmentation: 79 cases
Lesion Inpainting: 449 cases
Brain Tissue Segmentation: 4179 cases
Brain Volume Change Estimation: 1487 cases
WMH Lesion Activity: 116 cases
Cortical Lobar and Subcortical Structure Segmentation: Not applicable
Test-ReTest Dataset: Not applicable
Comprehensive Test Set: Not applicable
All data used for development were manually labelled to form the ground truth. The labelled items included sequence types (pre-contrast or post-contrast, T1-weighted or FLAR, brainbased or spine-based), skull and scaling factor based on skulls, skull stripped head, brain tissues (white matter, gray matter and CSF), white matter hyperintensities in patients, contrast-enhancing lesions, white matter hyperintensity lesion changes (new lesions or enlarged lesions), cortical gray matter lobes, thalamus, hippocampus, brain volume change. Specifically, the annotations for skullstripping, brain tissues, cortical gray matter lobes, thalamus, hippocampus, and brain volume changes were originally generated using widely used software items such as FSL and FreeSurfer, and the annotations were manually reviewed and accepted by trained neuroimaging analysts. The annotations for white matter hyperintensities (including pre-contrast ones, contrast-enhancing ones, and lesion changes) were manually annotated by trained neuroimaging analysts. All the annotations are further reviewed by senior neuroimaging analysts according to SNAC SOP.
When used for development, the data were split into training, validation, and test splits. For the training/validation/test splits of each task, the same subject was ensured to be included in only one of the three splits, therefore it is ensured the same subject was not used for training and evaluation.
Description of the test set, sample size, data source, and annotation protocol
The algorithms were developed based on SNAC in-house datasets collected from more than 1500 patients over a period of ten years, commencing July 2012, acquired from a wide range of clinical scanners, including 1.5T and 3T scanners from Siemens (1750 scans in total), GE (2214 scans in total) and Philips (1593 scans in total). Additionally, publicly available MRI datasets from healthy people were included for both training and testing, including 2629 scans obtained from 2159 subjects. The in-house dataset contains patients diagnosed with multiple sclerosis, and the patients are also with brain atrophy and related cognitive dysfunctions. The scans of healthy people were also included in the training process to avoid bias.
Test data sample sizes for each module:
Sequence Classification: 1207 cases
Brain Extraction: 458 cases
Scaling Factor Estimation: 527 cases
White matter hyperintensity Segmentation: 464 cases
Contrast-Enhancing Lesion Segmentation: 737 cases
Lesion Inpainting: 80 cases
Brain Tissue Segmentation: 500 cases
Brain Volume Change Estimation: 166 cases
WMH Lesion Activity: 64 cases
Cortical Lobar and Subcortical Structure Segmentation: 1504 cases
Test-ReTest Dataset: 120 cases
Comprehensive Test Set: 81 cases
All data used for development were manually labelled to form the ground truth. The labelled items included sequence types (pre-contrast or post-contrast, T1-weighted or FLAR, brainbased or spine-based), skull and scaling factor based on skulls, skull stripped head, brain tissues (white matter, gray matter and CSF), white matter hyperintensities in patients, contrast-enhancing lesions, white matter hyperintensity lesion changes (new lesions or enlarged lesions), cortical gray matter lobes, thalamus, hippocampus, brain volume change. Specifically, the annotations for skullstripping, brain tissues, cortical gray matter lobes, thalamus, hippocampus, and brain volume changes were originally generated using widely used software items such as FSL and FreeSurfer, and the annotations were manually reviewed and accepted by trained neuroimaging analysts. The annotations for white matter hyperintensities (including pre-contrast ones, contrast-enhancing ones, and lesion changes) were manually annotated by trained neuroimaging analysts. All the annotations are further reviewed by senior neuroimaging analysts according to SNAC SOP.
When used for development, the data were split into training, validation, and test splits. For the training/validation/test splits of each task, the same subject was ensured to be included in only one of the three splits, therefore it is ensured the same subject was not used for training and evaluation.
Besides the data used for development, 81 extra patients were reserved as the Comprehensive Test Set, which is used to evaluate the performance of all the analysis components.
Summary of Performance Studies
Non-clinical tests were conducted. Software "bench" testing was performed to evaluate the performance and functionality of the technical characteristics of iQ-solutions™. All testable requirement Specifications and the Risk Analysis have been successfully verified and traced. No clinical tests were conducted.
Performance results for various modules:
Sequence Classification: Accuracy = 100%
Brain Extraction: DICE = 0.982
Scaling Factor Estimation: STD = 0.0096
White matter hyperintensity Segmentation: DICE = 0.789
Contrast-Enhancing Lesion Segmentation: DICE = 0.790
Lesion Inpainting: PSNR = 30.79dB
Brain Tissue Segmentation: DICE = 0.972
Brain Volume Change Estimation: R² = 0.869
WMH Lesion Activity: R² = 0.833
Cortical Lobar and Subcortical Structure Segmentation: R² = 0.833
Substructure Volume Change Estimation: STD = 0.0102
All algorithms within iQ-solutions™ were verified to have generated results/metrics for compatible scans that met the pre-determined acceptance level(s).
