Search Results

ECG-less Cardiac streamlines patient preparation by enabling an alternative acquisition of cardiac CT images for general cardiac assessment without the need of a patient-attached ECG monitor. ECG-less Cardiac is for adults only.

ECG-less Cardiac is a software device that is an additional, optional cardiac scan mode that can be used on the Revolution Apex Elite, Revolution Apex, and Revolution CT with Apex edition systems. There is no change to the predicate device hardware to support the subject device. Currently, the available cardiac scan modes on the Revolution CT Family are Cardiac Axial and Cardiac Helical, which makes use of an ECG signal to physiologically trigger the cardiac acquisitions and/or to retrospectively gate the reconstruction.

ECG-less Cardiac is a third cardiac scan mode that introduces the ability to acquire cardiac images without the need of a patient-attached ECG monitor. Hence, an ECG signal from the patient is not utilized for this scan mode. The ECG-less Cardiac workflow leverages the full-heart coverage capability of 160 mm configurations, fast gantry speeds (0.28 and 0.23 s/rot), and existing cardiac software options of SmartPhase and SnapShot Freeze 2 (K183161) to acquire images that are suitable for coronary and cardiac functional assessment.

The ECG-less cardiac feature allows the user to acquire a cardiac CT scan without the need to complete the steps associated with utilizing an ECG monitor, such as attaching ECG electrodes to the patient, checking electrode impedance, and confirming an ECG trace is displayed on the operator console, thus optimizing the workflow.

ECG-less Cardiac may be best utilized in examinations where excluding the ECG connection would streamline the patient examination, including and unloading of the patient. This may result in an improved workflow for certain clinical presentations. ECG-less Cardiac may also increase access to cardiac assessment for patients that are difficult to receive an ECG signal from. Circumstances where the subject device is expected to increase cardiac access includes scenarios where trauma patient has a diagnostic ECG attached and/or other instrumentation, such that there is added difficulty of attaching ECG leads for a gated scan, and situations where it is challenging to get an ECG signal from a patient such as a patient's t-wave triggering the scan or R-peak being difficult to detect.

Here's a summary of the acceptance criteria and the study proving the device meets them, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria are implicitly derived from the study's conclusions, focusing on diagnostic utility and image quality. No specific quantitative thresholds for acceptance are explicitly stated in the document beyond "interpretable without a significant motion artifact penalty" and "of diagnostic utility."

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

• Sample Size for Test Set: The document does not explicitly state the exact number of cases or images included in the reader study (test set). It refers to "a reader study of sample clinical data" and "prospectively collected clinical data."
• Data Provenance: The data was prospectively collected clinical data from patients undergoing a routine cardiac exam. The country of origin is not specified.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

• Number of Experts: Three experts were used.
• Qualifications of Experts: They were board-certified radiologists who specialize in cardiac imaging. The document does not specify their years of experience.

4. Adjudication Method for the Test Set

The adjudication method is not explicitly stated. The document mentions that each image was "read by three board certified radiologists who specialize in cardiac imaging who provided an assessment of image quality." This suggests independent readings, but it does not detail a consensus or adjudication process (e.g., 2+1, 3+1).

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and the Effect Size of How Much Human Readers Improve with AI vs. Without AI Assistance

• No, an MRMC comparative effectiveness study involving human readers improving with AI vs. without AI assistance was not conducted or reported.
• The study was a reader study where experts assessed images generated by the ECG-less Cardiac system. The primary goal was to validate the diagnostic utility and image quality of the ECG-less Cardiac acquisitions themselves, not to assess human reader performance with or without an AI assist feature.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

Yes, a form of standalone performance was assessed in the engineering bench testing. This testing "assessed how simulated ECG-less Cardiac scan conditions performed against an ECG-gated 'ground truth' nominal phase location." This component evaluated the algorithm's ability to generate images comparable to traditional ECG-gated acquisitions without human interpretation being the primary focus.

7. The Type of Ground Truth Used

• For the engineering bench testing, the ground truth was an ECG-gated "ground truth" nominal phase location. This implies a comparison to a known, established reference standard for cardiac imaging synchronization.
• For the clinical reader study, the ground truth was effectively the expert consensus/assessment of the three board-certified radiologists regarding the interpretability, motion artifact, and diagnostic utility of the ECG-less images. There is no mention of pathology or outcomes data being used as ground truth for this part of the study.

8. The Sample Size for the Training Set

The document does not provide any information regarding the sample size used for the training set of the ECG-less Cardiac software.

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

The document does not provide any information on how the ground truth for the training set was established. It only discusses the testing (validation) phase.

Ask a specific question about this device

The Deep Learning Image Reconstruction option is a deep learning based reconstruction method intended to produce cross-sectional images of the head and whole body by computer reconstruction of X-ray transmission data taken at different angles and planes, including Axial, Helical (Volumetric), and Cardiac acquisitions, for all ages.

Deep Learning Image Reconstruction can be used for head, whole body, cardiac, and vascular CT applications.

Deep Learning Image Reconstruction is an image reconstruction method that uses a dedicated Deep Neural Network (DNN) that has been designed and trained specifically to generate CT Images to give an image appearance, as shown on axial NPS plots, similar to traditional FBP images while maintaining the performance of ASiR-V in the following areas: image noise (pixel standard deviation), low contrast detectability, high-contrast spatial resolution, and streak artifact suppression.

