Stránka sa načítava...
Progesterone Receptor (1E2) Image Analysis Software User Manual, v. 5.1
Progesteronrezeptor (1E2) Handbuch zur Bildanalyse-Software-Anwendung, v. 5.1
Manuel d’utilisation de l’application logicielle d’analyse d’images du récepteur de
progestérones (1E2), v. 5.1
Manual de usuario de la aplicación del software de análisis de imagen del receptor
de progesterona (1E2), v. 5.1
Manuale dell’utente del software per l’analisi delle immagini - Progesterone
Receptor (1E2), v. 5.1
Uživatelská příručka softwarové aplikace pro analýzu obrazu progesteronového receptoru
(1E2), v. 5.1
Brugervejledning for Softwareapplikation til progesteronreceptor (1E2)-billedanalyse, v. 5.1
Εγχειρίδιο χρήσης εφαρμογής του λογισμικού ανάλυσης εικόνας του υποδοχέα
προγεστερόνης (1E2), v. 5.1
Progeszteron receptor (1E2) képelemző szoftveralkalmazás használati útmutató, v. 5.1
Gebruikershandleiding voor de progesteronreceptor (1E2)
beeldanalyse-softwareapplicatie, v. 5.1
Brukerhåndbok for bildeanalyseprogram for progesteronreseptor (1E2), v. 5.1
Podręcznik użytkownika oprogramowania użytkowego do analizy obrazu receptora
progesteronu (1E2), v. 5.1
Manual do utilizador para o software de aplicação para a análise de imagem do receptor de
progesterona (1E2), v. 5.1
Priročnik za uporabo programske opreme za analizo slik progesteronskih
receptorjev (1E2), v. 5.1
Användarhandbok för programmet för bildanalys av progesteronreceptorn (1E2), v. 5.1
Not for distribution in the US
PS-001378/06449026001
Ventana Medical Systems, Inc. Part No.: 101407100 Rev. A
INTENTIONALLY BLANK
Companion Algorithm Progesterone Receptor (1E2)
Image Analysis Software User Manual, v. 5.1
Ventana Medical Systems, Inc. Part No.: 101407100 Rev. A
Companion Algorithm Progesterone Receptor (1E2) Image Analysis Software User Manual
Copyright
Copyright © 2014. Ventana Medical Systems, Inc. All rights reserved.
Trademarks
BENCHMARK, CONFIRM, COMPANION ALGORITHM, VENTANA ISCAN HT, ISCAN, VIRTUOSO,
VENTANA, and the VENTANA logo are trademarks of Roche. All other trademarks are the property of their respective
owners.
Companion Algorithm PR (1E2) image analysis software is licensed for use between Ventana Medical Systems, Inc. and
a licensee, and only users authorized there under are permitted to access and use the software. Unauthorized use and
distribution may result in civil and criminal penalties.
Open Source and Commercial Software
Refer to the Virtuoso Reference Guide for information on Open Source and Commercial Software programs.
Contact Information:
Ventana Medical Systems, Inc.
1910 E. Innovation Park Drive
Tucson, AZ 85755
USA
+1 520 887 2155
www.ventana.com
Ventana Medical Systems, Inc.
203 Ravendale Drive
Mountain View, CA 94043
USA
Roche Diagnostics GmbH
Sandhofer Strasse 116
D-68305 Mannheim
Germany
+49 621 7590
Part No.: 101407100 Rev. A i
Table of Contents
About the Companion Algorithm PR (1E2) image analysis software . . . . . . . . . . . .1
Topics Covered in this Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Who Should Read this Manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Related Document . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Technical Support. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Connection Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Cyber Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
User Authorization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Securing Networks and Servers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Protecting Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Chapter 1: Intended Use and Indications for Use . . . . . . . . . . . . . . . . . . . . . . . .5
Intended Use and Indications for Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Summary and Explanation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Test Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Warnings and Precautions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Pre-Analytical Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Required Materials Not Provided. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Performance Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Assay Cutoff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Chapter 2: PR Comparison and Reproducibility Studies . . . . . . . . . . . . . . . . . . . .9
PR Marker Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Staining Procedure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Study Devices and Samples for the BenchMark XT and iScan Coreo Slide Scanner . . . . . . . . 9
PR Concordance with BenchMark XT and iScan Coreo . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Intra-System and Inter-System Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
PR Concordance with BenchMark ULTRA Instrument and iScan Coreo Slide Scanner . . . . . 12
PR Concordance with BenchMark XT/ULTRA Instruments and VENTANA iScan HT Slide Scanner13
Appendix A: Reagents (Antibody) Package Inserts. . . . . . . . . . . . . . . . . . . . . . .19
Reagents (Antibody) Package Inserts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
ii Part No.: 101407100 Rev. A
Table of Contents PR (1E2) Image Analysis Software User Manual
INTENTIONALLY BLANK
Part No.: 101407100 Rev. A 1
About the Companion Algorithm PR (1E2) image analysis
software
Welcome to the Companion Algorithm Progesterone Receptor (PR) (1E2) image analysis software User Manual.