Key Metrics
For classification modules, accuracy is higher than 0.9 to be accepted; for segmentation modules, DICE score is higher than 0.6 (for brains the DICE should be higher than 0.9) or R² is higher than 0.7 to be accepted. For modules that are lack of ground truths (scaling factor estimation, lesion inpainting, and substructure volume change estimation), the metrics are better than previously accepted results to be accepted. For all models, the manual acceptable rate should be higher than 80% to call it acceptable.
Specific results:
Sequence Classification: Accuracy = 100%, Acceptable Rate = 100%
Brain Extraction: DICE = 0.982, Acceptable Rate = 99.3%
Scaling Factor Estimation: STD = 0.0096, Acceptable Rate = 100%
White matter hyperintensity Segmentation: DICE = 0.789, Acceptable Rate = 98.8%
Contrast-Enhancing Lesion Segmentation: DICE = 0.790, Acceptable Rate = 92.1%
Lesion Inpainting: PSNR = 30.79dB, Acceptable Rate = 97.5%
Brain Tissue Segmentation: DICE = 0.972, Acceptable Rate = 100%
Brain Volume Change Estimation: R² = 0.869, Acceptable Rate = 96.3%
WMH Lesion Activity: R² = 0.833, Acceptable Rate = 97.5%
Cortical Lobar and Subcortical Structure Segmentation: R² = 0.833, Acceptable Rate = 96.0%
Substructure Volume Change Estimation: STD = 0.0102, Acceptable Rate = 98.8%
Predicate Device(s)
Reference Device(s)
Not Found
Predetermined Change Control Plan (PCCP) - All Relevant Information
Not Found
§ 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).
0
Image /page/0/Picture/0 description: The image contains the logo of the U.S. Food and Drug Administration (FDA). On the left is the Department of Health & Human Services logo. To the right of that is the FDA logo, which is a blue square with the letters "FDA" in white. To the right of the blue square is the text "U.S. FOOD & DRUG ADMINISTRATION" in blue.
December 18, 2023
Sydney Neuroimaging Analysis Centre Pty Ltd % Belinda Dowsett Associate Director, Medical Devices / IVD PharmaLex Pty Ltd Suite 10.04. 1 Chandos Street St Leonards, NSW 2064 Australia
Re: K231929
Trade/Device Name: iQ-solutions™ Regulation Number: 21 CFR 892.2050 Regulation Name: Medical image management and processing system Regulatory Class: Class II Product Code: QIH, LLZ Dated: September 19, 2023 Received: September 19, 2023
Dear Belinda Dowsett:
We have reviewed your section 510(k) premarket notification of intent to market the device referenced above and have determined the device is substantially equivalent (for the indications for use stated in the enclosure) to legally marketed predicate devices marketed in interstate commerce prior to May 28, 1976, the enactment date of the Medical Device Amendments, or to devices that have been reclassified in accordance with the provisions of the Federal Food, Drug, and Cosmetic Act (the Act) that do not require approval of a premarket approval application (PMA). You may, therefore, market the device, subject to the general controls provisions of the Act. Although this letter refers to your product as a device, please be aware that some cleared products may instead be combination products. The 510(k) Premarket Notification Database available at https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm identifies combination product submissions. The general controls provisions of the Act include requirements for annual registration, listing of devices, good manufacturing practice, labeling, and prohibitions against misbranding and adulteration. Please note: CDRH does not evaluate information related to contract liability warranties. We remind you, however, that device labeling must be truthful and not misleading.
If your device is classified (see above) into either class II (Special Controls) or class III (PMA), it may be subject to additional controls. Existing major regulations affecting your device can be found in the Code of Federal Regulations, Title 21, Parts 800 to 898. In addition, FDA may publish further announcements concerning your device in the Federal Register.
1
Additional information about changes that may require a new premarket notification are provided in the FDA guidance documents entitled "Deciding When to Submit a 510(k) for a Change to an Existing Device" (https://www.fda.gov/media/99812/download) and "Deciding When to Submit a 510(k) for a Software Change to an Existing Device" (https://www.fda.gov/media/99785/download).
Your device is also subject to, among other requirements, the Quality System (QS) regulation (21 CFR Part 820), which includes, but is not limited to, 21 CFR 820.30. Design controls; 21 CFR 820.90. Nonconforming product; and 21 CFR 820.100, Corrective and preventive action. Please note that regardless of whether a change requires premarket review, the QS regulation requires device manufacturers to review and approve changes to device design and production (21 CFR 820.30 and 21 CFR 820.70) and document changes and approvals in the device master record (21 CFR 820.181).
