Search Results

The Deep Learning Image Reconstruction software 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, (Volumetric), and Cardiac acquisitions, for all ages.

Deep Learning Image Reconstruction software 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.

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' preference for the specific clinical need.

The DLR algorithm was modified on the Revolution CT/Apex platform for improved reconstruction speed and image quality and cleared in February 2022 with K213999. The same modified DLIR is now being ported to Revolution EVO (K131576) /Revolution Maxima (K192686), Revolution Ascend (K203169, K213938) and Discovery CT750 HD family CT systems including Discovery CT750 HD, Revolution Frontier and Revolution Discovery CT (K120833).

The provided text describes that the Deep Learning Image Reconstruction software was tested for substantial equivalence to a predicate device (K213999). The study performed was largely an engineering bench testing, comparing various image quality metrics between images reconstructed with Deep Learning Image Reconstruction (DLIR) and ASiR-V from the same raw datasets.

Here's a breakdown of the requested information based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The text indicates that the device aims to maintain the performance of ASiR-V in specific areas while offering an image appearance similar to traditional FBP images. The "acceptance criteria" can be inferred from the list of image quality metrics evaluated, with the performance goal being comparable or improved relative to ASiR-V.

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

• Test Set Sample Size: The text states "the identical raw datasets obtained on GEHC's Revolution Ascend, Revolution Frontier and Discovery CT750 HD systems". However, the number of cases or specific sample size for these datasets is not explicitly stated.
• Data Provenance: The raw datasets were "obtained on GEHC's Revolution Ascend, Revolution Frontier and Discovery CT750 HD systems". The country of origin is not specified, and it is stated that the study used retrospective raw datasets (i.e., existing data, not newly acquired data for the study).

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

The provided text focuses on engineering bench testing and image quality metrics. It does not mention the use of experts to establish ground truth for the test set or their qualifications. The evaluation primarily relies on quantitative image quality metrics.

4. Adjudication Method for the Test Set

Since experts were not explicitly used to establish ground truth, there is no mention of an adjudication method for the test set in the provided text.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done, and its effect size

An MRMC comparative effectiveness study was not performed according to the provided text. The study focused on technical image quality comparisons at the algorithm level, not human reader performance with or without AI assistance.

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

Yes, a standalone performance evaluation was done. The study described is primarily a standalone evaluation of the algorithm's image quality output (e.g., noise, resolution, artifacts, detectability) when compared to images reconstructed with ASiR-V from the same raw data.

7. The Type of Ground Truth Used

The "ground truth" for the test set was essentially:

• Quantitative Image Quality Metrics: Performance relative to ASiR-V for metrics like image noise, LCD, spatial resolution, streak artifact suppression, CT uniformity, CT number accuracy, CNR, spatial resolution (longitudinal), NPS, and artifact analysis.
• Reference Image Appearance: The stated goal was an image appearance similar to traditional FBP images shown on axial NPS plots.

There is no mention of pathology, expert consensus on clinical findings, or outcomes data being used as ground truth for this particular substantial equivalence study.

8. The Sample Size for the Training Set

The text states that the Deep Neural Network (DNN) is "trained on the CT scanner" and models the scanned object using "information obtained from extensive phantom and clinical data." However, the specific sample size for the training set is not provided.

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

The ground truth for the training set is implicitly established through the "extensive phantom and clinical data" mentioned as being used to train the DNN. The text indicates the DNN is trained to model noise propagation and identify noise characteristics to remove it, and to generate images with an appearance similar to traditional FBP while maintaining ASiR-V performance. This suggests the training involves learning from "ground truth" as defined by:

• Reference Image Quality: Likely images reconstructed with proven methods (e.g., FBP, ASiR-V) or images from phantoms with known properties.
• Noise Characteristics: The DNN is trained to understand and model noise.

However, the specific methodology for establishing this ground truth for the training data (e.g., expert annotation, simulated data, pathology confirmed disease) is not detailed in the provided text.

Ask a specific question about this device

The Deep Learning Image Reconstruction software 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 software 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: dose, 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 software support a normal throughput for routine CT.

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' preference for the specific clinical need.