Topics Covered in this Chapter
• Who Should Read this Manual (page 1)
• Introduction (page 1)
• Related Document (page 1)
• Technical Support (page 1)
• Connection Requirements (page 1)
• Cyber Security (page 2)
Who Should Read this Manual
System Administrators should read this user manual and use it for reference while operating the Virtuoso software.
Introduction
The Companion Algorithm Progesterone PR (1E2) image analysis software assists the pathologist in the semi-quantitative
measurement of PR in tissues stained with Ventana Medical Systems, Inc. CONFIRM anti-Progesterone Receptor (PR) (1E2) Rabbit
Monoclonal Primary Antibody (CONFIRM anti-PR (1E2)). This application generates an PR score that can be reviewed and
accepted by the pathologist, or if necessary, overridden by the pathologist. The image analysis application is an assist to the
pathologist in the scoring and interpretation of CONFIRM anti-PR (1E2) staining on breast cancer tissues.
The Virtuoso PR Digital Read Application allows the pathologist to view PR stained slides as images on a computer monitor, similar
to what can be viewed under a microscope. While reviewing the image, the pathologist may change magnification and move freely
about the image.
Related Document
For additional information on the Virtuoso software, see the following Ventana document:
• Virtuoso Reference Guide (PL-000123-EN)
Technical Support
Connection Requirements
Refer to the Virtuoso Reference Guide information on connection requirements.
Ventana Medical Systems, Inc.
1910 E. Innovation Park Drive
Tucson, AZ 85755
USA
Tel: +1 520 887 2155; for Technical Support, press 1
2 Part No.: 101407100 Rev. A
About the Companion Algorithm PR (1E2) image analysis software PR (1E2) Image Analysis Software User Manual
Cyber Security
Any device that is connected to a network (internally or externally) has the potential to be compromised by unauthorized access or
viruses. As with most devices, the software is designed to run on a computer utilizing Microsoft Windows and virus protection
software which requires the validation and implementation of the appropriate patches.
Some of the potential cyber security hazards are:
• Malicious software that alters the device software (such as viruses)
• Unauthorized access to the system that could compromise data safety
• Security of data transmitted over the Internet
Cyber security involves protecting data by preventing, detecting, and responding to malicious cyber attacks. Cyber attacks could
involve computer viruses which can completely erase data or hackers who alter files or even use the device as a host to attack other
devices. As serious as these hazards are, steps can be taken to maximize cyber security.
User Authorization
All software users must login with a valid user name and password. The user name and password are securely transmitted in
encrypted form over the Internet or Intranet. Once a user has logged in, the user remains active in the application until the user
explicitly logs out, closes the browser, or because the application closes after a period of inactivity.
Securing Networks and Servers
Network security consists of the provisions made in the computer network infrastructure, policies adopted by the network
administrator to protect the network, and the network resources that prevent unauthorized access.
The following are critical steps for securing a network server:
• Physical security (servers and network infrastructure behind locked doors)
• Use of robust passwords
• System and data backups (at regular intervals)
• Data protection
• Terminating unused services
• Restricting access to used services
The following are critical steps and methodologies used to secure network and servers:
• Data protection
• Data backups (at regular intervals)
• Refusal of automatic updates from off-the-shelf software
• Antivirus software for computers and servers
Protecting Data
Establishment of a network firewall and protection of the network against viruses using anti-virus software are effective methods to
protect data. Virus definitions should be kept up to date and regular scans of computers for spyware should be performed using a
legitimate anti-spyware application. If viruses or spyware are found, remove them immediately.
