(26 days)
Lung CARE CT is a self-contained image analysis software package for evaluating CT volume data sets. Combining enhanced commercially available digital image processing tools with optimized workflow and reporting tools, the software is designed to support the physician in confirming the presence of absence of physician-identified lung lesions (eg. nodules) in addition to evaluation, documentation and follow-up of any such lesions using standard or low-dose spiral CT scanning. This evaluation tool allows for volumetric analysis of pulmonary nodule or lesion size over time, helping the Physician to assess the changes in their growth. It is also designed to help the physician classify conspicuous regions of tissue unambiguously, with respect to their size, dimensions, shape and position.
This premarket notification covers Siemens LungCARE CT software package. It is based on Siemens syngo software platform. Lung CARE CT is a self-contained image analysis software package for cvaluating CT volume data sets. Combining enhanced commercially available digital image processing tools (MIP, MPR, SSD, VRT), evaluation tools (volumetric estimation using consistent standardized measurement protocol, comparator tool for nodule matching by synchronization of two datasets, classification of nodules using configurable descriptors) and reporting tools (targeted presets, saved lesion) with optimized workflow palette, the software package is designed to support the physician in confirming the presence of physician identified lung lesions (eg. nodulcs) in addition to evaluation, documentation and follow-up of any such lesions using standard or low-dose spiral CT scanning. This visualization tool allows for volumetric analysis of pulmonary nodule or lesion size over time, helping the Physician to assess the changes in their growth. It is also designed to help the physician classify conspicuous regions of tissue unambiguously, with respect to their size, dimensions, shape and and position.
This document describes the LungCARE CT software package, a 3D CT reconstruction software designed to assist physicians in evaluating lung lesions. The submission includes a summary of pre-clinical and clinical information, and a declaration of substantial equivalence to previously cleared devices.
Here's an analysis of the provided information concerning acceptance criteria and the supporting studies:
1. Table of Acceptance Criteria and Reported Device Performance
The submission does not explicitly state specific quantitative acceptance criteria for performance metrics (e.g., minimum accuracy, sensitivity, or reproducibility thresholds). Instead, the studies focus on demonstrating the capabilities and reproducibility of the device's volumetric measurements under various conditions.
| Acceptance Criterion (Implied) | Reported Device Performance |
|---|---|
| Reproducibility of volumetric measurements (phantom study) | Influencing factors: Imaging parameters (scanning and reconstruction) influence reproducibility. |
| Specific findings: Independent of reconstructed field of view. | |
| Slightly dependent on reconstruction kernel. | |
| Better reproducibility with thin slice collimations. | |
| Normal dose vs. low dose imaging showed no additional benefit to volume estimation. | |
| Recommendation: End-users should consistently use thin slice collimations, medium kernel, low radiation dose with a full field of view reconstruction, and the same protocol for best reproducibility. | |
| Reproducibility of volumetric measurements (clinical study) | Nodule Type Affects Reproducibility: Clearly defined, compact pulmonary nodules showed better volume reproducibility than ill-defined nodules with multiple connections to pleura and/or vessels. |
| User Intervention: The proposed segmentation results were not modified by the user in this study, and the authors concluded that the method allows for reliable estimation of volume growth, but cautioned users to carefully evaluate and critically assess visual representation of segmentation results, especially for ill-defined nodules. | |
| Ability to support physicians in confirming the presence/absence of lesions and their evaluation, documentation, and follow-up. | The device provides tools for: |
- Volumetric estimation using standardized measurement protocol.
- Nodule matching by synchronization of two datasets (comparator tool).
- Classification of nodules using configurable descriptors.
- Targeted presets and saved lesion reporting tools.
- Visualization for volumetric analysis of pulmonary nodule or lesion size over time to assess growth.
- Classification of conspicuous tissue regions by size, dimensions, shape, and position. (Implied, as these are features of the device, not an evaluated performance metric in the studies described). |
2. Sample Sizes and Data Provenance
A. Lung Phantom Bench Testing Study (Kohl et al.):
- Sample Size: Not explicitly stated, but it's a bench testing study using a lung phantom, implying manufactured phantoms rather than patient data.
- Data Provenance: Not applicable as it's a phantom study.
B. Clinical Evaluation (Wormanns et al.):
- Sample Size (Test Set): 10 patients with pulmonary metastatic disease. A total of 150 pulmonary nodules were manually marked and evaluated across these patients.
- Data Provenance: Not explicitly stated (e.g., country of origin), but it is a clinical evaluation, therefore retrospective patient data since the study was conducted to evaluate the software.
3. Number of Experts and their Qualifications for Ground Truth
- A. Lung Phantom Bench Testing Study (Kohl et al.): Not applicable, as this was a phantom study and did not involve human interpretation or ground truth establishment in the traditional sense for medical imaging. The "ground truth" would be the known physical dimensions or changes in the phantom.
- B. Clinical Evaluation (Wormanns et al.): Not explicitly stated how many experts were involved in manually marking the 150 pulmonary nodules. The qualifications of these individuals are also not specified (e.g., "radiologist with 10 years of experience").
4. Adjudication Method
- The information provided does not specify an adjudication method for either study (e.g., 2+1, 3+1, none).
- For the clinical study, it states that "150 pulmonary nodules were manually marked and then evaluated," which implies a single expert or a non-adjudicated process unless specified otherwise.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- No MRMC comparative effectiveness study is described where human readers' performance with and without AI assistance is compared. The clinical study focused on the reproducibility of the software's volumetric measurements, not on the human reader's improvement with the software.
6. Standalone (Algorithm Only) Performance
- Yes, a standalone study was performed. The "Wormanns et al." clinical evaluation assessed the reproducibility of volumetric measurements using LungCARE CT where "The proposed segmentation results, provided by the software package, were not modified by the user." This indicates the study evaluated the algorithm's performance in segmenting and measuring nodules without human intervention to adjust its output.
7. Type of Ground Truth Used
- A. Lung Phantom Bench Testing Study (Kohl et al.): The ground truth would be based on the known physical properties and dimensions of the phantom and the controlled changes applied to it.
- B. Clinical Evaluation (Wormanns et al.): The ground truth for the 150 pulmonary nodules was established by manual marking. This implies expert human identification and marking of the nodules, which then serves as a reference for the software's measurements. This is a form of expert consensus or expert-derived ground truth, though the number and qualifications of experts are not specified, nor is an explicit consensus process mentioned.
8. Sample Size for the Training Set
- The document does not provide any information about the training set used for the LungCARE CT software. The studies described are evaluation studies of an already developed product.
9. How the Ground Truth for the Training Set Was Established
- Since there is no information about the training set, there is also no information on how its ground truth was established.
§ 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.