Zenit G Sight System Experience G. MOROZZI Università degli Studi di Siena
Scenario Immunofluorescent antinuclear antibody test is still the "gold standard" in the diagnosis of CTD BUT: Standardization is a critical issue due to: subjective interpretation Microscope. Zenit G Sight The novel automated system for image acquisition and interpretation of cell/tissue IFF tests
Hardware PC LED Xeon quadcore 2W Power 4GB RAM B2A filter 4x500GB RAID 10 Ex: 450-490nm Monitor DM: 505 nm LCD Eizo 24 16/9 Color Camera Screen Res. : 1920x1200 QiCam QImaging 12bit Objective Lens 4x 40x Motorized Revolver
Main Advantages No Fluorescence Decay Objective evaluation Analysis and pattern recognition for ANA HEp-2 assays Digitization for IIF substrates wells Navigable wells as a virtual microscope Workflow management File storage and easy retrieval of scanned wells Data and Result exchange with other LABs
Block diagram of a scanning process Scanning Process The scanner transforms each well of a glass slide in a digital file Digital Slide Viewing The digital wells are viewed with the use of a virtual microscope
Well Image Generation Each tile of the area is acquired by the camera to create a unique well image
Navigate the wells of the slides as a Virtual Microscope with the PC monitor
Virtual Well Viewing 10x 20x 40x Wells can be navigated Resolution of well viewing can be changed
Scanning Advantages (1/2) After that a well is digitalized it is always mantained without fluorescence decay effects The specific fluorescence level of a well should be observed also after a long time LED illumination does not photobleach signal as fast as mercury lamps do
Scanning Advantages (2/2) Positive and negative tiles are scanned without out-of-focus No additional filters (DAPI) are required to search focus on negatives Evans Blue doesn t effect the discrimation capabilities of software Scan well area can be reduced in order to save time during acquisition (5-1 ) Positive/Negative - Patterns Discrimination are performed while the system is scanning
Workflow Management Woklist loading (automatic or manual) The worklist contains the number of slides to process, the test to be carried out for each slide and sample data for each well Scanning and automatic analysis of samples The system automatically acquires images and analyze them Validation The operator examine every digital image and navigate the virtual slides. The system diagnosis can be conformed or modified Refertation (to print or export file) The system allows report printing and creation of an output worklist file
AVAILABLE IN JUNE
IN PROGRESS
IN PROGRESS
Our experience..
593 Sera examined in collaboration with UDINE lab. 518 sera for ANA test 75 selected for particular IIF patterns or specificities
Samples selected SSA SSB DNA SP100 Cenp-B AMA-M2 Gp210 RNP Scl70 Jo1 PCNA PMscl Sm P-ribosomiale
ANALYSIS TRADITIONAL MICROSCOPY: G-SIGHT Hep-2000 Analysis by 2 expert operators Pattern and titre Hep-2000 automated system for image acquisition and Interpretation of pos/neg and patterns
Pattern Recognition System has be trained to detect and recognize five patterns: Nucleolar Speckled Homogenuos Centromere Mitocondrial..this topic is in progress
TITRE vs SCORE SCORE: a measure of probability is given as value of confidence that analyzed well is positive AT THE MOMENT NO CORRELATION WITH IIF TITRE!! (in progress) SCORE : 0 100 Algorithm response is: negative-certain (N); negative-uncertain (N?); Positive (P) System has been trained in order to avoid that positive samples are classified as negative (HIGHT SENSITIVITY)
Hep-2000 G-SIGHT(1) SCORE (1) G-SIGHT (2) SCORE (2) ALTRO pattern titolo OMOG 5120 P 100 P 99 NEG N N? 10 NEG N N 1 NUCLEOL 160 N 1.48 N? 20 NEG N N? 2,5 FIN PUNT + NUCLEOL 320 P 82 P 84 FIN PUNT E NUCLEOL 80 N 4.25 N? 29 FIN PUNT 160 N? 37 P 58 FIN PUNT 80 N 6.83 N? 33 NUCLEOL 160 N? 15.03 N? 44 FIN PUNT E OMOG 640 P 99 P 98 PCNA 160 N? 46.7 P 64 MATR NUCL 320 P 98 P 95 GRGR 160 N? 15.2 N? 43 SP100 CENTR GRGR AMA 1280 P 99 P 97 SP100 M2 MEMBRANA NUCL 640 P P 77 anti-gp210 PCNA 640 N? P 59 PCNA OMOG E CITOPL 2560 P P 98 P-RIBOSOMI FUSO MITOTICO 320 N 0.16 N? 8
RESULTS N = 593 Neg POS N? G-SIGHT* 86 (14.5%) 2 ev 161 (27%) 396 (67%) 1 ev 130 (22%) 345 (58%) 65 (11%) Conv. microscopy 397 (67%) 196 (33%) * 1 well UD
Pos P+N? Pos P Neg N False pos N? (pos) pos 506 161** 86* 311* : 291 (N?) 20 (as P) 54 ** 0 (neg) *IIF neg **IIF pos
LOW TITRE Conv. 1:80 1:160 mycroscopy 43 51 (58%) No positive samples have been classified as negative Negative score: 0 2
Results in term of positivity ( P) hom speck nucl centr Nucl. memb 52/77 78% 94/125 75% 21/37 57% 9/10 90% 3/5 60% Nucl matrix 3 /4 75% PCNA cit mit 2 /2 100% 6/ 6 100% 2/2 100% SSA/ SSB 24/26 92% Scl70 CENP Sm RNP 7/7 100% 4/4 100% PCNA/ simil PM/ Scl GP210 Sp100 GrGr M2 Jo1 5/5 2/3 1/1 4/6 1/10 4/4 0/1* * IIF neg
Accuratezza pattern singoli e misti (media 96%) 0 Sconosciuto = 87.5% 1 Centromerico = 100.0% 2 Mitocondriale = 100.0% 3 Nucleolare = 100.0% 4 Omogeneo = 100.0% 5 Granulare = 100.0% Omogeneo-Granulare = 97.7% Nucleolare-Citoplasm. Omogeneo-Citoplasm. = 100.0% Granulare-GRGR = 100.0% Granulare-Nucleolare = 94.1% = 0.0% (1 campione analizzato)
In conclusion, my questions: By automatic systems. Is it useful to define ANA pattern?... (in my opinion ) Is it useful to define the fluorescence titre?... (in my opinion ) Is it useful to to avoid that positive samples are classified as negative? (in my opinion. On the other hands.higher sensitivity or specificity?