Please be advised that FDA's issuance of a substantial equivalence determination does not mean that FDA has made a determination that your device complies with other requirements of the Act or any Federal statutes and regulations administered by other Federal agencies. You must comply with all the Act's requirements, including, but not limited to: registration and listing (21 CFR Part 807); labeling (21 CFR Part 801); medical device reporting of medical device-related adverse events) (21 CFR Part 803) for devices or postmarketing safety reporting (21 CFR Part 4, Subpart B) for combination products (see https://www.fda.gov/combination-products/guidance-regulatory-information/postmarketing-safety-reportingcombination-products); good manufacturing practice requirements as set forth in the quality systems (QS) regulation (21 CFR Part 820) for devices or current good manufacturing practices (21 CFR Part 4, Subpart A) for combination products; and, if applicable, the electronic product radiation control provisions (Sections 531-542 of the Act); 21 CFR Parts 1000-1050.
Also, please note the regulation entitled, "Misbranding by reference to premarket notification" (21 CFR 807.97). For questions regarding the reporting of adverse events under the MDR regulation (21 CFR Part 803), please go to https://www.fda.gov/medical-device-safety/medical-device-reportingmdr-how-report-medical-device-problems.
2
For comprehensive regulatory information about mediation-emitting products, including information about labeling regulations, please see Device Advice (https://www.fda.gov/medicaldevices/device-advice-comprehensive-regulatory-assistance) and CDRH Learn (https://www.fda.gov/training-and-continuing-education/cdrh-learn). Additionally, you may contact the Division of Industry and Consumer Education (DICE) to ask a question about a specific regulatory topic. See the DICE website (https://www.fda.gov/medical-device-advice-comprehensive-regulatoryassistance/contact-us-division-industry-and-consumer-education-dice) for more information or contact DICE by email (DICE@fda.hhs.gov) or phone (1-800-638-2041 or 301-796-7100).
Sincerely,
Daniel M. Krainak, Ph.D. Assistant Director DHT8C: Division of Radiological Imaging and Radiation Therapy Devices OHT8: Office of Radiological Health Office of Product Evaluation and Quality Center for Devices and Radiological Health
Enclosure
3
DEPARTMENT OF HEALTH AND HUMAN SERVICES Food and Drug Administration
Indications for Use
Form Approved: OMB No. 0910-0120 Expiration Date: 06/30/2023 See PRA Statement below.
510(k) Number (if known) K231929
Device Name iO-solutionsTM
Indications for Use (Describe)
iQ-solutions™ is a software medical device intended for automatic annotation, visualization, and quantification of segmentable brain structures from a set of brain MRI scans. It is intended to accelerate and improve the quantification of brain structures that would otherwise require a manual process of identifying, and measuring regions of interest in brain MRI scans.
iQ-solutions TM consists of both cross-sectional and longitudinal analysis pipelines.
· The cross-sectional pipeline is intended to conduct structure segmentation and volume analysis on brain MRI scans at a single time point.
· The longitudinal pipeline is intended to conduct volume change analysis on a single patient's MRI scans acquired on the same scanner, with consistent image acquisition protocol and with contrast at two different time points. The results of the cross-sectional pipeline cannot be compared with the results of the longitudinal pipeline.
Type of Use (Select one or both, as applicable)
Prescription Use (Part 21 CFR 801 Subpart D)
Over-The-Counter Use (21 CFR 801 Subpart C)
CONTINUE ON A SEPARATE PAGE IF NEEDED.
This section applies only to requirements of the Paperwork Reduction Act of 1995.
DO NOT SEND YOUR COMPLETED FORM TO THE PRA STAFF EMAIL ADDRESS BELOW.
The burden time for this collection of information is estimated to average 79 hours per response, including the time to review instructions, search existing data sources, gather and maintain the data needed and complete and review the collection of information. Send comments regarding this burden estimate or any other aspect of this information collection, including suggestions for reducing this burden, to:
Department of Health and Human Services Food and Drug Administration Office of Chief Information Officer Paperwork Reduction Act (PRA) Staff PRAStaff@fda.hhs.gov
"An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB number."