PR (1E2) Image Analysis Software User Manual About the Companion Algorithm PR (1E2) image analysis software
Part No.: 101407100 Rev. A 3
Evaluate Your Software Settings
The default settings of most software enable all available functionality. However, hackers may be able to take advantage of this
functionality to access devices. It is especially important to check the settings for software that connects to the Internet (browsers,
email clients, etc.). Apply the highest level of security available that still provides needed functionality.
Backup and Recovery
In order to develop a successful backup and recovery plan, comprehension of data accessibility needs and the potential impact of
data loss is essential. Automatic backup procedures need to be adopted using a data backup utility.
4 Part No.: 101407100 Rev. A
About the Companion Algorithm PR (1E2) image analysis software PR (1E2) Image Analysis Software User Manual
INTENTIONALLY BLANK
Part No.: 101407100 Rev. A 5
Chapter 1: Intended Use and Indications for Use
This chapter shows comparison and reproducibility studies for the PR marker.
Intended Use and Indications for Use
The Virtuoso system provides automated digital slide creation, management, analysis, and viewing. It is intended for in vitro
diagnostic use as an aid to the pathologist in the display, detection, counting, review and classification of tissues and cells of clinical
interest based on particular morphology, color, intensity, size, pattern and shape.
The IHC PR (1E2) Digital Read and Image Analysis applications are intended for use as an aid to the pathologist in the detection
and quantitative measurement of progesterone receptor (PR) protein in formalin-fixed, paraffin-embedded normal and neoplastic
tissue. PR results are indicated for use as an aid in the management, prognosis and prediction of therapy outcomes in breast cancer.
When used with Ventana Medical Systems, Inc. CONFIRM anti-Progesterone Receptor (PR) (1E2) Rabbit Monoclonal Primary
Antibody, it is indicated for use as an aid in the assessment of breast cancer patients for whom endocrine treatment is being
considered (but is not the sole basis for treatment).
Note: NOTE: The IHC PR (1E2) Digital Read and Image Analysis applications are adjunctive computer-assisted methodologies for the
qualified pathologist in the acquisition and measurement of images from microscope glass slides of breast cancer specimens
stained for the presence of PR protein. The pathologist should verify agreement with the Image Analysis software application
score. The accuracy of the test results depends on the quality of the immunohistochemical staining. It is the responsibility of
a qualified pathologist to employ appropriate morphological studies and controls as specified in the instructions for the
CONFIRM anti-Progesterone Receptor (PR) (1E2) Rabbit Monoclonal Primary Antibody assay used to assure the validity of the
iScan System for IHC PR Digital Read and Image Analysis scores. The actual correlation of CONFIRM anti-Progesterone Receptor
(PR) (1E2) Rabbit Monoclonal Primary Antibody to clinical outcome has not been established.
Summary and Explanation
The Virtuoso system for PR (1E2) is an instrument and software system designed to assist the qualified pathologist in the consistent
quantitative assessment of protein expression in immunohistochemically (IHC) stained histologic sections from formalin-fixed,
paraffin-embedded (FFPE) normal and neoplastic tissues. The Virtuoso system can be used for review of digitized images of
histologic sections with image analysis algorithms (Companion Algorithm Virtuoso image analysis applications), or without
Companion Algorithm image analysis algorithms (Virtuoso Digital Read applications).
Digital Read applications present images on the computer screen in the same manner as one would see with a manual microscope,
inclusive of the pathologist's ability to select any areas of interest and the option of various magnification levels. For the Companion
Algorithm image analysis applications, the pathologist may use the system software to select and outline one or several field of views,
(FOVs), and each FOV may be viewed at various magnifications and then analyzed by the software; a count of the total number of
target cells and the number interpreted by the algorithm as positive and negative is generated. The pathologist can accept the score
provided by the algorithm, or may override the score with a pathologist score. The system requires competent human intervention
at all steps in the analysis process, and the software makes no independent interpretations of the data.
