(106 days)
CT:VQ software is a non-invasive image post-processing technology, using CT lung images to provide clinical decision support for thoracic disease diagnosis and management in adult patients. It utilizes two non-contrast chest CT studies to quantify and visualize ventilation and perfusion.
Quantification and visualizations are provided as DICOM images. CT:VQ may be used when Radiologists, Pulmonologists, and/or Nuclear Medicine Physicians need a better understanding of a patient's lung function and/or respiratory condition.
CT:VQ is a Software as a Medical Device (SaMD) technology, which can be used in the analysis of a paired (inspiratory/expiratory) non-contrast Chest CT. It is designed to measure regional ventilation (V) and regional perfusion (Q) in the lungs.
The Device provides visualization and quantification to aid in the assessment of thoracic diseases. These regional measures are derived from the lung tissue displacement, the lung volume change, and the Hounsfield Units of the paired (inspiratory/expiratory) chest CT.
The Device outputs DICOM images containing the ventilation output and perfusion output, consisting of a series of image slices generated with the same slice spacing as the expiration CT. In each slice the intensity value for each voxel represents either the value of ventilation or the value for perfusion, respectively, at that spatial location. Additional Information sheet is also generated containing quantitative data, such as lung volume.
Here's a breakdown of the acceptance criteria and the study details for the CT:VQ device, based on the provided FDA 510(k) summary:
Acceptance Criteria and Device Performance
The acceptance criteria for CT:VQ are implicitly demonstrated through its strong performance in clinical studies, showing agreement with established gold standards. While explicit numerical acceptance criteria are not provided in a table format within the summary, the narrative describes the goals of the study:
- Consistency/Agreement with Nuclear Medicine Imaging (SPECT/CT): The device's regional ventilation and perfusion measurements should align well with SPECT/CT findings.
- Correlation with Pulmonary Function Tests (PFTs): CT:VQ metrics should statistically correlate with standard PFTs like DLCO and FEV1/FVC ratio.
- Interpretability and Clinical Actionability: The outputs should be clear, understandable, and useful for clinicians.
- Safety and Effectiveness Profile: The device should have a safety and effectiveness profile similar to the primary predicate device.
Table of Acceptance Criteria and Reported Device Performance (as inferred from the text):
| Acceptance Criterion (Inferred) | Reported Device Performance |
|---|---|
| Strong regional agreement with SPECT VQ (Ventilation) | CT:VQ showed strong regional agreement with SPECT VQ across lobar distributions of ventilation. In the Reader Performance Study, clinicians consistently rated CT:VQ outputs as having good to excellent agreement with SPECT across all lung regions. |
| Strong regional agreement with SPECT VQ (Perfusion) | CT:VQ showed strong regional agreement with SPECT VQ across lobar distributions of perfusion. In the Reader Performance Study, clinicians consistently rated CT:VQ outputs as having good to excellent agreement with SPECT across all lung regions. |
| Correlation with Gas Transfer Impairment (DLCO) | Quantitative perfusion heterogeneity metrics derived from CT:VQ demonstrated stronger associations with gas transfer impairment (DLCO) than those derived from SPECT, suggesting improved physiological sensitivity. There was a statistically significant correlation between the CT:VQ and PFT outputs. |
| Correlation with Airway Obstruction (FEV1 and FEV1/FVC % predicted) | Ventilation heterogeneity metrics from CT:VQ correlated well with FEV1 and FEV1/FVC % predicted. There was a statistically significant correlation between the CT:VQ and PFT outputs. |
| Interpretability and Clinical Actionability by Intended Users | The Reader Performance Study affirmed that CT:VQ outputs are interpretable and clinically actionable by intended users. |
| Inter-reader variability similar to SPECT | Inter-reader variability was not significantly different for CT:VQ than for SPECT. |
| Feasibility of generating reliable and consistent data | The clinical studies successfully demonstrated the feasibility of generating valid data that is reliable and consistent with Nuclear Medicine Ventilation imaging results. |
| Safety and effectiveness profile similar to predicate device | Based on the clinical performance, CT:VQ was found to have a safety and effectiveness profile that is similar to the primary predicate device. It also demonstrated the capability to provide information without contrast agents (unlike some alternative perfusion methods). |
| Robustness across various CT inputs | Verification testing demonstrated that the Device was robust within acceptable performance limits across the entire range of inputs (CT scanners, institutions, varying lung volumes, image properties affecting voxel size and SNR). Specific performance limits are not quantified in the summary, but the general claim of robustness is made. |
Study Details
Here's a breakdown of the specific information requested about the studies:
1. Sample sizes used for the test set and the data provenance:
- Test Set Sample Sizes:
- Reader Performance Study: n=77
- Standalone Performance Assessment: n=58 (a subset of the overall clinical studies data)
- Data Provenance:
- Country of Origin: Not explicitly stated, but the submission is from 4DMedical Limited in Australia, and the FDA clearance is in the US. The description mentions "clinically-acquired data included paired chest CTs acquired on CT scanners across a range of manufacturers and models and at different institutions, across a diverse range of patients." This suggests multi-institutional data, potentially from various geographic locations, but this is not confirmed.