4
510(k) Summary – iQ-solutions™ 5
General Information 5.1
| 510(k) Sponsor: | Sydney Neuroimaging Analysis Centre Pty Ltd (SNAC) |
|---|---|
| Address: | Level 1, 94 Mallett Street |
| Camperdown NSW 2050 | |
| Australia | |
| Contact Person: | Dr Tim (Chenyu) Wang |
| Director of Operations, Sydney Neuroimaging Analysis Centre | |
| Contact Information: | Email: tim@snac.com.au |
| Phone: + 61 2 9114 4187 | |
| Date Prepared: | 16 Dec 2023 |
5.2 Device Name and Classification
| Device Name: | iQ-solutions™ |
|---|---|
| Trade Name: | iQ-solutions™ |
| Common Name: | Medical Image Processing Software |
| Classification Name | System, Image processing, Radiological |
| Classification Panel: | Radiology |
| CFR Section: | 21 CFR §892.2050 |
| Device Class: | Class II |
| Product Code | QIH, LLZ |
5.3 Predicate Device
| Product Name: | icobrain |
|---|---|
| Device Trade Name: | Icobrain |
| 510(k) Number | K192130 |
| Clearance Date: | December 13, 2019 |
| Manufacturer: | icometrix NV |
| Common Name | Medical Image Processing Software |
| Classification Name: | System, Image processing, Radiological |
| Classification Panel: | Radiology |
| CFR Section | 21 CFR §892.2050 |
5
| Device Class | Class II |
|---|---|
| Product Code | LLZ |
5.4 Standards Applied
iQ-solutions™ software has been designed, developed, and tested by the following product and process standards:
- . ISO 13485:2016 Medical devices — Quality management systems — Requirements for regulatory purposes
- ISO 14971:2019 Medical devices – Application of risk management to medical devices
- IEC 62304:2006 + A1:2015 Medical device software software life cycle processes ●
- IEC 62366-1:2015 Medical devices – Part 1: Application of usability engineering to medical devices
- ISO 12052:2006 Digital Imaging and communication in Medicine (DICOM)
- CFR 21 Part 820 Quality System Regulation for Medical Devices
Device Description 5.5
The following flowchart describes the iQ-solutions™ process and analysis:
Image /page/5/Figure/13 description: The image is a flowchart that shows the steps involved in image processing. The first step is to input images in DICOM format. The next steps are pre-processing, image processing, and output generation. The final step is to generate a final report in PDF format, DICOM output images, and dynamic reports in a webpage format.
iQ-solutions™ is a standalone software device that processes brain MRI scans to outline and quantify the brain structures described in the intended use. The iQ-solutions™ software interacts with the user's picture archiving and communication system (PACS) to receive scans and returns the results to the same destination or other user-specified radiology software system or database.
iQ-solutions™ analysis module consists of pre-trained convolutional neural networks (CNN) that have been verified and validated to segment the specific brain structures and create binary masks accordingly using the incoming head MRI scans. Each convolutional neural network is coupled with a pre-processing component that transforms the MRI scans to the standard position and contrast; and a post-processing component that prepares for the output annotations, statistical calculation, and the input of the next analysis component in either analysis pipeline.
iQ-solutions™ operates together with a medical image routing software or an integration platform that can connect with iQ-solutions™ and a PACS system that can fulfill the requirements of iQsolutions™.
MRI scans are sent to iQ-solutions™ by means of transmission functions within the user's PACS system. Upon completion of processing, iQ-solutions™ returns results to the user's PACS or other userspecified radiology software system or database.
The results from iQ-solutions™ are not intended to be used as a diagnostic tool, and the interpretation of measurements contained in the iQ-solutions™ report remains entirely the responsibility of the qualified radiologists or physicians managing the individual patient's care.
6
iQ-solutions™ takes as input brain MRI scans. During image processing, cross-sectional and longitudinal segmentation is performed, and measurements of the segmented brain structures are provided. The resulting outputs are generated and provided in the form of quantitative reports and the annotation masks of specific segmentations in DICOM format. The reports are in PDF format which are returned to the PACS, together with DICOM images that display colored overlays of specific brain substructures and, where relevant, white matter hyperintensities superimposed on duplicates of the original DICOM images (with watermarks to inform users that these images are not to be used for diagnostic purposes). iQ-solutions™ also provide dynamic reports in a webpage format in which the scan overview panel becomes scrollable, allowing the user to browse through different slices of the scan in the sagittal, coronal, and axial planes.
In this release of iQ-solutions™, the two types of analysis pipelines (cross-sectional study and longitudinal study) are conducted, and different reports are generated together with the generated masks, including:
iQ-MS™ PDF report:
- Normalized volume and volume change of the whole brain (sum of white matter and gray matter)
- Normalized volume and volume change of gray matter
- Normalized volume and volume change of the thalami
- Unnormalized volume and volume changes of FLAIR white matter hyperintensities (if any)
- Unnormalized volume of T1 post-contrast white matter hyperintensities
- Brain and brain substructure volumes compared with a healthy cohort.
iQ-COG™ PDF report:
- Normalized volume and volume change of the whole brain (sum of white matter and gray matter)
- Normalized volume and volume change of cortical gray matter
- Normalized volume of the cortex in brain lobes (frontal, parietal, temporal, occipital)
- . Normalized volume and volume change of the hippocampi
- Asymmetry index of left and right cortex in lobes and left and right hippocampi
- Unnormalized volume and volume changes of FLAIR white matter hyperintensities (if any)
- . Brain and brain substructure volumes compared with a healthy cohort.