The images produced are branded as "TrueFidelity™ CT Images". Reconstruction times with Deep Learning Image Reconstruction support a normal throughput for routine CT.

Deep Learning Image Reconstruction was trained specifically on the Revolution CT/Apex platform (K163213, K133705, K19177). The deep learning technology is integrated into the scanner's existing raw data-based image reconstruction chain to produce DICOM compatible "TrueFidelity™ CT Images".

The system allows user selection of three strengths of Deep Learning Image Recon: Low, Medium or High. The strength selection will vary with individual users' preferences and experience for the specific clinical need.

The DLR algorithm is now being modified on the Revolution family CT systems (K133705, K163213, K19177) for improved reconstruction speed and image quality, thus triggering this premarket notification.

The provided document describes the Deep Learning Image Reconstruction (DLIR) device, its acceptance criteria, and the study conducted to prove it meets these criteria.

Here's a breakdown of the requested information:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state numerical acceptance criteria with pass/fail thresholds for each metric. Instead, it focuses on demonstrating non-inferiority or improvement compared to a predicate device (ASiR-V) and ensuring diagnostic quality. The reported device performance is qualitative, indicating "significantly better subjective image quality" and "diagnostic quality images."

However, based on the non-clinical and clinical testing sections, we can infer the performance metrics evaluated.

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

• Sample Size: 40 retrospectively collected clinical cases.
• Data Provenance: Retrospectively collected clinical cases. The country of origin is not specified in the provided text.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts

• Number of Experts: 6 board-certified radiologists.
• Qualifications of Experts: Board-certified radiologists with "expecialty areas that align with the anatomical region of each case."

4. Adjudication Method for the Test Set

The document describes a reader study where each of the 40 cases (reconstructed with both ASiR-V and DLIR) was read by 3 different radiologists independently. They provided an assessment of image quality using a 5-point Likert scale. There's no explicit mention of an adjudication process (e.g., 2+1, 3+1) if there were disagreements among the three readers, as the focus seems to be on independent assessment and overall subjective preference comparison.

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

Yes, a multi-reader multi-case (MRMC) comparative effectiveness study was done. Human readers compared images reconstructed with DLIR (AI-assisted reconstruction) against images reconstructed with ASiR-V (without DLIR).

• Effect Size: The study confirmed that DLIR (the subject device) produced diagnostic quality images and "have significantly better subjective image quality" than the corresponding images generated with the ASiR-V reconstruction algorithm. The text doesn't provide a specific numerical effect size (e.g., a specific improvement percentage or statistical metric), but it qualitatively states a "significant" improvement based on reader preference for image noise texture, image sharpness, and image noise texture homogeneity.

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

Yes, extensive standalone (algorithm-only) performance testing was conducted. This is detailed in the "Additional Non-Clinical Testing" section, where DLIR and ASiR-V reconstructions of identical raw datasets were compared for various objective image quality metrics without human interpretation during these specific tests.

7. The Type of Ground Truth Used

The ground truth for the clinical reader study was established through expert consensus/assessment of image quality and preference by the participating radiologists. For the non-clinical bench testing, the ground truth was based on objective physical measurements and established phantom data with known properties.

8. The Sample Size for the Training Set

The document mentions that the Deep Neural Network (DNN) for DLIR was "trained specifically on the Revolution CT/Apex platform." However, it does not specify the sample size (number of images or cases) used for the training set.

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

The text states that the DNN was "trained specifically to generate CT Images to give an image appearance, as shown on axial NPS plots, similar to traditional FBP images while maintaining the performance of ASiR-V." It also notes that the DNN "models the scanned object using information obtained from extensive phantom and clinical data."

While the exact method for establishing ground truth for training isn't explicitly detailed, it implies a process where:

• Reference Images: Traditional FBP (Filtered Back Projection) and ASiR-V images likely served as reference or target outputs for the DNN, specifically regarding image appearance, noise characteristics, and spatial resolution.
• "Extensive phantom and clinical data": This data, likely corresponding to various anatomical regions, pathologies, and dose levels, was fed into the training process. The ground truth would involve teaching the network to reconstruct images that, when compared to conventionally reconstructed images (FBP/ASiR-V), exhibit desired image quality attributes (e.g., reduced noise while preserving detail).
• Noise Modeling: The training process characterized "the propagation of noise through the system" to identify and remove it, suggesting a ground truth related to accurate noise modeling and reduction.

Ask a specific question about this device

The GE Discovery MI Gen2 is a PET/CT system for producing attenuation corrected PET images. It is intended to be used by qualified health care professionals for imaging the distribution and localization of any positron-emitting radiopharmaceutical in a patient, for the assessment of metabolic (molecular) and physiologic function in patients, with a wide range of sizes and extent of disease, of all ages.

Discovery MI Gen2 is intended to image the whole body, head, heart, bone, the gastrointestinal and lymphatic systems, and other organs. The images produced by the system may be used by physicians to aid in radiotherapy treatment planning, therapy guidance and monitoring, and in interventional radiology procedures. The images may also be used for precise functional and anatomical mapping (localization, registration, and fusion).