The Virtuoso system consists of a slide scanner, computer, monitor, keyboard, mouse, Companion Algorithm image analysis
algorithms and software with a user interface. Virtuoso is an end-to-end digital pathology software solution that allows pathology
laboratories to acquire, manage, view, analyze, share, and report on digital images of pathology specimens. Using the Virtuoso
software, the pathologist can view digital images at various magnifications (as previously stated), add annotations, make
measurements, perform image analysis, and generate reports.
For in vitro diagnostic (IVD) use
6 Part No.: 101407100 Rev. A
Chapter 1: Intended Use and Indications for Use PR (1E2) Image Analysis Software User Manual
Test Principles
The Companion Algorithm PR (1E2) image analysis software employs image analysis techniques to obtain PR scoring. Pre-defined
parameters are used to obtain PR scores. The identification of the nucleus is carried out automatically by the image analysis
algorithms. The steps involved in the analysis algorithms are:
1. Enhancing the image. This process increases the contrast to make the image more suitable for analysis.
2. Identifying the epithelial area. The epithelial area is the region of the image where there is the possibility of epithelial cells
being present.
3. Identifying the nucleus.
4. Classifying the cells based on extent, intensity, and thickness of nuclear staining.
5. Computing the score.
Warnings and Precautions
It is important that glass slides with acceptable staining quality be used.
Pre-Analytical Variables
Tissue preparation and staining should follow the recommendations provided in the CONFIRM anti-PR (1E2) assay package insert.
For optimal image capture using the Virtuoso software, it is recommended that the tissue be free of folds and be placed on the slide
with a minimum of 2 mm boundary from the edge on all sides. The cover slip and slide label (if present) should not overhang the
edges of the slide. For further information on scanning, please refer to the appropriate iScan slide scanner reference guide.
Procedure
Refer to the Virtuoso Reference Guide.
It is recommended that at least three field of view be selected for analysis when using the Companion Algorithm PR (1E2) software.
Required Materials Not Provided
The Virtuoso System for PR 1E2 requires use of CONFIRM anti-PR (1E2), and any additional material or supplies listed in the
Ventana package insert, to stain tissues prior to analysis.
The iScan Coreo slide scanner or the VENTANA iScan HT slide scanner is required for scanning of the slides.
Results
The Companion Algorithm PR 1E2 image analysis software produces images and a staining score. The pathologist views the image
and the instrument score, makes an assessment, and reports a score which may not be the same as the instrument score. Refer to
the Virtuoso Reference Guide for an example of a report.
Limitations
The algorithms are designed to work for PR cell nuclei staining. The test results are only as good as the quality and accuracy of the
immunohistochemistry slide that is imaged, and the subsequent image that is analyzed. The pathologist must validate the PR staining
run by examination of the PR control images to verify that the expected results have been obtained before images from patient slides
PR (1E2) Image Analysis Software User Manual Chapter 1: Intended Use and Indications for Use
Part No.: 101407100 Rev. A 7
are analyzed. The pathologist must follow the manufacturer's recommendations for CONFIRM anti-PR (1E2) including using all
the positive and negative quality control materials for each staining run. If the control slides are not acceptable, the patient tissues
need to be re-stained with acceptable results. (See the CONFIRM anti-PR (1E2) package insert for details about quality control
recommendations.) The pathologist must follow the Ventana recommendations for surveying the entire breast cancer specimen to
assess any heterogeneity in the CONFIRM anti-PR (1E2) staining, the degree of background staining, cytoplasmic staining, edge
effect, etc. as recommended in the CONFIRM anti-PR (1E2) user's manual (available at www.ventana.com). If the images captured
have different staining (nuclear, cytoplasm, etc.), incorrect results will be generated. Incorrect results will also be generated if the
image quality cannot be analyzed. The software algorithms determine whether the quality of an image can be analyzed, based on
pre-defined parameters. Refer to the Virtuoso Reference Guide for more information.
The PR (1E2) image analysis algorithm will reject nuclei that are elongated regardless of the overall shape of the cell. For this reason,
tumors containing large numbers of cells with elongated nuclei may need to be evaluated manually. In addition, performance of the
Virtuoso System with the following types of breast cancers has not been evaluated: carcinoma in situ, carcinosarcoma, comedo
carcinoma, cystosarcoma phylloides, medullary carcinoma of the breast, mucinous variants of breast cancer, and spindle cell
carcinoma.