5.6 Indications for Use
iQ-solutions™ is a software medical device intended for automatic annotation, visualization, and quantification of segmentable brain structures from a set of brain MRI scans. It is intended to accelerate and improve the quantification of brain structures that would otherwise require a manual process of identifying, annotating, and measuring regions of interest in brain MRI scans.
iQ-solutions™ consists of both cross-sectional and longitudinal analysis pipelines.
- . The cross-sectional pipeline is intended to conduct structure segmentation and volume analysis on brain MRI scans at a single time point.
- . The longitudinal pipeline is intended to conduct volume change analysis on a single patient's MRI scans acquired on the same scanner, with consistent image acquisition protocol and with consistent contrast at two different time points.
The results of the cross-sectional pipeline cannot be compared with the results of the longitudinal pipeline.
7
5.7 Comparison with Predicate Device
iQ-solutions™ (subject device) has the same intended use as icobrain 5.0 (K192130) (predicate device) as both are intended for automatic labelling, visualization and volumetric quantification of segmentable brain structures based on three-dimensional medical images. The devices both take 3D images of the brain as input and generate an electronic report with similar quantitative information. The principal difference between the devices is that, in addition to MRI images, the predicate icobrain 5.0 (K192130) is able to process non-contrast CT (NCCT) images (for 'icobrain cross'). This difference does not result in a new intended use; and does not impact the safety or effectiveness of the subject device.
The subject device and the predicate device have different technological characteristics with respect to software features, functionalities and core algorithms. These different technological characteristics do not impact the safety or effectiveness of the subject device: specifically, performance data confirms that the subject device is as safe and effective as the predicate device.
Table 2 shows the comparison between the subject device and its predicate and no major technological differences were found between the two systems that raise new issues of safety and/or effectiveness. Thus, the iQ-solutions™ is substantially equivalent to the predicate device.
8
Table 2: Comparison table for iQ-solutions™ as the subject device and icobrain 5.0 (K192130) as the Predicate Device.
| Characteristic | Subject device:
iQ-solutions™ | Predicate Device:
icobrain 5.0 (K192130) | Summary |
|-------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Intended Use | iQ-solutions™ is a software medical device
intended for automatic annotation,
visualization, and quantification of
segmentable brain structures from a set of
brain MRI scans. It is intended to accelerate
and improve the quantification of brain
structures that would otherwise require a
manual process of identifying, annotating, and
measuring regions of interest in brain MRI
scans.
iQ-solutions™ consists of both cross-sectional
and longitudinal analysis pipelines.
The cross-sectional pipeline is intended to
conduct structure segmentation and
volume analysis on brain MRI scans at a
single time point. The longitudinal pipeline is intended to
conduct volume change analysis on a single
patient's MRI scans acquired on the same
scanner, with a consistent image
acquisition protocol and with consistent
contrast at two different time points. The results of the cross-sectional pipeline
cannot be compared with the results of the
longitudinal pipeline. | icobrain 5.0 is intended for automatic labelling,
visualization and volumetric quantification of
segmentable brain structures from a set of
MRI or Non-Contrast CT (NCCT) images. This
software is intended to automate the current
manual process of identifying, labelling and
quantifying the volume of segmentable brain
structures identified on MRI or NCCT images.
icobrain 5.0 consists of two distinct image
processing pipelines: icobrain 5.0 cross and
icobrain 5.0 long.
icobrain 5.0 cross is intended to provide
volumes from MRI or NCCT images acquired
at a single brain time point. icobrain 5.0 long is intended to provide
changes in volumes between two MRI
images that were brain acquired on the
same scanner, with the same image
acquisition protocol and with the same
contrast at two different time points. The results of the icobrain 5.0 cross cannot be
compared with the results of icobrain 5.0 long. | Intended use is the same.
There is a difference in functionality that does
not constitute a major difference.
Both devices are intended for automatic
labelling, visualization and quantification of
segmentable brain structures.