When used with radiopharmaceuticals approved by the regulatory in the country of use, the raw and image data is an aid in; detection, localization, evaluation, diagnosis, staging, monitoring, and/or follow up, of abnormalities, lesions, tumors, inflammation, infection, organ function, disorders, and/or disease, such as, but not limited to, those in oncology, cardiology, and neurology. Examples of which are:

Cardiology:

• Cardiovascular disease
• Myocardial perfusion
• Myocardial viability
• Cardiac inflammation
• Coronary artery disease

Neurology:

• Epilepsy
• Dementia, such as Alzheimer's disease, Lewy body dementia, Parkinson's disease with dementia, and frontotemporal dementia
• Movement disorders, such as Parkinson's and Huntington's disease
• Tumors
• Inflammation
• Cerebrovascular disease such as acute stroke, chronic and acute ischemia
• Traumatic Brain Injury (TBI)

Oncology/Cancer:

• Non-Small Cell Lung Cancer
• Small Cell Lung Cancer
• Breast Cancer
• Prostate Cancer
• Hodgkin's disease
• Non-Hodgkin's lymphoma
• Colorectal Cancer
• Melanoma

Discovery MI Gen2 is also intended for stand-alone, diagnostic CT imaging in accordance with the stand-alone CT system's cleared indications for use.

GE's Discovery MI (DMI) Gen2, same as the unmodified predicate device, is a hybrid digital PET/CT diagnostic imaging system combining a GE Positron Emission Tomography (PET) System and a GE Computed Tomography (CT) System. The DMI Gen2 is intended for CT attenuation corrected, anatomically localized PET imaging of the distribution of positron-emitting radiopharmaceuticals. lt is intended to image the whole body, head, heart, brain, lung, breast, bone, the gastrointestinal and lymphatic systems, and other organs. The system is also intended for stand-alone, diagnostic CT imaging.

GE has modified the cleared Discovery MI (K161574) within our design controls to include a 6ring configuration that provides 30 cm Axial Field of View (AFOV) coverage. DMI Gen2 employs the same detector design architecture and manufacturing process as in the predicate to offer scalable ring configurations (3-ring, 4-ring, 5-ring and 6-ring) to have scalable AFOV coverage (15cm, 20cm, 25cm and 30cm) and corresponding imaging performances.

The provided text is a 510(k) Summary of Safety and Effectiveness for the GE Discovery MI Gen2 PET/CT system. It does not include a description of acceptance criteria or a detailed study proving the device meets specific performance metrics in a clinical setting. Instead, it states that clinical testing was not required due to the nature of the changes to the device and the use of established engineering and physics-based performance testing.

Therefore, many of the requested items related to clinical study design and ground truth are explicitly stated as not applicable or not performed in this submission.

Here's a breakdown based on the provided text, highlighting what is present and what is absent:

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

• Absent. The document does not provide specific acceptance criteria or reported performance metrics in a tabular format for clinical outcomes. It focuses on engineering and image performance evaluation testing, but no specific values or acceptance thresholds are given. The mention of "better detectability of small lesions" and "higher AFOV coverage... allows a patient to be scanned using fewer field of views" are general claims of improvement, not specific performance metrics against an acceptance criterion.

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

• Not applicable. This was a non-clinical study. The summary states: "Clinical Testing: Discovery MI Gen2 is designed and built entirely from existing and cleared systems, subsystems, components, and technologies of its Predicate Device (Discovery MI). This type of change in Discovery MI Gen 2 is supported using scientific, established/standardized, engineering/physics-based performance testing, without inclusion of clinical images, to demonstrate that the device is as safe and as effective as the predicate devices. Given the above information and the type and scope of the changes, particularly the addition of the 30 cm, 6-ring, AFOV configuration, clinical testing is not required to demonstrate that the Discovery MI Gen 2 is as safe and as effective as the legally marketed predicate device."
• "Image Performance evaluation testing used a variety of test methods and phantoms covering a broad base of relevant imaging performance and image quality test cases..." This indicates the test set consisted of phantoms, not patient data.

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)

• Not applicable. No human experts were used for ground truth because the testing was non-clinical, using phantoms, and relied on "mathematical and physics analysis" and "scientific methods that are standardized (e.g. NEMA, FDA Guidance), well established, and/or reviewed in previous GE's PETCT or Nuclear Medicine clearances."

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

• Not applicable. No human review or adjudication was performed as it was a non-clinical performance study.

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 study was not performed as clinical testing with human readers was not part of this 510(k) submission. No AI component is mentioned.

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

• Partially applicable, but not for a clinical algorithm. The testing was "algorithm only" in the sense that it assessed the device's performance using phantoms and engineering methods, independent of human interpretation in a clinical setting. However, it's a PET/CT system, not an AI algorithm in the context of diagnostic assistance. The document refers to "Deep Learning Image Reconstruction (DLIR) K193170" for the CT System component, suggesting an AI component is involved in image generation, but this submission focuses on the full PET/CT system and does not detail performance data specific to DLIR.

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

• Engineering/Physics-based Standards and Phantoms. The ground truth was established through "scientific methods that are standardized (e.g. NEMA, FDA Guidance)" and "phandoms" [sic] with known properties, along with "mathematical and physics analysis."