Performance Characteristics
Performance of the staining agent is described in the Ventana package insert for the CONFIRM anti-PR (1E2) staining process. See
Comparison and Reproducibility Studies in Chapter 2 for a description of the performance of the software.
Assay Cutoff
The performance of the Virtuoso System with PR 1E2 was evaluated at the CAP/ASCO recommended clinical cutoffs: 0-0.99%
was considered a negative test result and 1-10% and >10% were considered positive test results.
8 Part No.: 101407100 Rev. A
Chapter 1: Intended Use and Indications for Use PR (1E2) Image Analysis Software User Manual
INTENTIONALLY BLANK
Part No.: 101407100 Rev. A 9
Chapter 2: PR Comparison and Reproducibility Studies
This chapter shows comparison and reproducibility studies for the PR marker.
PR Marker Studies
Staining Procedure
Refer to the CONFIRM anti-PR (1E2) package insert for BenchMark XT, ULTRA and GX instruments, ultraView, and iView
detection.
Study Devices and Samples for the BenchMark XT and iScan Coreo Slide Scanner
The PR comparison and reproducibility studies for the Virtuoso Digital Read and Companion Algorithm image analysis software
consisted of 120 de-identified archived breast carcinoma sections immunohistochemically stained with CONFIRM anti-PR (1E2).
Study test samples covered the ranges of 0-0.99%, 1-10% and >10%, and were interpreted at three different sites by three different
pathologists. All test slides were scanned at 20X magnification and all images were output in bif file format. For the image analysis
evaluation, a minimum of three fields of view were selected by each pathologist independently for each image.
PR Concordance with BenchMark XT and iScan Coreo
The table below shows the concordance results for CONFIRM anti-PR (1E2) staining interpretation among three different sites:
1. Digital Read vs. Manual Method.
Table 2-1 Agreement - Digital Read vs. Manual (manual = true score).
Confusion Matrix
Digital
Site 1 Site 2 Site 3
(n =112) (n =114) (n =116)
Neg Pos Neg Pos Neg Pos
Manual
Neg (0-0.99%) 50 3 51 0 52 1
Pos (1-10%, >10%) 3 56 3 60 8 55
% Agreement 95% 97% 92%
(95% CI) (89% -98%) (93% - 99%) (86% - 96%)
10 Part No.: 101407100 Rev. A
Chapter 2: PR Comparison and Reproducibility Studies PR (1E2) Image Analysis Software User Manual
2. Image Analysis vs. Manual Method.
Intra-System and Inter-System Studies
The study was designed to demonstrate inter- and intra-Virtuoso system reproducibility for Virtuoso Digital Read and Companion
Algorithm image analysis applications. A designated subset of 40 cases that span the range of the PR scoring categories (0-0.99%,
negative, 1-10%, positive, and >10%, positive) were used.
1. Intra-pathologist/Inter-Day (pair-wise comparisons, Session1 vs. Session 2, Session1 vs. Session 3, Session 2 vs. Session 3).
Table 2-2 Agreement - Image Analysis vs. Manual (manual = true score).
Confusion Matrix
Image Analysis
Site 1 Site 2 Site 3
(n = 112) (n = 115) (n = 114)
Neg Pos Neg Pos Neg Pos
Manual
Neg (0-0.99%) 50 2 51 2 51 1
Pos (1-10%, >10%) 7 53 2 60 8 54
% Agreement 92% 97% 92%
(95% CI) (85% - 96%) (91% -99%) (86% - 96%)
Table 2-3 Intra- Pathologist Digital Read.
Confusion Matrix
Intra-Pathologist Digital
Session 2 Session 3 Session 3
Neg Pos Neg Pos Neg Pos
19 20 21 18 21 18
Session 1
Neg 20 17 2 18 1
Pos 20 2 18 3 17
Session 2
Neg 19 17 2
Pos 20 416
% Agreement 90% 90% 85%
(95% CI) (76% - 96%) (76% - 96%) (70% - 93%)
PR (1E2) Image Analysis Software User Manual Chapter 2: PR Comparison and Reproducibility Studies
Part No.: 101407100 Rev. A 11
2. For Intra-Pathologist Image Analysis.
3. Inter-pathologist (pair-wise comparisons, Pathologist 1 vs. Pathologist 2, Pathologist 1 vs. Pathologist 3, Pathologist 2 vs.