Both devices have two pipelines that are
intended for cross-sectional analysis and
longitudinal analysis, which are intended for
images from a single time point and different
time points respectively. The results from the
pipelines cannot be compared.
iQ-solutions™ works with MRI images only,
while icobrain 5.0 works with both MRI
images and NCCT images (only for icobrain 5.0
cross). This difference does not result in a new
intended use; rather, use with NCCT only
broadens the application to scenarios in which
NCCT is the preferred input. This difference
does not affect the safety or effectiveness of
the subject device. |
| Technological
Characteristics | Software package, with off-the-shelf
software
Operates on off-the-shelf hardware
(multiple vendors) | Software package, with off-the-shelf
software
Operates on off-the-shelf hardware
(multiple vendors)
DICOM compatible | Same. |
| Characteristic | Subject device:
iQ-solutions™ | Predicate Device:
icobrain 5.0 (K192130) | Summary |
| | DICOM compatible Segmentation by machine learning and
deep learning algorithms (supervised
segmentation with Convolutional Neural
Networks) Image pre-processing components Measurement calculation Report generation techniques | Segmentation by classical machine learning
and deep learning (supervised voxel
classification with Convolutional Neural
Networks) Image pre-processing components Measurement calculation Report generation techniques | |
| Anatomical Region
of Interest | Head | Head | Same |
| Product Input
(Data Acquisition
Protocol) | Pre-contrast T1-weighted MRI scans from
single or multiple time points, Fluid-attenuated inversion recovery (FLAIR)
MRI scans from single or multiple time
points, and Post-contrast T1-weighted MRI scans from
a single time point. | T1-weighted and fluid-attenuated inversion
recovery (FLAIR) MRI images from a single
or multiple time points Non-contrast CT (NCCT) from a single time
point. | Similar.
Both the subject device and the predicate
device process T1 and FLAIR MRI images from
single or multiple time points.
NCCT may be used with icobrain 5.0 but not the
subject device; lack of this additional input
compatibility does not impact the safety or
clinical function of the subject device. |
| Product Output | Multiple electronic reports with volumetric
information on brain structures and white
matter hyperintensities (if any) Automatically compares results to
reference percentile data and to prior scans
when available Annotated DICOM images that can be
displayed on DICOM workstations and
Picture Archive and Communications
Systems (PACS) | Multiple electronic reports with volumetric
information on brain structures and
hyperintensities and midline shift (for NCCT
only) Automatically compares results to a
reference healthy population data and to
prior scans when available Annotated DICOM images that can be
displayed on DICOM workstations and
Picture Archive and Communications
Systems (PACS) | Similar.
icobrain 5.0 additionally outputs the degree of
'midline shift' but this is calculated on NCCT
(relevant to specific clinical scenarios only). As
iQ-solutions™ does not analyze NCCT, this
difference does not impact the safety of the
subject device and is not clinically significant |
| Characteristic | Subject device:
iQ-solutions™ | Predicate Device:
icobrain 5.0 (K192130) | Summary |
| Findings Covered | MR:
Normalized volume and volume changes of the whole brain Normalized volume and volume changes of gray matter Normalized volume and volume changes of cortical gray matter Normalized cortical volumes of cortex in brain lobes (frontal, temporal, parietal and occipital) Normalized volume and volume changes in the hippocampi Normalized volume and volume changes of the thalamus Asymmetry index of structures, including cortex in lobes and hippocampus Unnormalized volume and volume changes of FLAIR white matter hyperintensities Unnormalized volume and volume changes of T1-w white matter hyperintensities (in post-contrast T1-w only) | MR:
Normalized volume and volume changes of the whole brain Normalized volume and volume changes of gray matter Normalized volume and volume changes of cortical gray matter Normalized volume and volume changes of cortex in frontal, temporal, parietal and occipital lobes Normalized volume and volume changes in the hippocampi Normalized volume and volume change of left and right hippocampus Asymmetry index of structures, including cortex in lobes and hippocampus Unnormalized volume and volume changes of FLAIR white matter hyperintensities FLAIR white matter hyperintensities in different regions (juxtacortical, periventricular, deep white matter and infratentorial). Unnormalized volume and volume changes of T1-w white matter hyperintensities and hypointensities CT: Normalized volumes of the whole brain Normalized volumes of the lateral ventricles Normalized volumes of basal cisterns | Similar.
iQ-solutions™ does not process NCCT images and therefore does not report findings for CT images.
For head MRI analysis, iQ-solutions™ is similar to icobrain 5.0, except that the subject device iQ-solutions™ does not report statistics related to regional (lobar) cortical volume change or T1-w image-derived hypo-intensities. iQ-
solutions™ additionally reports volume analysis of the thalamus.
These differences are not expected to affect the safety or effectiveness of the subject device or to be clinically significant in the proposed intended use. |
| Characteristic | Subject device:
iQ-solutions™ | Predicate Device:
icobrain 5.0 (K192130) | Summary |
| | | Measurement of midline shift
Volume of hyperintensities | |
| Removal of Cases
from the Worklist
Queue | No | No | Same |
| Performance
measurement
testing | Accuracy (for MRI only)
Brain segmentable structure
volumes/volume changes compared to FSL
generated and manually verified ground
truth. Reproducibility (for MRI only) Brain segmentable structure
volumes/volume changes compared to test-
retest images | Accuracy for MRI measurement Brain segmentable structure volumes/volume changes compared to simulated and/or manually labelled ground truth. Accuracy for CT measurement lesions and midline shift compared to manually labelled ground truth lateral ventricles and the whole brain: MRI images segmented by cleared icobrain 3.0 software taken as ground truth Reproducibility for MRI measurement Brain segmentable structure volumes/volume changes compared to test- retest images. Reproducibility for CT measurements Simulation study | Same for MRI measurements.