8. The sample size for the training set

• Not applicable / Not explicitly stated. This document describes the clearance of a medical imaging device (PET/CT system), not a new AI algorithm that requires a separate training set. While the CT component mentions "Deep Learning Image Reconstruction (DLIR)," the training data and methods for DLIR (K193170) are outside the scope of this specific 510(k) summary. For the overall PET/CT system, there isn't a "training set" in the sense of a machine learning model, but rather a design and development process based on existing technologies.

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

• Not applicable / Not explicitly stated. As above, no training set for a new AI algorithm specific to this 510(k) is discussed. For the DLIR component (from K193170, mentioned as part of the CT system), the ground truth for its training would have been established in its own separate clearance, likely through high-quality, low-noise CT scans.

Ask a specific question about this device

The deep learning based Max Field-of-View (MaxFOV 2) is a CT image reconstruction method intended to produce images of the head and whole-body using Axial, Helical, and Cine acquisitions.

MaxFOV 2 is designed to extend the nominal display field of view (DFoV) for cases where patient size and positioning requirements result in a portion of the patient's body to be outside of the nominal DFoV.

These extended FoV images are intended for use in radiation therapy planning and are clinically useful for the simulation and planning of radiation therapy for the treatment of cancer for patients. They can also be used for visualization of patient anatomy for cases not involving therapy planning. MaxFOV 2 is intended for patients of all ages, especially bariatric patients.

The MaxFOV 2 is an enhanced Extended Field of View (EFOV) reconstruction option for GE's CT scanners. The MaxFOV 2 utilizes a new deep learning algorithm to extend the display field of view (DFOV) beyond the CT system's scan field of View (SFOV) of 50cm to up to 80cm depending on the bore size of the CT system. CT scanners use the EFOV reconstruction algorithms to visualize tissue truncated due to large patient habitus and/or off-center patient positioning. Same as the Wide View option on the predicate, the MaxFOV 2 is designed to enable a clinically useful visualization of the skin line and CT Number of human body parts located outside of the SFOV. EFOV images are especially useful for radiation therapy planning and they can also be used for visualization of patient anatomy outside of the SFOV for routine CT imaging. This DL enabled new MaxFOV2 EFOV reconstruction process offers improved performance over the existing WideView option on the predicate device.

The DL MaxFOV2 algorithm was designed and tested for GE's multiple CT scanner platforms of various bore sizes from 70cm to 80cm. These CT systems with the integrated MaxFOV 2 option remain compliant with the same standards as base CT systems.

This option is commercially marketed as MaxFOV2.

The provided text describes the MaxFOV 2, a CT image reconstruction method that utilizes deep learning to extend the display field of view. The document outlines the device's indications for use, technological characteristics, and the summary of non-clinical and clinical testing performed to support its substantial equivalence to a predicate device.

Here's an analysis of the acceptance criteria and the study that proves the device meets them, based on the provided text:

Acceptance Criteria and Reported Device Performance

The document doesn't explicitly state quantitative acceptance criteria in a table format with pass/fail thresholds. Instead, it describes various tests and a clinical reader study designed to demonstrate performance that is "consistent and acceptable," does not "raise new or different questions of safety and effectiveness," and supports "substantial equivalence and performance claims." The performance is evaluated relative to the predicate device, "Wide View (K023332)".

Assuming the core acceptance criteria revolve around image quality, accuracy of patient contour (skin line), and CT number accuracy in the extended FOV, the reported performance is qualitative but positive:

Study Details

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

• Sample Size for Test Set: 49 CT exams were used in the clinical reader study.
• Data Provenance: The exams were "acquired from different GE CT system platforms." The text does not specify the country of origin or whether the data was retrospective or prospective. It states the exams "represent typical and challenging RTP-relative scenarios where the MaxFOV2 will likely be used."

3. Number of Experts and Qualifications for Ground Truth

• Number of Experts: 5 external, clinical readers.
• Qualifications of Experts: The specific qualifications (e.g., years of experience, subspecialty) are not explicitly stated, beyond them being "external, clinical readers."

4. Adjudication Method for the Test Set

The text indicates that readers scored images using a 5-point Likert scale. It does not mention an explicit adjudication method (e.g., 2+1, 3+1 consensus) for establishing a single "ground truth" score per image. It seems the readers' individual scores were aggregated and analyzed to support the claims.

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

• Was it done?: A clinical reader study was performed, which is a type of MRMC study, though the primary goal seems to be demonstrating substantial equivalence rather than a direct comparison of human readers with and without AI assistance to quantify an "improvement effect size." The study evaluates the output of the AI model (the reconstructed images) through human reader assessment.
• Effect Size of Human Improvement with AI vs. without AI assistance: Not explicitly quantified in terms of human reader improvement. The study assesses the quality of images produced by MaxFOV2, which uses a deep learning component. The improvement is implied for the images produced by MaxFOV2 over the predicate device. "This DL enabled new MaxFOV2 EFOV reconstruction process offers improved performance over the existing WideView option on the predicate device." However, this statement refers to the device's performance, not specifically how human readers improve their diagnostic accuracy or efficiency when using the AI-assisted images compared to reading images generated without the AI.

6. Standalone Performance (Algorithm Only)

The text describes "engineering bench testing using phantoms." These tests (MaxFOV 2 Patient contour (Skin line) accuracy and CT Number accuracy, MaxFOV 2 IQ Performance Evaluation using a very large phantom, MaxFOV 2 Performance Evaluation Using an anthropomorphic phantom) appear to be standalone performance assessments of the algorithm and its output, independent of a human reader's interpretation in a diagnostic context. These tests would evaluate the algorithm's output directly against known phantom characteristics.

