(189 days)
LSN (Liver Surface Nodularity) is an image analysis software application intended to assist radiologists and other trained healthcare professionals in analyzing and reporting on the liver morphology depicted in computed tomography (CT) images for use in assessment of chronic liver disease. LSN is designed to assist the user in the evaluation and documentation of liver morphology, specifically liver surface nodularity, provided that the surface nodularity is adequately depicted on the CT images.
LSN provides quantitative metrics related to liver fibrosis by automating segmentation of the liver surface within userdefined Regions of Interest (ROIs) and calculating distances and means related to the liver surface nodularity. LSN also offers reporting capabilities for documenting user-confirmed results, thereby facilitating communication with other trained healthcare professionals and assessment of changes over time.
LSN is intended to provide image-related information that is interpreted by a trained professional, but it does not directly generate any diagnosis. The information provided by LSN should not be used in isolation when making patient management decisions.
LSN is not intended for use with or for the diagnostic interpretation of mammography images.
LSN (Liver Surface Nodularity) is a post-processing software application which assists trained professionals in evaluating DICOM computed tomography image studies of patients with chronic liver disease. The software provides tools to enable the user to make quantitative measurements related to liver surface nodularity as depicted on CT images.
The generated information consists of a LSN Score (reported in tenths of a millimeter), a quantitative measure of the surface nodularity based on a set of user-defined ROIs sampling the liver surface. LSN calculates the distance between the detected liver edge and a smoothed polynomial line (spline) on a pixel-by-pixel basis inside ROIs and reports the mean of these distances on a per-slice basis as well as an overall LSN Score for the imaging series.
LSN provides the user with information the progression of chronic liver disease. LSN does not make clinical decisions and the information provided by LSN must not be used in isolation when making patient management decisions. The LSN Score may provide value by standardizing terminology used to describe surface neporting, thereby facilitating communications between radiologists and other clinicans invalient's treatment planning. In addition, standardized reporting metrics may also be helpful in assessing changes for the same patient over time.
LSN functions by displaying a DICOM CT abdominal series to the user paints a broad region of interest (RO)) delineating the liver edge on a subset of image slices. Then, for the painted region on each slice, the edge is detected using multiple algorithms. For each detected edge, a spline is fit to the edge and the shortes from each edge pixel to the spline are calculated and averaged, resulting in a potential LSN value. The maximum LSN value calculated for an edge is reported as the LSN values for all slices on which ROIs have been painted are then averaged to determine the overall LSN score.
The core LSN algorithms are implemented in platform-independent code, and have been integrated into both a standalone PC research application and a Mac-based viewer plugin for clinical use. Both platforms produce an equivalent LSN score; the sthe algorithm to require less re-work by the user. The clinical version also produces a report containing images of the scores for each slice, and the overall LSN score. The report is produced in both PDF and DICOM formats and is ready for upload to PACS.
The provided text describes the acceptance criteria and a study to prove the device meets these criteria for the LSN (Liver Surface Nodularity) software.
Here's the breakdown of the information requested:
1. A table of acceptance criteria and the reported device performance:
The document describes the acceptance criteria in terms of the results of testing done. While it doesn't present a formal table of quantitative acceptance criteria and corresponding performance metrics, it states general criteria that were met.
| Acceptance Criteria (Inferred from "Testing Information and Performance" section) | Reported Device Performance |
|---|---|
| All product specifications verified. | "All product specifications were verified." |
| Product meets user needs. | "the product to meet user needs was validated." |
| Testing performed according to internal company procedures. | "Testing was performed according to internal company procedures." |
| Software testing and validation conducted according to written testing procedures. | "Software testing and validation were conducted according to written testing was conducted." |
| Test results reviewed by design personnel before software release. | "Test results were reviewed by designals before software proceeded to release." |
| Validation test results support design intent. | "Validation test results support the conclusion that actual device performance satisfies the design intent." |
| Functional verification met design requirements. | "functional verfication...all met design requirements." |
| Licensing met design requirements. | "licensing...all met design requirements." |
| Labeling met design requirements. | "labeling...all met design requirements." |
| Feature functionality met design requirements. | "feature functionality all met design requirements." |
| Arithmetic accuracy verified and validated. | "Arithmetic ... accuracy was veilied and validated by comparison to alternative calculation mechanisms." |
| Report accuracy verified and validated. | "report accuracy was veilied and validated by comparison to alternative calculation mechanisms." |
| Clinical operation validated through usability testing. | "clinical operation was validated through usability testing." |
| LSN output is repeatable for different CT imaging and reconstruction parameters. | "LSN output is repeatable for different CT imaging and reconstruction parameters." |
| LSN output is reproducible across different CT scanner types and vendors. | "reproducible across different CT scanner types and vendors." |
| Intra-observer measurement variability is low. | "the intra- and inter-observer measurement variability is low." |
| Inter-observer measurement variability is low. | "the intra- and inter-observer measurement variability is low." |
| Risk analysis completed and risk control implemented to mitigate unacceptable hazards. | "The LSN risk analysis was completed and risk control mere implemented to mitigate unacceptable hazards." |
| Verification testing results supported claims of substantial equivalence. | "Verfication testing results supported the claims of substantial equivalence." |
2. Sample size used for the test set and the data provenance:
- Test Set Sample Size: The document does not explicitly state the sample size (number of cases or images) used for the testing/validation set.
- Data Provenance: The document does not specify the country of origin of the data or whether it was retrospective or prospective. It generally refers to "different CT imaging and reconstruction parameters" and "different CT scanner types and vendors."
- Racial Backgrounds: "LSN has not been evaluated with images from patients of all ethnicities. It has been primarily evaluated with White and Black racial backgrounds. LSN has not been evaluated with images from pediatric patients."
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
The document mentions "usability testing" and "user expertise" but does not specify a number of experts used to establish ground truth or their specific qualifications (e.g., "radiologist with 10 years of experience"). It only generally refers to "highly-trained healthcare professionals such as radiologists and medical imaging technologists."
4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:
The document does not describe any specific adjudication method (like 2+1 or 3+1) for establishing ground truth on the test set.
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:
The document does not describe an MRMC comparative effectiveness study directly measuring human reader improvement with AI assistance. The study focuses on the device's technical performance and consistency, stating that it "assists radiologists" and "does not directly generate any diagnosis."
6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
The document implies that the device's output (LSN score calculation, segmentation, etc.) was tested for repeatability, reproducibility, and variability, suggesting a standalone component for these technical measurements. However, the overall device function is described as "intended to assist radiologists," meaning it's not purely standalone in its intended clinical use. The "arithmetic and report accuracy" validation could be considered aspects of standalone performance proof.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
The document speaks of "verification" and "validation" against "design requirements" and "alternative calculation mechanisms" for arithmetic accuracy. For "clinical operation," it mentions "usability testing." While it implies the existence of a 'correct' or 'intended' output, it does not explicitly state the specific type of ground truth (e.g., expert consensus readings, histopathology confirmation) used for validating the LSN score itself or the segmentation accuracy.
8. The sample size for the training set:
The document does not mention the sample size for any training set. It primarily discusses "bench testing" and "validation" of the final product.
9. How the ground truth for the training set was established:
Since no training set is mentioned or implied, no information is provided on how its ground truth might have been established. The focus of this document is on the validation of the device for regulatory submission, not its development process.
§ 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.