- Retrospective or Prospective: Not explicitly stated whether the studies were retrospective or prospective. The description "clinical studies were also conducted to demonstrate the safety and efficacy... in the context of clinical care" and comparing with "gold-standard and best practice measures for respiratory diagnosis" often implies retrospective analysis of existing data combined with prospective data collection, but this is not definitive in the text.
2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Number of Experts: Not explicitly stated for establishing ground truth, although for the Reader Performance Study, "clinicians with expertise in thoracic imaging and pulmonary care" were involved in rating the outputs. The implication is that these experts, along with SPECT/CT and PFT results, contributed to the ground truth.
- Qualifications of Experts: "Clinicians with expertise in thoracic imaging and pulmonary care." No specific number of years of experience or board certifications (e.g., radiologist with 10 years of experience) is provided.
3. Adjudication method (e.g. 2+1, 3+1, none) for the test set:
- Adjudication Method: Not explicitly stated. The summary mentions "Inter-reader variability was not significantly different for CT:VQ than for SPECT," which implies multiple readers, but the method for resolving discrepancies or establishing a final ground truth from multiple readers is not detailed.
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:
- MRMC Study: A "Reader Performance Study" was conducted with n=77 cases, involving "clinicians with expertise in thoracic imaging and pulmonary care." This aligns with the characteristics of an MRMC study.
- Effect Size of Human Reader Improvement with AI vs. without AI assistance: The summary does not provide an effect size or direct comparison of human reader performance with CT:VQ assistance versus without it. The study focused on assessing:
- Agreement between CT:VQ outputs and SPECT.
- Interpretability and clinical actionability of CT:VQ outputs.
- Inter-reader variability of CT:VQ vs. SPECT.
It does not quantify an improvement in reader accuracy or efficiency due to AI assistance.
5. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- Standalone Performance: Yes, a "Standalone Performance Assessment" was performed with a subset of 58 cases. The findings indicated strong regional agreement between CT:VQ and SPECT VQ measurements and stronger associations of CT:VQ perfusion metrics with DLCO compared to SPECT.
6. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- Type of Ground Truth: A combination of established clinical diagnostics was used:
- Nuclear Medicine Imaging (Single photon emission computed tomography, SPECT/CT): Used as a "gold-standard and best practice measure" for regional ventilation and perfusion.
- Pulmonary Function Tests (PFTs): Specifically Diffusing capacity of the lung for carbon monoxide (DLCO) and FEV1/FVC ratio, used to correlate with CT:VQ outputs.
- Clinical Diagnosis/Findings: Implied through "Case Studies further illustrated key advantages of CT:VQ... successfully replicated the diagnostic findings of SPECT."
7. The sample size for the training set:
- Training Set Sample Size: Not explicitly stated in the provided text. The summary only mentions the sample sizes for the clinical validation studies (test sets).
8. How the ground truth for the training set was established:
- Training Set Ground Truth Establishment: Not explicitly stated how the ground truth for the training set was established, as the training set size and characteristics are not detailed. Typically, it would involve similar rigorous processes (e.g., expert annotation, gold-standard imaging modalities, clinical outcomes) as the test set, but this information is absent in this document.
§ 892.1750 Computed tomography x-ray system.
(a)
Identification. A computed tomography x-ray system is a diagnostic x-ray system intended to produce cross-sectional images of the body by computer reconstruction of x-ray transmission data from the same axial plane taken at different angles. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II.