The provided text is a 510(k) summary for the GE Healthcare Japan Corporation's "Deep Learning Image Reconstruction" device. It outlines the device's technical characteristics, intended use, and comparison to predicate devices for substantial equivalence determination. However, it does not include detailed information regarding specific acceptance criteria, a comprehensive study proving the device meets these criteria, or specific performance metrics in a tabular format. The document focuses on establishing substantial equivalence based on the fundamental technology being unchanged from the predicate and successful completion of design control testing and quality assurance measures.

Therefore, I cannot extract all the requested information. Here's what can be inferred and what is missing:

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

This information is not provided in the document. The document states: "Design verification and validation, including IQ bench testing, demonstrate that the Deep Learning Image Reconstruction (DLIR) software met all of its design requirement and performance criteria." However, it does not specify what those "design requirement and performance criteria" are or the reported performance data against them.

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

This information is not provided in the document. The document mentions "IQ bench testing" and "System Testing" including "Image Performance Testing (Verification)" and "Simulating Use Testing (Validation)," but does not detail the sample sizes or data provenance used for these tests.

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)

This information is not provided in the document.

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

This information is not provided in the document.

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

This information is not provided in the document. The document describes the device as a "deep learning based reconstruction method" that produces images with "similar to traditional FBP images while maintaining the performance of ASiR-V in the following areas: dose, image noise (pixel standard deviation), low contrast detectability, high-contrast spatial resolution, and streak artifact suppression." This implies a comparison to other reconstruction methods, but not a MRMC study involving human readers with and without AI assistance.

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

Yes, based on the description, the primary testing described is "standalone" algorithm performance. The device is a "deep learning based reconstruction method" and the testing described, such as "IQ bench testing" and "Image Performance Testing," focuses on the intrinsic image quality outputs of the algorithm. There is no mention of human-in-the-loop performance in the context of effectiveness studies.

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

This information is not explicitly stated in the document. Given the context of "IQ bench testing" and performance metrics like "image noise," "low contrast detectability," and "spatial resolution," it's highly likely that objective phantom studies and potentially established image quality metrics (which could be considered a form of "ground truth" for image quality, validated against known physical properties) were used. However, expert consensus on clinical diagnostic accuracy or pathology is not mentioned as a ground truth.

8. The sample size for the training set

This information is not provided in the document. It mentions that the device "uses a dedicated Deep Neural Network (DNN) that has been designed and trained specifically to generate CT Images," but the details of the training set are not disclosed.

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

This information is not provided in the document. While it states 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," the method for establishing the ground truth for this training is not detailed.

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 Deep Learning Image Reconstruction software 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 software 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: dose, 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 software support a normal throughput for routine CT.

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.

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

The core of the acceptance criteria revolves around demonstrating that the Deep Learning Image Reconstruction (DLIR) on the Revolution Ascend system is substantially equivalent to its predicate device (DLIR on Revolution EVO) and performs at least as well as, or better than, ASiR-V reconstruction in key image quality metrics.

Study Details

1. Sample size used for the test set and the data provenance:

   • Sample Size: A total of 60 retrospectively collected clinical cases were used.
   • Data Provenance: The data was retrospectively collected. The country of origin is not explicitly stated in the provided text.

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

   • Number of Experts: 9 board-certified radiologists.
   • Qualifications: These radiologists had "expertise in the specialty areas that align with the anatomical region of each case."

3. Adjudication method for the test set:

   • Each image was read by 3 different radiologists.
   • The readers completed their evaluations independently and were blinded to the results of the other readers' assessments.
   • The text doesn't explicitly state an adjudication method like 2+1 or 3+1 for discrepancies. It implies a consensus or agreement was sought, or that individual assessments contributed to the overall conclusion of substantial equivalence. Given they provided an assessment on a Likert scale and then compared images, it seems individual reader assessments were aggregated, rather than a discrepancy resolution process.

4. 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, an MRMC study was implicitly done, as 9 radiologists evaluated 60 cases, with each case being read by 3 different radiologists. The study involved a comparison between ASiR-V reconstructions and Deep Learning Image Reconstruction (DLIR) images.
   • Effect Size: The document does not provide a specific effect size (e.g., percentage improvement in accuracy or AUC) of how much human readers improved with DLIR assistance compared to ASiR-V. It states that the study results "support substantial equivalence and performance claims" and that readers assessed image quality and compared noise texture and sharpness, implying favorable or equivalent performance.