7. Type of Ground Truth Used for Test Set

• For Bench Testing: The ground truth for bench testing (e.g., image quality, skin line accuracy, CT number accuracy) would be established by the known physical properties and measurements of the phantoms used.
• For Clinical Reader Study: The "ground truth" for clinical acceptability in the reader study is based on the expert consensus/opinion of the 5 external clinical readers using a Likert scale. This is a subjective assessment of image quality and clinical utility rather than an objective "pathology" or "outcomes" ground truth. The study evaluates how well the generated images aid human interpretation.

8. Sample Size for the Training Set

The document does not explicitly state the sample size used for the training set of the deep learning algorithm ("CNN"). It only mentions that the MaxFOV2 "uses a CNN which is trained on multiple CT scanners."

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

The document does not provide details on how the ground truth for the training set was established. It simply states that the CNN was "trained on multiple CT scanners." For deep learning image reconstruction, the training might involve paired data (e.g., truncated vs. full FOV images, or images generated with existing algorithms as a reference), or simulated data, but the specific method of ground truth establishment is not described.

Ask a specific question about this device

Critical Care Suite is a computer aided triage and notification device that analyzes frontal chest x-ray images for the presence of prespecified critical findings (pneumothorax). Critical Care Suite identifies images with critical findings to enable case prioritization or triage in the PACS/workstation.

Critical Care Suite is intended for notification only and does not provide diagnostic information beyond the notification. Critical Care Suite should not be used in-lieu of full patient evaluation or solely relied upon to make or confirm a diagnosis. It is not intended to replace the review of the x-ray image by a qualified physician.

Critical Care Suite is indicated for adult-size patients.

Critical Care Suite is a software module that employs Al-based image analysis algorithms to identify pre-specified critical findings (pneumothorax) in frontal chest X-ray images and flag the images in the PACS/workstation to enable prioritized review by the radiologist.

Critical Care Suite employs a sequence of vendor and system agnostic AI algorithms to ensure that the input images are suitable for the pneumothorax detection algorithm and to detect the presence of pneumothorax in frontal chest X-rays:

• The Quality Care Suite algorithms conduct an automated check to confirm that the input image is compatible with the pneumothorax detection algorithm and that the lung field coverage is adequate;

• the PTX Classifier determines whether a pneumothorax is present in the image.

If a pneumothorax is detected, Critical Care Suite enables case prioritization or triage through direct communication of the Critical Care Suite notification during image transfer to the PACS. It can also produce a Secondary Capture DICOM Image that presents the Al results to the radiologist.

When deployed on a Digital Projection Radiographic Systems such as Optima XR240amx, Critical Care Suite is automatically run after image acquisition. Quality Care Suite algorithms produce an on-device notification if the lung field has atypical positioning to give the technologist the opportunity to make correction before sending the image to the PACS. The Critical Care Suite output is then sent directly to PACS upon exam closure where images with a suspicious finding are flagged for prioritized review by the Radiologist.

In parallel, an on-device, technologist notification is generated 15 minutes after exam closure, indicating which cases were prioritized by Critical Care Suite in PACS. The technologist notification is contextual and does not provide any diagnostic information. The on-device, technologist notification is not intended to inform any clinical decision, prioritization, or action.

The Digital Projection Radiographic System intended use remains unchanged in that the system is used for general purpose diagnostic radiographic imaging.

Here's a breakdown of the acceptance criteria and the study that proves the device meets them, based on the provided text:

Acceptance Criteria and Device Performance

1. Table of Acceptance Criteria and Reported Device Performance:

2. Sample size and Data Provenance for the Test Set:

• Sample Size: 804 frontal chest X-ray images (N=376 for pneumothorax present; N=428 for pneumothorax absent).
• Data Provenance: Collected in North America, representative of the intended population. The text does not explicitly state if it was retrospective or prospective, but the nature of a "collected dataset" for evaluation typically implies retrospective analysis of existing images.

3. Number of Experts and Qualifications for Ground Truth of the Test Set:

• Number of Experts: 3 independent US-board certified radiologists.
• Qualifications: "US-board certified radiologists." No specific years of experience or subspecialty are mentioned beyond board certification.

4. Adjudication Method for the Test Set:

• The text states the ground truth was "established by 3 independent US-board certified radiologists." It does not explicitly detail a specific adjudication method like 2+1 or 3+1. This implies a consensus-based approach where the radiologists independently reviewed images to establish the ground truth, likely resolving discrepancies through discussion to reach a final determination.

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

• No, a multi-reader multi-case (MRMC) comparative effectiveness study directly comparing human readers with AI assistance vs. without AI assistance was not reported in this summary. The study focused on the standalone diagnostic performance of the AI algorithm.

6. Standalone (Algorithm Only) Performance Study:

• Yes, a standalone performance study of the algorithm without human-in-the-loop was done. The reported metrics (ROC AUC, Sensitivity, Specificity) are direct measurements of the algorithm's performance against the established ground truth.

7. Type of Ground Truth Used:

• Expert Consensus: The ground truth was established by "3 independent US-board certified radiologists." This indicates an expert consensus approach.

