August 3, 2011 Digital Pathology Diagnosis Assistance System Computer s s eye and Pathologist s s eye Ogura M, Saito A, Graf HP, Marugame A, Kiyuna T, Yamashita Y, Cosatto E, Malon C & Fukumoto M. Innovative Solutions Promotion Div., NEC Corp. & Dept pathology, IDAC, Tohoku Univ JAPAN
Tohoku University
1. Earthquake
3. Explosion of NPP 2. Tsunami
Trends in Digital Pathology Olympus Philips Aperio Olympus USA 3DHISTECH(ZEISS) GE Healthcare BioImagene Nikon Leica SCN400 DMetrix
Background of Device Social factors Development -Aging of of Population -Improved early detection by by demands for for the the patient care -National initiatives Clinical affairs -Delay in in computerization -Increase of of tests due due to to the the advancement of of screening -Law suit suit risks for for malpractice Advance in in life life science -Disease mechanism discovery Personalized Medicine -Molecular target drugs eg. eg. Herceptin ( HER2 test) Cancer Diagnosis Support Business System and Service Advance in in technology -Screening equipment -Virtual slide equipment -Image processing (NEC s core- competence)
Outline of NEC System Slides Pathologist NEC Hardware Server, Storage, etc. In maximum, 210 slides Image Analysis Sub-system Reporting Gastric Colorectal Breast Prostate (under development) Report image etc. Hamamatsu NDP Digital image Image Management Sub-system (Databases) GUI Bar-code information Image Clinical Analysis results Customization on the basis of user s workflow Clinical Info. Tissue types, staining, etc.
e-pathologist Technology (1) Mapping of tissue area (2) Analysis of structural features at low magnification (x10) Region of Interest (ROI) (3) Analysis nuclear atypia at high magnification (x40) (4) Integration of results
Gastric Biopsy Image Analysis
Signet-ring Cell Detector of Gastric Cancer
Breast IHC Image Analysis Automatic ROI selection and counting Collaboration with MGH
Pathologists and e-pathologist Pathologist e-pathologist Measurement relative absolute Information circumstantial direct Thinking top-down bottom-up Base case tissue Qualitative Subjective Experience Ideal Yearning Emotional Hate Quantitative Objective Study
What expected from Second opinion Prescreening Telepathology Specimen Archive Quality control/quality assurance
What others? Quantitative measurement tool for assisting diagnosis Navigation for pathologist Solution for reduction of time consuming process Standardization of diagnosis Tool for cost reduction A part of full automatic lab. systems
Current Status in Japan Request for re-exam by a 2nd pathologist Decision by pathologist Case information Pathology report Disagree Approved report Slides Main flow Pathologist Comparison by operator Agree Batch processing Analysis results Image Scanner Hamamatsu Photonics NDP Image Files e-pathologist Marking on the cancerous area
Next step (hardware) Database of Cases Image files Analysis results Full-time Pathologists (Labo A) Sample information e-pathologist Internet Full-time Pathologists (Labo B) Slides Part-time pathologist (access from local office) Pathology report Remote access Approved report Check (technician) - Re-check, second opinion - Need consultation
In the Future: Cloud Service Hospital B Hospital A NEC Cloud Computing Center Clinical info. Slides client ASP server (optional) Internet e-pathologist e-pathologist (local) Pathologists Approved report -Pre-diagnostic -Interactive for second opinion Measurements Remote access Analysis server
In the Future (Software) Related needs in cancer area Image Filing Image Filing and Storage and Storage Telepathology, Telepathology, Remote Remote access access Link Link to to Medical Medical Records Records Market (Clients) 1.Clincal Lab 2. Cancer Center (Prefectural and Regional) 3.Academic Hospital Image Filing Organ Pathology sample Virtual Slide Digital Image Translation between Research and Clinical Areas Value added diagnosis IHC, FISH, ISH (ER,PR,Her2, ) Feature detection Grading/ Therapeutics Non-cancer applications Skin Skin pathology pathology Pharmaceutical Pharmaceutical Animal pathology Animal pathology Pathologist Assistance Feature detection & analysis Lesion detection characterization reporting Gene exp. Protein exp. Chromozome : Research platform Clinical data
Problem to be Solved 1. No standard image format dependent on the maker of scanners 2. Image resolution, Color balances depend on the digitalizing equipment 3. File size too large
Reference timing of analytical results Slides Preparation Scanning e-pathologist Analyze Analysis Results Image Library F A B C D E A: As the tool for Pre-screening tool B: QC/QA for screening process C: ROI selector for navigation D: Checking process E: Final QC/QA F: selection for image archiving Sign out Pre-Screening Time Pathologists Dx e-pathologist is not the substitution for pathologists but a friend of pathologists!
Before Tsunami After Tsunami
Further Discussion & NEC Demonstration Go to Booth 10