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

   • Yes, extensive standalone (algorithm only) non-clinical engineering bench testing was performed. This included evaluations of:
     • Low contrast detectability (LCD)
     • Image Noise (pixel standard deviation)
     • High contrast spatial resolution (MTF)
     • Streak Artifact Suppression
     • Spatial Resolution
     • Noise Power Spectrum (NPS) and Standard Deviation of noise
     • CT Number Accuracy and Uniformity
     • Contrast to Noise (CNR) ratio
     • Artifact analysis - metal objects, unintended motion, truncation
     • Pediatric Image Quality Performance
     • Low Dose Lung Cancer Screening

6. The type of ground truth used:

   • For the clinical reader study, the ground truth was based on expert assessment/consensus (implying the "gold standard" for diagnostic image quality, noise texture, and sharpness was the radiologists' expert opinion). The cases were "retrospectively collected clinical cases," suggesting the presence of a known clinical diagnosis or outcome, but the specific ground truth for disease presence/absence is not explicitly stated as the primary output of the DLIR evaluation. The evaluation focused more on image quality attributes and comparison between reconstruction methods rather than diagnostic accuracy against a separate definitive truth.

7. The sample size for the training set:

   • The document states the Deep Neural Network (DNN) was "trained on the Revolution family CT Scanners" but does not provide the specific sample size (number of images or cases) used for training.

8. How the ground truth for the training set was established:

   • The text does not explicitly detail how the ground truth for the training set was established. It mentions the DNN was "designed and trained specifically to generate CT Images to give an image appearance... similar to traditional FBP images while maintaining the performance of ASiR-V in the following areas: dose, image noise (pixel standard deviation), low contrast detectability, high-contrast spatial resolution, and streak artifact suppression." This suggests the training likely involved pairing raw CT data with expertly reconstructed ASiR-V or FBP images as a reference for image quality characteristics. The ground truth in this context would be the desired output image characteristics (e.g., low noise, high resolution) that the DLIR algorithm was optimized to reproduce.

Ask a specific question about this device

The Deep Learning Image Reconstruction for Gemstone Spectral Imaging option is a deep learning based CT reconstruction method intended to produce cross-sectional images by computer reconstruction of dual energy X-ray transmission data acquired with Gemstone Spectral Imaging, for all ages. Deep Learning Image Reconstruction for Gemstone Spectral Imaging can be used for whole body, vascular, and contrast enhanced head CT applications.

Deep Learning Image Reconstruction for Gemstone Spectral Imaging (DLIR-GSI) is the next step in CT reconstruction advancement. Like its predicate device (DLIR), DLIR-GSI is an image reconstruction method that uses a dedicated Convolution Neural Network (CNN) that has been designed and trained specifically to reconstruct CT GSI Images to give an image appearance similar to traditional FBP images while maintaining or improving the performance of ASiR-V. The DLIR-GSI can generate monochromatic images (MC), material decomposition images (MD), and virtual unenhanced images (VUE). Multiple MD images such as lodine, Water, Calcium, Hydroxyapatite (HAP), Fat, Uric Acid can be prescribed by the user and generated by the subject device. DLIR-GSI demonstrates same or better Imaging performance as compared to ASiR-V in the following areas: low contrast detectability (LCD), image noise, contrast to noise ratio (CNR), high contrast spatial resolution, CT number accuracy, MD quantification accuracy and metal artifact reduction. Reconstruction times with DLIR-GSI support a normal throughput for routine CT.

The device is marketed as Deep Learning Image Reconstruction for Gemstone Spectral Imaging and the images produced are branded as "TrueFidelity™ CT Images".

Deep Learning Image Reconstruction for Gemstone Spectral Imaging is compatible with dual energy scan modes using the standard kernel and was trained specifically on the Revolution CT family of systems (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 DLR-GSI: Low, Medium, or High. The strength selection will vary with individual users' preference for the specific clinical need.

As compared to the predicate device, the intended use of Deep Learning Image Reconstruction for Gemstone Spectral Imaging does not change (head and whole-body CT image reconstruction). Both algorithms are designed to produce cross-sectional images of the head and body by computer reconstruction of X-ray transmission data, for all ages.

