Digital Pathology and Image Analysis. Queen s University Department of Pathology and Molecular Medicine Shakeel Virk

Transcription

1 Digital Pathology and Image Analysis Queen s University Department of Pathology and Molecular Medicine Shakeel Virk

2 Outline Digital Pathology and Image Analysis capabilities at Queen s Laboratory for Molecular Pathology Thinking outside the box and how digital pathology and image analysis can help your research?

3 Queen s Laboratory for Molecular Pathology Part of Department of Pathology and Molecular Medicine Provides pathology services to researchers, students, clinicians, and private sector groups

4 Queen s Laboratory for Molecular Pathology Services offered by the QLMP include: Histology (block processing, sectioning, H&E staining, IHC, IF) Tissue Microarray construction, FISH, Electron Microscopy, Nanostring, Flow Cytometry, etc.. Digital Pathology Image Analysis

5 Digital Pathology Aperio CS whole slide scanner. Scans bright field images at 20x and 40x. Text box With shadow Aperio FL whole slide scanner. Scans fluorescence images at 20x and 40x Text box With shadow Spectrum Image Management System Aperio Image Analysis tools AQUA Image Analysis Indica Labs HALO image analysis tools

6 Image Analysis Quantitative immunofluorescence as well as bright-field image analysis are available using the Halo software package. Morphological or multiplexed protein expression analysis are carried out across tissue sections or many TMA cores in a manner that permits the exploration of potential associations between biomarker expression, cell type, morphology or clinical parameters.

7 Data Management Users around the world Slide Scanners High Speed Data Transfer Main Data Centre - Web Image Server - Image Analysis Software - TMA Software - Image Management Database - RAID Storage Space - Security - Backup

8 Digital Pathology: Thinking outside the box Thinking outside the box and how digital pathology and image analysis can help your research?

9 Digital Pathology: Thinking outside the box Text box With shadow

10 Thinking outside the box Score Slides Microscope Pathologist Digital Image Score Digital Scanner Image Server Pathologists Image Analysis Score Hospital/Remote Location

11 Example #1: Dr. Anthony Reiman Dr. Anthony Reiman - University of New Brunswick: 20 NSCLC TMA sections ICH stained were mailed to QLMP for scanning Slide were scanned and images were stored in a secure account on the image server Dr. Reiman and his staff log-in and view and manipulate images: Zoom, pan Add annotations Take high resolution snap shots Etc.

12 Example #1: Dr. Anthony Reiman Dr. Reiman wanted to score the slides using image analysis With an account he could run any image analysis already in the system. Unfortunately the existing Image Analysis algorithms were not appropriate for the slide set. Dr. Reiman s student set up a webex call with the QMLP. A custom algorithm was designed for the TMA slide set and uploaded to the image server. Dr. Reiman can now use the custom image analysis at any time on any slides on the server.

13 Example #2: Dr.Harriet Feilotter / CIMTEC (London) TMAs stained with many different antibodies. The antibodies were worked up, scanned and placed online. Pathologists and researchers were given secure access for viewing, annotating and image analysis. Regular teleconferences were held to discuss the project with the shared images.

14 Example # 3: Dr. Johanne Weberpals (Ottawa) Dr. Weberpals obtained CCTG OV.16 clinical trial TMA sections for 300 patients and stained for PTEN and BRCA1 using IF. Images were scanned and analyzed using AQUA image analysis software. BRCA 1 Signal Cytokeratin Signal BRCA 1 and Cytokeratin Signal Gynecologic Oncology Volume 140, Issue 3, pages , 2016.

15 Practical Considerations Text box With shadow

16 Practical Considerations Image Analysis requires a pathologist/expert to ensure the image analysis is measuring what it is supposed to measure. Digital Slide Scanner range in price from $150K to 300K. Storage Space: A typical 20x scanned slide is 500 megabytes (jpeg compression). For approximately 2000 slides 1 terabyte of space is required. Data backup, Image Management System, and IT support. Image analysis software

17 Conclusion In summary, digital pathology and image analysis helps permit the exploration of potential associations between biomarker expression, cell type, morphology or clinical parameters. There already exist nodes across Ontario who have invested and setup digital pathology and image analysis solutions to take advantage of.