As iQ-solutions™ is not designed to analyze CT
images, the measurements for CT input are not
applicable in the subject device. |
| Environment of
Use | iQ-solutions™ is used by trained professionals
in hospitals, imaging centers or in image-
processing labs | icobrain 5.0 is used by trained professionals in
hospitals, imaging centers or image-processing
labs | Same |
| Tested Items | Product Risk Assessment
Software verification tests
Software validation tests | Product Risk Assessment
Software verification tests
Software validation tests | Same |
9
10
11
12
SNAC has determined that the predicate icobrain 5.0 (K192130) has similar technological and performance characteristics with respect to the segmentation and quantification of normalized volume and volume changes of different brain structures, including the whole brain, the gray matter, the cortical gray matter and the cortex in the brain lobes; the unnormalized volume and volume changes of FLAIR white matter hyperintensities; and the volume of contrast-enhancing T1 white matter hyperintensities. The differences between the subject device and the predicate device are not considered to be substantial or significant in the proposed intended use.
The conclusions from the verification and validation data support the assertion that the safety and effectiveness of the subject and predicate devices are equivalent; and that any differences in the technological characteristics and functionalities of the subject device do not alter/impact its intended use relative to the predicate product or have any additional implications with respect to safety or efficacy. We (SNAC) conclude that iQ-solutions™ is substantially equivalent to icobrain 5.0.
5.8 Performance Data
Each algorithm was tested using the test dataset, which was further verified by trained neuroimaging analysts and expert radiologists. All algorithms within iQ-solutions™ were verified to have generated results/metrics for compatible scans that met the pre-determined acceptance level(s).
5.9 Data Preparation and Demographics Info
The algorithms were developed based on SNAC in-house datasets collected from more than 1500 patients over a period of ten years, commencing July 2012, acquired from a wide range of clinical scanners, including 1.5T and 3T scanners from Siemens (1750 scans in total), GE (2214 scans in total) and Philips (1593 scans in total). Additionally, publicly available MRI datasets from healthy people were included for both training and testing, including 2629 scans obtained from 2159 subjects. The demographic information is included in Table 3.
| Datasets | Diagnosis
(subject/scans) | Number of
Subjects (F/M) | Age (mean ± std) | Number of Scans
(Vendor:1.5T/3T) |
|-----------------|--------------------------------------------|-----------------------------|-----------------------------|---------------------------------------------------------------------------------------------|
| Public
Data | Healthy (1823/2225),
Dementia (336/404) | 2159 (1146/1013) | 18y to 95y
(44.84±20.63) | 2629, including SIEMENS
(347/1449), Philips
(307/208), GE (79/221)
and others. |
| Private
Data | Multiple sclerosis
(1570/5664) | 1570 (1069/501) | 18y to 93y
(44.02±13.49) | 5664, including SIEMENS
(698/1052), Philips
(243/1350), GE
(1008/1206) and others. |
Table 3. Demographic Summary for Used Data
The in-house dataset contains patients diagnosed with multiple sclerosis, and the patients are also with brain atrophy and related cognitive dysfunctions. The scans of healthy people were also included in the training process to avoid bias. Table 4 summarizes the number of cases and involved subjects used in training and testing all the algorithms.
13
| Analysis Modules/Datasets | Number of Cases
(Train/Validation) | Number of
Cases (Test) | Number of
Subjects (All) |
|----------------------------------------------------------|---------------------------------------|---------------------------|-----------------------------|
| Sequence Classification | 4862 | 1207 | 2784 |
| Brain Extraction | 1843 | 458 | 1454 |
| Scaling Factor Estimation | 2986 | 527 | 3513 |
| White matter hyperintensity Segmentation | 1858 | 464 | 1026 |
| Contrast-Enhancing Lesion Segmentation | 79 | 737 | 310 |
| Lesion Inpainting | 449 | 80 | 529 |
| Brain Tissue Segmentation | 4179 | 500 | 3313 |
| Brain Volume Change Estimation | 1487 | 166 | 1648 |
| WMH Lesion Activity | 116 | 64 | 180 |
| Cortical Lobar and Subcortical Structure
Segmentation | Not applicable ** | 1504 | 1436 |
| Test-ReTest Dataset | Not applicable ** | 120 | 3 |
| Comprehensive Test Set | Not applicable ** | 81 | 81 |
Table 4. Summary of Number of Cases used in Developing/Evaluating Analysis Modules
5.10 Data Annotation and Acceptance Criteria
All data used for development were manually labelled to form the ground truth. The labelled items included sequence types (pre-contrast or post-contrast, T1-weighted or FLAR, brainbased or spine-based), skull and scaling factor based on skulls, skull stripped head, brain tissues (white matter, gray matter and CSF), white matter hyperintensities in patients, contrast-enhancing lesions, white matter hyperintensity lesion changes (new lesions or enlarged lesions), cortical gray matter lobes, thalamus, hippocampus, brain volume change. Specifically, the annotations for skullstripping, brain tissues, cortical gray matter lobes, thalamus, hippocampus, and brain volume changes were originally generated using widely used software items such as FSL and FreeSurfer, and the annotations were manually reviewed and accepted by trained neuroimaging analysts. The annotations for white matter hyperintensities (including pre-contrast ones, contrast-enhancing ones, and lesion changes) were manually annotated by trained neuroimaging analysts. All the annotations are further reviewed by senior neuroimaging analysts according to SNAC SOP.