8. Sample Size for the Training Set:

• The document does not explicitly state the sample size for the training set. It mentions the algorithm was "trained on annotated medical images" but provides no further details on the quantity of images used for training.

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

• The document states the device utilizes a "deep learning algorithm trained on annotated medical images." While it doesn't explicitly describe the method for establishing ground truth for the training set, it is implied that these "annotated medical images" had pre-existing labels or were labeled by experts for the purpose of training the AI.

Ask a specific question about this device

The Deep Learning Image Reconstruction option is a deep learning based reconstruction method intended to produce cross-sectional images of the head and whole body by computer reconstruction of X-ray transmission data taken at different angles and planes, including Axial, Helical (Volumetric), and Cardiac acquisitions, for all ages.

Deep Learning Image Reconstruction can be used for head, whole body, cardiac, and vascular CT applications.

Deep Learning Image Reconstruction is an image reconstruction method that uses a dedicated Deep Neural Network (DNN) that has been designed and trained specifically to generate CT lmages to give an image appearance, as shown on axial NPS plots, similar to traditional FBP images while maintaining the performance of ASiR-V in the following areas: image noise (pixel standard deviation), low contrast detectability, high-contrast spatial resolution, and streak artifact suppression.

The images produced are branded as "TrueFidelity™ CT Images". Reconstruction times with Deep Learning Image Reconstruction support a normal throughput for routine CT.

Deep Learning Image Reconstruction was trained specifically on the Revolution CT family of systems (K163213, K133705). The deep learning technology is integrated into the scanner's existing raw data-based image reconstruction chain to produce DICOM compatible "TrueFidelity™ CT Images".

The system allows user selection of three strengths of Deep Learning Image Recon: Low, Medium or High. The strength selection will vary with individual users' preferences and experience for the specific clinical need.

Acceptance Criteria and Device Performance for Deep Learning Image Reconstruction (K183202)

The Deep Learning Image Reconstruction (DLIR) device, developed by GE Medical Systems, LLC, was evaluated for substantial equivalence to its predicate device, ASiR-V, as part of its 510(k) submission (K183202). The acceptance criteria for the DLIR are implicitly defined by its performance being equivalent to or better than the predicate device across various image quality metrics relevant to CT imaging.

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria for DLIR are based on maintaining the performance of ASiR-V in key imaging characteristics while achieving an image appearance similar to traditional FBP images. The "reported device performance" refers to the demonstrated performance of DLIR relative to ASiR-V during non-clinical and clinical testing, confirming it met the unspoken expectation of non-inferiority or improvement.

2. Sample Size for the Test Set and Data Provenance

The clinical reader study used 60 retrospectively collected clinical cases. The raw data from these cases were reconstructed with both ASiR-V and Deep Learning Image Reconstruction. The data provenance is not explicitly stated in terms of country of origin but is implied to be from standard clinical practice given the retrospective collection of cases.

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

Nine board-certified radiologists were used for the clinical reader study (test set). Their qualifications included:

• Expertise in specialty areas aligning with the anatomical region of each case.
  • Three radiologists specialized in body and extremity anatomy.
  • Three radiologists specialized in head/neck anatomy.
  • Three radiologists specialized in cardiac/vascular anatomy.

A single board-certified radiologist performed a final evaluation of low contrast and small lesions in the abdominal and pelvis region.

4. Adjudication Method (Test Set)

Each image in the clinical reader study was read by 3 different radiologists independently. These radiologists provided an assessment of image quality related to diagnostic use according to a a 5-point Likert Scale. There is no explicit mention of an adjudication process (e.g., 2+1, 3+1), but for the direct comparison part, readers were asked to compare ASiR-V and DLIR images directly.

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

Yes, a multi-reader multi-case (MRMC) study was performed. The study involved 9 radiologists reading 60 cases reconstructed with both ASiR-V and DLIR.
The exact effect size of how much human readers improve with AI (DLIR) vs. without AI (ASiR-V, as it's also an advanced reconstruction) assistance is not explicitly quantified in terms of specific metrics like diagnostic accuracy improvement or reading time reduction. However, the study's results are stated to "support substantial equivalence and performance claims." Readers were also asked to directly compare ASIR-V and Deep Learning Image Reconstruction images according to three key metrics of image quality preference – image noise texture, image sharpness, and image noise texture homogeneity, implying a preference or at least equivalence for DLIR.

6. Standalone (Algorithm Only) Performance

Yes, standalone (algorithm only) performance was done as part of the engineering bench testing. This included objective measurements of various image quality metrics using identical raw datasets on a GE Revolution CT, then applying DLIR or ASiR-V reconstruction. The results from this testing demonstrated the algorithm's performance in:

• Low Contrast Detectability (LCD)
• Image Noise (pixel standard deviation)
• High-Contrast Spatial Resolution (MTF)
• Streak Artifact Suppression
• Spatial Resolution, longitudinal (FWHM)
• Low Contrast Detectability/resolution (statistical)
• Noise Power Spectrum (NPS) and Standard Deviation of noise
• CT Number Uniformity
• CT Number Accuracy
• Contrast to Noise (CNR) ratio
• Artifact analysis

7. Type of Ground Truth Used

• Non-Clinical Testing: Ground truth for non-clinical testing was established via physical phantoms (e.g., MITA/FDA low contrast phantoms, uniform phantoms, quality assurance phantoms with tungsten wire, oval uniform polyethylene phantoms with embedded objects) and model observers for objective measurements.
• Clinical Testing: Ground truth for the clinical reader study was based on expert consensus/opinion from board-certified radiologists using a 5-point Likert scale for image quality assessment and for direct comparison of image quality preference attributes. A final evaluation by one board-certified radiologist confirmed diagnostic quality against established clinical standards.