Pathologist 3).
4. Inter-Pathologist Image Analysis.
Table 2-4 Intra-Pathologist Image Analysis.
Confusion Matrix
Intra-Pathologist Image Analysis
Session 2 Session 3 Session 3
Neg Pos Neg Pos Neg Pos
17 22 18 21 18 21
Session 1
Neg 18 17 1 17 1
Pos 21 0 21 1 20
Session 2
Neg 17 17 0
Pos 22 121
% Agreement 97% 95% 97%
(95% CI) (87% - 100%) (83% - 99%) (87% - 100%)
Table 2-5 Inter-Pathologist Digital Read.
Confusion Matrix
Inter-Pathologist Digital
Site 2 Site 3 Site 3
Neg Pos Neg Pos Neg Pos
54 60 60 56 60 56
Site 1
Neg 53 51 1 50 2
Pos 59 2 57 7 52
Site 2
Neg 54 50 2
Pos 60 654
% Agreement 97% 92% 93%
(95% CI) (92% - 99%) (85% - 96%) (87% - 96%)
Table 2-6 Inter-Pathologist Image Analysis.
Confusion Matrix
Inter-Pathologist Image Analysis
Site 2 Site 3 Site 3
Neg Pos Neg Pos Neg Pos
53 62 60 55 60 55
Site 1
Neg 57 50 6 54 2
Pos 55 0 55 3 52
Site 2
Neg 53 52 1
Pos 62 854
% Agreement 95% 95% 92%
(95% CI) (89% - 97%) (90% - 98%) (86% - 96%)
12 Part No.: 101407100 Rev. A
Chapter 2: PR Comparison and Reproducibility Studies PR (1E2) Image Analysis Software User Manual
PR Concordance with BenchMark ULTRA Instrument and iScan Coreo Slide Scanner
Study Devices and Samples for BenchMark ULTRA Instrument and iScan Coreo Slide Scanner
These PR (1E2) comparison studies for the Virtuoso Digital Read and Companion Algorithm Image Analysis software consisted of
120 de-identified archived breast carcinoma sections immunohistochemically stained with CONFIRM anti-PR (1E2) on the
BenchMark ULTRA instrument. Study test samples covered the ranges of 0-0.99%, 1-10% and >10%, and were interpreted by one
pathologist at one site. All test slides were scanned at 20X magnification and all images were output in bif file format. Three fields
of view were selected by the pathologist for each image when using the image analysis application. 113 of 120 cases were able to be
evaluated and were included in the analyses below.
The tables below shows the concordance results for CONFIRM anti-PR (1E2) assay stained on the BenchMark ULTRA instrument.
Digital Read vs. Manual
Image Analysis vs. Manual
Table 2-7 Agreement (ULTRA Instrument) - Digital Read vs. Manual
Manual Microscope Read
Digital Read
Positive Negative Total
Positive (2+, 3+) 68 1 69
Negative (0, 1+) 4 40 44
Total 72 41 113
Positive Percent Agreement (PPA) n/N (%) (95% CI) 68/72 (94.4%) (86.6-97.8%)
Negative Percent Agreement (NPA) n/N (%) (95% CI) 40/41 (97.6%) (87.4-99.6%)
Overall Percent Agreement (OPA) n/N (%) (95% CI) 108/113 (95.6%) (90.1-98.1%)
Table 2-8 Agreement (ULTRA Instrument) - Image Analysis vs. Manual
Manual Microscope Read
Image Analysis Read
Positive Negative Total
Positive (2+, 3+) 71 1 72
Negative (0, 1+) 3 39 42
Total 74 40 114
Positive Percent Agreement (PPA) n/N (%) (95% CI) 71/74 (95.5%) (88.7-98.6%)
Negative Percent Agreement (NPA) n/N (%) (95% CI) 39/40 (97.5%) (87.1-99.6%)
Overall Percent Agreement (OPA) n/N (%) (95% CI) 110/114 (96.5%) (91.3-98.6%)
PR (1E2) Image Analysis Software User Manual Chapter 2: PR Comparison and Reproducibility Studies
Part No.: 101407100 Rev. A 13
PR Concordance with BenchMark XT/ULTRA Instruments and VENTANA iScan HT Slide
Scanner
Study Devices and Samples for BenchMark XT/ULTRA Instruments and iScan HT Slide Scanner
The PR comparison and reproducibility studies for the Virtuoso Digital Read and Companion Algorithm image analysis software
consisted of 30 de-identified archived breast carcinoma sections immunohistochemically stained with CONFIRM anti-PR (1E2).