The Pass/Fail criteria are based on both the metrics and acceptable rate of the manual review. For classification modules, accuracy is higher than 0.9 to be accepted; for segmentation modules, DICE score is higher than 0.6 (for brains the DICE should be higher than 0.9) or R² is higher than 0.7 to be accepted. For modules that are lack of ground truths (scaling factor estimation, lesion inpainting, and substructure volume change estimation), the metrics are better than previously accepted results to be accepted. For all models, the manual acceptable rate should be higher than 80% to call it acceptable.
5.11 Data Split
When used for development, the data were split into training, validation, and test splits. For the training/validation/test splits of each task, the same subject was ensured to be included in only one of the three splits, therefore it is ensured the same subject was not used for training and evaluation. The splits of each analysis component have been reviewed before the development process.
14
Besides the data used for development, 81 extra patients were reserved as the Comprehensive Test Set, which is used to evaluate the performance of all the analysis components.
5.12 Verification and Validation Result
The performance data demonstrates continued conformance with special controls for medical devices containing software. Non-clinical tests were conducted on the subject device during product development.
Software "bench" testing was performed to evaluate the performance and functionality of the technical characteristics of iQ-solutions™. All testable requirement Specifications and the Risk Analysis have been successfully verified and traced.
| Analysis Modules | Metric Result | Acceptable Rate | Pass/Fail |
|---|---|---|---|
| Sequence Classification | Accuracy = 100% | 100% | Pass |
| Brain Extraction | DICE = 0.982 | 99.3% | Pass |
| Scaling Factor Estimation | STD = 0.0096 | 100% | Pass |
| White matter hyperintensity Segmentation | DICE = 0.789 | 98.8% | Pass |
| Contrast-Enhancing Lesion Segmentation | DICE = 0.790 | 92.1% | Pass |
| Lesion Inpainting | PSNR = 30.79dB | 97.5% | Pass |
| Brain Tissue Segmentation | DICE = 0.972 | 100% | Pass |
| Brain Volume Change Estimation | $R^2 = 0.869$ | 96.3% | Pass |
| WMH Lesion Activity | $R^2 = 0.833$ | 97.5% | Pass |
| Cortical Lobar and Subcortical Structure | |||
| Segmentation | $R^2 = 0.833$ | 96.0% | Pass |
| Substructure Volume Change Estimation | STD = 0.0102 | 98.8% | Pass |
Table 5: Summary of Performances of all the Analysis Modules
The conclusions from the verification and validation data support the assertion that the safety and effectiveness of the subject and predicate devices are equivalent; and that any differences in the technological characteristics and functionalities of the subject device do not alter/impact its intended use relative to the predicate product or have any additional implications with respect to safety or efficacy. SNAC concludes that iQ-solutions™ is substantially equivalent to icobrain 5.0 (K192130).
5.13 Non-Clinical Tests
Non-clinical tests were conducted to test the functionality of iQ-solutions™. Software validation and bench testing have been conducted to assess the performance claims as well as the claim of substantial equivalence to the predicate device. A summary of tests is provided in Table 5 above.
iQ-solutions™ has been tested to meet the requirements of conformity to multiple industry standards. Non-clinical performance testing demonstrates that iQ-solutions™ complies with the FDA guidance document, "Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices" (May 11, 2005).
15
5.14 Clinical Tests
No clinical tests were conducted to test the performance and functionality of the subject device as verification and validation of the subject device have been performed through non-clinical bench testing. The data from non-clinical bench testing activities were used to support the subject device and the substantial equivalence argument. No animal testing has been performed on the subject device.
5.15 Conclusion
The performance testing presented above establishes that iQ-solutions™ is equivalent concerning safety and effectiveness of the predicate device. The differences in the iQ-solutions™ technological characteristics do not affect the intended use of the iQ-solutions™ device is substantially equivalent to the currently marketed device icobrain 5.0 (K192130).