8. Sample Size for the Training Set

The document states that Deep Learning Image Reconstruction was trained specifically on the Revolution CT family of systems (K163213, K133705). It also mentions that the Deep Neural Network (DNN) "models the scanned object using information obtained from extensive phantom and clinical data." However, an exact sample size (number of images or cases) for the training set is not provided in the provided text.

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

The ground truth for the training set, which involved "extensive phantom and clinical data," was established through the inherent characteristics of CT imaging data. For example, for phantom data, the known physical properties and structures within the phantoms serve as ground truth. For clinical data, the "ground truth" for the training process would implicitly be derived from high-quality, typically higher-dose or reference-standard reconstructions (e.g., traditional FBP or fully iterative reconstructions) that the DNN aims to emulate or improve upon, often by learning to remove noise while preserving diagnostic information. The DNN was designed to generate CT images to give an image appearance, as shown on axial NPS plots, similar to traditional FBP images while maintaining the performance of ASiR-V.

Ask a specific question about this device

SnapShot Freeze 2 is designed for use with gated cardiac acquisitions to reduce cardiac induced motion artifacts.

SnapShot Freeze 2 (SSF 2) is a post processing software, which can be delivered on general purpose computing platforms. SnapShot Freeze 2 is an automated motion correction algorithm designed for use with gated cardiac acquisitions from GE CT scanners to reduce cardiac induced motion artifacts in the whole heart. It is based on the same fundamental algorithm as the predicate product commercially marketed under the name CardIQ Xpress 2.0 with SnapShot Freeze Option (K120910, AKA SSF 1). Same as its predicate device the SSF 2 algorithm works on multi-phase, gated cardiac CT DICOM compatible image data and produces a new image series in which motion artifact is reduced.

The provided text does not contain a detailed study proving the device meets specific acceptance criteria with reported device performance metrics in a table. Instead, it describes internal validation and testing to support the product's substantial equivalence to a predicate device.

However, based on the information provided, here's an attempt to extract relevant details and structure them according to your request, with disclaimers that much of the quantitative information you're asking for (like specific acceptance criteria values and reported performance against them) is not present in the given document.

Device Name: SnapShot Freeze 2
Intended Use: Motion correction of gated cardiac image series.
Indications for Use: Designed for use with gated cardiac acquisitions to reduce cardiac induced motion artifacts.

Overview of Device Performance and Acceptance (as inferred from the document):

The document describes engineering bench testing and a clinical assessment to demonstrate that SnapShot Freeze 2 is "as safe and effective, and performs in a substantially equivalent manner to the predicate device CardIQ Xpress 2.0 with SnapShot Freeze Option." The main improvement of SnapShot Freeze 2 is its ability to extend motion correction to the "whole heart" beyond just coronary vessels.

1. Table of Acceptance Criteria and Reported Device Performance:

The document does not provide a clear table of predefined acceptance criteria with corresponding numerical performance results. However, it mentions qualitative performance improvements and claims of "effective temporal resolution" based on phantom testing.

2. Sample Size and Data Provenance for Test Set:

• Sample Size for Clinical Assessment: A "representative clinical sample image set of 60 CT cardiac cases."
• Data Provenance: The document states this data is "representative of routine clinical imaging from a cardiac acquisition perspective," but "intentionally includes data from subjects with elevated heart rates or those with heart rate variability which represent more challenging cases." It does not specify the country of origin, nor explicitly state if it was retrospective or prospective, but the phrasing "representative clinical sample image set" and "routine clinical imaging" suggests it was likely a retrospective collection of existing patient data.

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

• Number of Experts: Three board certified radiologists.
• Qualifications: "Board certified radiologists." No further details on years of experience are provided.

4. Adjudication Method for the Test Set:

• The document states, "A representative clinical sample image set of 60 CT cardiac cases... were assessed by three board certified radiologists using 5-point Likert scale."
• It does not specify an adjudication method (e.g., 2+1, 3+1 consensus, or independent reading). It only states they "assessed" the cases.

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

• No MRMC comparative effectiveness study demonstrating how human readers improve with AI vs. without AI assistance is explicitly described. The clinical assessment was to demonstrate the "diagnostic capability of the motion corrected images by SSF 2," not a direct comparison of human reader performance with and without the tool.

6. Standalone (Algorithm Only) Performance:

• The document implies that the effective temporal resolution metrics (29 ms, 24 ms) derived from "mechanical and mathematical cardiac phantom testing" represent the standalone performance of the algorithm in terms of motion correction capability, independent of human interpretation.
• The clinical assessment of "diagnostic capability" also evaluates the output of the algorithm, suggesting an evaluation of its quality for diagnostic purposes.

7. Type of Ground Truth Used for the Test Set:

• The "ground truth" for the clinical assessment appears to be the expert consensus/assessment of the three board-certified radiologists using a 5-point Likert scale to determine the "diagnostic capability" of the motion-corrected images. No mention of pathology or outcomes data as direct ground truth for image quality/diagnostic utility is made for the 60 clinical cases.