Acceptance Criteria and Study Details for Deep Learning Image Reconstruction for Gemstone Spectral Imaging (DLIR-GSI)

1. Table of Acceptance Criteria and Reported Device Performance:

The study concluding that the device meets the acceptance criteria is based on:

• Bench Testing: Performed on the identical raw datasets obtained on GE's Revolution CT family of systems, applying both DLIR-GSI and ASiR-V reconstructions for comparison.
• Clinical Reader Study: A retrospective study involving radiologists evaluating images reconstructed with both ASiR-V and DLIR-GSI.

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

• Test Set Sample Size:
  • Main Reader Study: 40 retrospectively collected cases.
  • Additional Clinical Evaluation: 7 additional retrospectively collected clinically challenging 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:

• Main Reader Study: 5 board-certified radiologists.
  • Qualifications: Expertise in the specialty areas that align with the anatomical region of each case. (e.g., three readers for body/extremity, three for contrast-enhanced head/neck, one qualified for both). Specific years of experience are not mentioned.
• Additional Clinical Evaluation: 1 board-certified radiologist.
  • Qualifications: Expertise in the specialty area that aligns with all cases containing small, low-contrast objects. Specific years of experience are not mentioned.

4. Adjudication Method for the Test Set:

• Main Reader Study: Each image was read by 3 different radiologists. The radiologists provided an assessment of image quality using a 5-point Likert scale.
  • Adjudication Method: Implicitly, a consensus or agreement among the 3 readers would have been used for the assessment of diagnostic quality and noise texture preference. The document states, "The result of this reader study confirmed that the DLIR-GSI (the subject device) produce diagnostic quality images and have preferred noise texture than the reference device ASiR-V," suggesting that the collective findings of the readers led to this confirmation. Explicit details of a 2+1 or 3+1 adjudication are not provided.
• Additional Clinical Evaluation: A single board-certified radiologist evaluated the 7 challenging cases. No adjudication method was applicable as there was only one reader.

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

• Yes, an MRMC study was performed (the "Clinical Testing" described).
• Effect Size (Improvement with AI vs. without AI assistance): The document states that readers confirmed DLIR-GSI produced "diagnostic quality images" and had "preferred noise texture" compared to ASiR-V (considered without AI assistance in this context, or a lesser AI-powered reconstruction). The additional evaluation confirmed "adequately visualized for diagnostic use" in challenging cases. However, specific quantitative effect sizes (e.g., a percentage improvement in diagnostic accuracy, a specific change in AUC, or a numerical metric of improvement in reader performance) are not provided in the given text.

6. Standalone (Algorithm Only) Performance Study:

• Yes, a standalone performance study was done through engineering bench testing. This testing compared DLIR-GSI (algorithm only) against ASiR-V (reference/control algorithm) using identical raw datasets. Metrics like LCD, image noise, CNR, spatial resolution, CT number accuracy, material decomposition accuracy, iodine detection, and metal artifact reduction were evaluated directly from the reconstructed images without human interpretation.

7. Type of Ground Truth Used:

• Bench Testing: The ground truth for metrics like LCD, image noise, spatial resolution, CT number accuracy, etc., would have been based on physical phantom measurements and known parameters of the phantoms used in the engineering tests.
• Clinical Reader Study: The ground truth for image quality and diagnostic usability was established by expert consensus/interpretation from the board-certified radiologists. The text doesn't mention pathology or outcomes data as the primary ground truth for the reader study, but rather the radiologists' assessment of diagnostic quality and visualization.

8. Sample Size for the Training Set:

• The document states that the neural network was "trained specifically to reconstruct CT GSI Images" using "single energy acquired images on the CT Scanner" (for the predicate) and "dual energy acquired images on the CT Scanner" (for the proposed device). It also mentions "information obtained from extensive phantom and clinical data" was used for noise characteristics.
• However, the specific sample size (number of images or cases) for the training set is NOT provided in the text.