Study test samples covered the ranges of 0-0.99% (negative), 1-10% (positive) and >10% (positive), and were interpreted at by one
pathologist. All test slides were scanned at 20X magnification and all images were output in bif file format. Three fields of view were
selected by the pathologist for each image when using the image analysis application.The tables below shows the concordance results
for CONFIRM anti-PR (1E2) assay stained on the BenchMark ULTRA instrument.
Digital Read vs. Manual
Image Analysis vs. Manual
Table 2-9 Agreement - Digital Read vs. Manual Microscopic Read
Digital Read
Manual Microscope Read
Negative Positive Total
Negative 10 1 11
Positive 0 17 17
Total 10 18 28
Positive Percent Agreement (PPA) n/N (%) (95% CI) 17/17 (100%) (82-100%)
Negative Percent Agreement (NPA) n/N (%) (95% CI) 10/11 (91%) (62-98%)
Overall Percent Agreement (OPA) n/N (%) (95% CI) 27/28 (96%) (82-99%)
Table 2-10 Agreement - Image Analysis vs. Manual Microscopic Read
Image Analysis Read
Manual Microscope Read
Negative Positive Total
Negative 10 1 11
Positive 1 16 17
Total 11 17 28
Positive Percent Agreement (PPA) n/N (%) (95% CI) 16/17 (94%) (73-99%)
Negative Percent Agreement (NPA) n/N (%) (95% CI) 10/11 (91%) (62-98%)
Overall Percent Agreement (OPA) n/N (%) (95% CI) 26/28 (93%) (77-98%)
14 Part No.: 101407100 Rev. A
Chapter 2: PR Comparison and Reproducibility Studies PR (1E2) Image Analysis Software User Manual
PR Concordance with BenchMark GX Instrument, iScan Coreo, and iScan HT Slide Scanners
Study Devices and Samples for BenchMark GX Instrument, iScan Coreo, and iScan HT Slide Scanners
The PR comparison and reproducibility studies for the Virtuoso Digital Read and Companion Algorithm image analysis software
consisted of 30 de-identified archived breast carcinoma sections immunohistochemically stained with CONFIRM anti-PR (1E2).
Study test samples covered the ranges of 0-0.99% (negative), 1-10% (positive) and >10% (positive), and were interpreted by two
pathologists. All test slides were scanned at 20X magnification and all images were output in bif file format. For the image analysis
evaluation, a minimum of three fields of view were selected by each pathologist independently for each image.
The results were analyzed for each pathologist individually and also averaged across the two pathologists. The average overall
percent agreement (OPA) was determined as the weighted average for the two pathologists who independently scored the same
cases, where the weight of a pathologist's individual OPA rate was determined by the number of cases considered evaluable by that
pathologist.
Digital Read vs. Manual Microscopic Read, iScan Coreo Slide Scanner
The average OPA for Digital Read vs. Manual Microscopic Read for iScan Coreo slide scanner was 92.2%.
Table 2-11 Pathologist #1 Agreement - Digital Read vs. Manual Microscopic Read, iScan
Coreo Slide Scanner
Digital Read
Manual Microscope Read
Negative Positive Total
Negative 12 3 15
Positive 1 8 9
Total 13 11 24
Overall Percent Agreement (OPA) n/N (%) (95% CI) 83.3%
Table 2-12 Pathologist #2 Agreement - Digital Read vs. Manual Microscopic Read,
iScan Coreo Slide Scanner
Digital Read
Manual Microscope Read
Negative Positive Total
Negative 11 0 11
Positive 0 16 16
Total 11 16 27
Overall Percent Agreement (OPA) n/N (%) (95% CI) 100.0%
/
