Search Results

Thoracic VCAR is a CT, non-invasive image analysis software package, which may be used in conjunction with CT lung images to aid in the assessment of thoracic disease. The software provides automatic segmentation of the lungs and automatic segmentation and tracking of the airway tree. Thoracic VCAR also provides quantification of Hounsfield units and display by color of thresholds within a segmented region.

Thoracic VCAR also supports Gemstone Spectral Imaging (GSI) acquisitions for the evaluation of pulmonary perfusion. It provides additional information to aid in visualization of variations of perfusion within the lungs and to quantitatively assess lung volumes. It is intended to be used as an adjunct to current standard methods utilizing color coded displays of iodine attenuation differences in the lungs to aid in identifying segments of relative perfusion differences which may be useful in assessing thoracic disease. Thoracic diseases that may be associated with changes in perfusion include pulmonary embolism and COPD.

Thoracic VCAR is a software analysis package for the Advantage Workstation (AW) platform, CT Scanner, Cloud or PACS stations which can be used in the analysis of CT images. It is designed for the analysis and processing of volumetric CT chest data. It provides quantitative information to aid in the assessment of thoracic diseases.

The primary features of the software are: lung and lobe segmentation to obtain threshold-based volume measurements; bronchial tree segmentation and tracking to determine wall thickness measurements; lung maps based on HU values to help the physician in determining the location and extent of disease across both lungs a well as each lobe. Additionally, GSI datasets can be used for the evaluation of relative perfusion within the lungs.

This document describes the Thoracic VCAR with GSI Pulmonary Perfusion device and its acceptance criteria and supporting studies.

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria are not explicitly stated in a quantitative manner (e.g., minimum sensitivity, specificity, or F1 score). Instead, the document describes the validation in terms of demonstrating effectiveness and improvement in diagnostic value.

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

• Test Set (Clinical): 15 CT cases.
• Data Provenance: The document states this data is "representative of routine clinical imaging from thoracic acquisition perspective." It does not specify the country of origin, but given the FDA submission, it's likely from the US or a region with comparable clinical practices. It is a retrospective assessment as radiologists reviewed existing cases.
• Test Set (Phantom): Not explicitly quantified, but described as "a phantom test with a test setup that simulates regions of relative perfused and non-perfused volumes."

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

• Number of Experts: Three board-certified radiologists.
• Qualifications: "board certified radiologists." No specific years of experience are mentioned.

4. Adjudication Method for the Test Set

The document does not explicitly describe an adjudication method (e.g., 2+1 or 3+1) for the clinical test set to establish a definitive "ground truth" or to resolve discrepancies between readers. It states that the radiologists "assessed" the cases using a 5-point Likert scale, implying their individual assessments were used to determine improvement.

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

• Was an MRMC study done? Yes, a form of clinical assessment with multiple readers (three radiologists) was performed to evaluate the diagnostic value, reader confidence, and efficiency.
• Effect Size of Human Readers' Improvement with AI vs. Without AI Assistance: The document states the assessment "demonstrated the proposed device improves diagnostic value, reader confidence and efficiency." However, it does not provide a specific quantitative effect size (e.g., AUC increase, specific percentage improvement in diagnosis, or a statistical measure of confidence increase). The Likert scale was used, but the results of this scale (e.g., average Likert score, p-values) are not presented in the summary provided. It's implied the improvement is qualitative in nature based on the radiologists' feedback.

6. Standalone (Algorithm Only) Performance Study

• Was a standalone study done? Yes, the phantom study can be considered a standalone performance test for the algorithm's core function of identifying and separating iodine/water distributions. It verifies the algorithm's capability to correctly segment/threshold modeled lung regions based on perfusion differences.

7. Type of Ground Truth Used

• Phantom Study: Simulated ground truth, based on known iodine concentrations in designed regions of the anthropomorphic lung phantom.
• Clinical Study: Expert consensus or individual expert opinion using a 5-point Likert scale. Given there's no mention of adjudication, it's likely based on the aggregated individual expert opinions or a qualitative assessment of general improvement agreed upon by the experts. It's not explicitly stated that an independent "true" ground truth (e.g., pathology, follow-up outcomes) was established for these clinical cases against which the AI or human reader performance was objectively measured.

8. Sample Size for the Training Set

The document does not provide information on the sample size used for the training set of the algorithm.

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

The document does not provide information on how the ground truth for the training set was established.

Ask a specific question about this device