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

• The text implies that the neural network was trained to produce an "image appearance similar to traditional FBP images while maintaining or improving the performance of ASiR-V." This suggests that the "ground truth" for training was implicitly the characteristics of high-quality CT images, likely leveraging existing FBP and ASiR-V reconstructed images from "extensive phantom and clinical data."
• For noise modeling, ground truth was based on "characterization of the photon statistics as it propagates through the preprocessing and calibration imaging chain" and using a trained neural network that "models the scanned object using information obtained from extensive phantom and clinical data."
• Specific details on how the ground truth was rigorously established for the training data (e.g., expert annotations, pathology correlation, quantitative metrics derived from known phantoms) are NOT explicitly described. The training appears to be focused on matching or improving upon established reconstruction methods using a large dataset.

Ask a specific question about this device

The Deep Learning Image Reconstruction software 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 software 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 software support a normal throughput for routine CT.

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.

Deep Learning Image Reconstruction software was initially introduced on the Revolution CT systems (K133705, K163213). The DLR algorithm is now ported to Revolution EVO (K131576), which offers 64 detector row and up to 40mm collimation, and ASIR-V reconstruction option.

Here's a breakdown of the acceptance criteria and study details based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The document doesn't explicitly state quantitative "acceptance criteria" in a pass/fail format with numerical thresholds. Instead, it describes performance goals relative to the predicate device (ASiR-V) or traditional FBP images. The reported device performance generally indicates "as good as or better than" the reference.

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

• Sample Size: 60 retrospectively collected clinical cases.
• Data Provenance: Retrospective. The origin country is not explicitly stated, but the submitter is GE Healthcare Japan Corporation, so it's possible some or all cases originated from Japan or a region where GE Healthcare Japan Corporation operates.

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

• Number of Experts: 7 board-certified radiologists.
• Qualifications: Board-certified radiologists with expertise in the specialty areas that align with the anatomical region of each case. The document does not specify years of experience.

4. Adjudication Method for the Test Set

• Adjudication Method: Each image was read by 3 different radiologists who provided independent assessments of image quality. The readers were blinded to the results of other readers' assessments. There is no explicit mention of an adjudication process (e.g., 2+1 or 3+1 decision) for discrepant reader opinions; it appears the individual assessments were analyzed.

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

• MRMC Study: Yes, a clinical reader study was performed where 7 radiologists read images reconstructed with both ASiR-V (without DLIR) and DLIR.
• Effect Size of Human Reader Improvement: The document states that readers were asked to "compare directly the ASIR-V and Deep Learning Image Reconstruction (DLIR) images according to three key metrics of image quality preference – image noise texture, image sharpness, and image noise texture homogeneity." It reports that the results support substantial equivalence and performance claims and implies a preference for DLIR images, but does not quantify the effect size of how much human readers "improve" with AI assistance in terms of diagnostic accuracy or efficiency. The study primarily focused on radiologists' preference for image quality characteristics.

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

• Standalone Performance: Yes, extensive non-clinical engineering bench testing was performed where DLIR and ASiR-V reconstructions were compared using identical raw datasets. This included objective metrics such as Low Contrast Detectability (LCD), Image Noise (pixel standard deviation), High-Contrast Spatial Resolution (MTF), Streak Artifact Suppression, Noise Power Spectrum (NPS), CT Number Accuracy and Uniformity, and Contrast to Noise (CNR) ratio. This constitutes a standalone (algorithm-only) performance evaluation.

7. The Type of Ground Truth Used

• For the Reader Study (Clinical Performance): The ground truth for evaluating diagnostic use was based on the assessment of image quality related to diagnostic use according to a 5-point Likert Scale by board-certified radiologists. This is a form of expert consensus on image quality suitable for diagnosis, rather than a definitive "truth" established by pathology or patient outcomes.
• For the Bench Testing (Technical Performance): The "ground truth" was the objective measurement of various image quality metrics (e.g., pixel standard deviation for noise, MTF for spatial resolution) in phantoms, which have known properties.

8. The Sample Size for the Training Set

• The document states that the Deep Neural Network (DNN) used in Deep Learning Image Reconstruction was "trained specifically" but does not disclose the sample size of the training set.

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

• The document implies that the DNN was trained to generate CT Images to give an image appearance similar to traditional FBP images while maintaining ASiR-V performance in certain areas. This suggests that existing "traditional FBP images" or images reconstructed with "ASiR-V" served as a reference or a form of "ground truth" for the training process. However, the exact methodology for establishing ground truth during the training phase (e.g., using paired low-dose/high-dose images, or simulated noise reduction) is not detailed in the provided text.

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