8. Sample Size for the Training Set:

• The document does not specify a sample size for the training set. It describes the algorithm's fundamental similarity to its predicate (SSF1) and states that SSF2 "extends the motion correction capability to the whole heart." This suggests the core algorithm was already established, and the "enhancement" for whole-heart motion correction might have involved further development or refinement without explicitly detailing a new, distinct "training set" in this submission summary.

9. How Ground Truth for Training Set was Established:

• The document does not describe how the ground truth for any potential training set was established. Given that the algorithm is based on and an extension of a predicate device, it's plausible that the underlying algorithm was developed and validated earlier, and the current submission focuses on the safety and effectiveness of the extended capability.

Ask a specific question about this device

GenIQ is an automated post-processing software option that is indicated for use on dynamic magnetic resonance imaging data sets to generate parametric image intensity variations over time. This dynamic change in signal intensity is used to calculate functional parameters related to tissue flow and leakage of the contrast agent from the intravascular to the extracellular space.

GenIQ provides information that when interpreted by a trained physician, can be useful for assessing tissue vascular properties.

GenIQ is a software application used for the pharmacokinetic analysis of Dynamic Contrast Enhanced (DCE) MRI data sets. The application is used to perform a General Kinetic Model (GKM)-based pharmacokinetic modeling of DCE-MRI data. The goal of GenIQ is to extract functional parameters describing tissue vascular properties such as forward and backward transfer constants, plasma volume, and volume of extra-cellular space.

Acceptance Criteria and Study for GenIQ

The provided document describes the GenIQ, an automated post-processing software option for dynamic Magnetic Resonance Imaging (MRI) data sets. It calculates functional parameters related to tissue flow and contrast agent leakage.

1. Acceptance Criteria and Reported Device Performance

The document does not explicitly present a table of acceptance criteria with corresponding performance metrics like sensitivity, specificity, accuracy, or AUC. Instead, it states that "Simulated use testing was performed on digital phantom data referenced by Quantitative Imaging Biomarkers Alliance (QIBA). This validation demonstrated good implementation of the General Kinetic Model."

This implies the acceptance criterion for the phantom study was the accurate implementation of the General Kinetic Model (GKM) as validated against QIBA reference data. The reported performance is that the "validation demonstrated good implementation" of this model.

For clinical data, the document states, "anonymized MR contrast-enhanced images were used as clinical datasets to validate the GenIQ application." Again, specific performance metrics against an acceptance criterion are not detailed, but the overall conclusion is that "GE Healthcare considers the GenIQ application to be as safe, as effective, and performance is substantially equivalent to the predicate device." This suggests the clinical validation aimed to demonstrate performance comparable to its predicate (Philips MR Permeability Software - K130278), which would inherently imply meeting certain effectiveness and safety standards.

Table of Acceptance Criteria and Reported Device Performance (Inferred):

2. Sample Size and Data Provenance for the Test Set

• Sample Size for Test Set:
  • Digital Phantom Data: Not specified.
  • Clinical Datasets: Not specified, only described as "anonymized MR contrast-enhanced images."
• Data Provenance: The document does not explicitly state the country of origin.
  • Digital Phantom Data: Referenced by the Quantitative Imaging Biomarkers Alliance (QIBA).
  • Clinical Data: Described as "anonymized MR contrast-enhanced images." The study is likely retrospective as it used existing "anonymized" images.

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

This information is not provided in the document. For the digital phantom data, the ground truth is inherently defined by the QIBA reference data and the mathematical model itself. For the clinical datasets, the method of establishing ground truth (e.g., expert consensus, pathology) and the number/qualifications of experts are not described.

4. Adjudication Method (Test Set)

This information is not provided in the document.

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

An MRMC comparative effectiveness study is not explicitly mentioned. The validation focused on the software's ability to implement the GKM and its equivalence to a predicate device. The document states, "GenIQ provides information that when interpreted by a trained physician," implying human-in-the-loop, but there's no study comparing human readers with and without AI assistance to quantify an effect size.

6. Standalone Performance

The evaluation primarily describes the "GenIQ application" as an automated post-processing software option, suggesting a focus on its standalone (algorithm-only) performance in calculating pharmacokinetic parameters from DCE-MRI data. While the output is "interpreted by a trained physician," the validation described (GKM implementation on phantom data, and validation on clinical datasets) focuses on the algorithm's accuracy in producing these parameters.

7. Type of Ground Truth Used

• Digital Phantom Data: The ground truth was based on the Quantitative Imaging Biomarkers Alliance (QIBA) reference data and the theoretical correctness of the General Kinetic Model (GKM). These are essentially simulated/known data sets designed to test the mathematical implementation.
• Clinical Datasets: The document does not specify the type of ground truth used for the clinical validation. It only states that images were used "to validate the GenIQ application." This could imply comparison to a reference standard established by expert consensus, other imaging modalities, or clinical outcomes, but it is not detailed.

8. Sample Size for the Training Set

This information is not provided in the document. The document describes testing and validation, but not the development or training of the software.

9. How Ground Truth for the Training Set was Established

This information is not provided in the document as the training set details are absent.

Ask a specific question about this device