WHOLE SLIDE IMAGE SCANNER STUDIES
USCAP Poster Presentation
Optimization of the number of Z-stack layers for whole-slide images of urine cytology with different cellularity by an artificial intelligence algorithm
Resources and News
Educational resources and news from AIxMed to keep you informed
and up-to-date on our AI empowered digital cytology solutions.
and up-to-date on our AI empowered digital cytology solutions.
.png)
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
AIxURO URINE CYTOLOGY CLINICAL STUDIES
March 2025
USCAP 2025 Poster Presentation
Enhancing Bladder Cancer Diagnosis Through the Integration of an AI-Driven Cytology System
USCAP 2025 Poster Presentation
Enhancing Bladder Cancer Diagnosis Through the Integration of an AI-Driven Cytology System
STUDY DESIGN:
- De-identified ThinPrep urine cytology slides (200) were retrospectively selected. Two cytopathologists (CP) provided consensus diagnoses (ground truth, GT) for all cases: 100 Negative for High-Grade Urothelial Carcinoma (NHGUC), 35 Atypical Urothelial Cells (AUC), 32 Suspicious for HGUC (SHGUC), and 33 HGUC.
- Slides were digitized into WSIs utilizing Mikroscan SLxCyto and customized Huron WSI imagers and examined using AI-assisted WSI review (AIxURO)
- 1 cytopathologist (CP) and 2 cytologists (CT) blindly reviewed slides with a 2-week washout period between research arms:
o Arm 1- Microscopy only
o Arm 2- AIxURO Mikroscan SLxCyto’s WSI review
o Arm 3- AIxURO customized Huron’s WSI review
- Performance Metrics:
o Comparison of study diagnosis with ground truth diagnosis for 3 Arms using following thresholds:
(1) AUC+ (AUC, SHGUC and HGUC) cases as positive; NHGUC cases as negative
(2) SHGUC+ (SHGUC and HGUC) cases as positive; NHGUC and AUC cases as negative
o The slide evaluation time (SET) in each arm was documented and made comparison.
RESULTS:
- Diagnostic performance using the AUC+ threshold, AIxURO WSI review (Arm 2 and Arm 3) demonstrated higher sensitivity than microscopy (85.0% and 88.3% vs 79.3% overall). However, AIxURO WSI review exhibited lower specificity than microscopy (85.7% and 82.7% vs. 94.3%).
- When using the SHGUC+ threshold, AIxURO WSI review demonstrated higher overall sensitivity (74.9% and 86.2 vs 76.9%) and slightly lower specificity (96.0% and 92.3% vs 97.5% overall) compared to microscopy.
- AIxURO WSI review markedly reduced the SET versus microscopy (35.9s and 36.4s vs 102.6s). SETs for AUC+ are 45s and 43.6s vs 116.4s, while SET for negative cases are 26.5s and 29.1s vs 88.9s.
CONCLUSIONS:
AIxURO WSI review demonstrated higher sensitivity but lower specificity than microscopy for AUC+ and SHGUC+ thresholds. Notably, AIxURO WSI review reduced slide evaluation time by at least 64.5%, offering substantial efficiency gains. These findings highlight AIxURO’s potential to enhance workflow efficiency in settings withstaffing shortages while maintaining diagnostic performance.
THE PARIS SYSTEM FOR URINE CYTOLOGY NUCLEAR TO CYTOPLASMIC RATIO
March 2025
USCAP 2025 Oral Presentation
Quantitative Analysis of the Paris System Cytomorphological Criteria in Digital Urinary Cytology Using Artificial Intelligence Algorithm
USCAP 2025 Oral Presentation
Quantitative Analysis of the Paris System Cytomorphological Criteria in Digital Urinary Cytology Using Artificial Intelligence Algorithm
STUDY DESIGN:
- 296 urine cytology slides from confirmed biopsy-positive cases within six months (70 AUC, 65 SHGUC,161 HGUC)
- Slides digitized into WSIs with Leica AT2 scanner
- AI algorithm identified and categorized cells with high (suspicious) or low (atypical) malignancy risk based on The Paris System for Reporting Urinary Cytology (TPS) features; Top-24 category order-ranks 24 most abnormal cells
- Performance Metrics: Mean values of suspicious cells, atypical cells, and the top-24 cells (cells ranked by highest probability of malignancy) with TPS features compared across the three categories (HGUC, SHGUC, and AUC).
RESULTS:
- HGUC cases had significantly higher mean number of abnormal cells for top-24 cells (12.2 vs. 5.5 vs. 2.5), suspicious cells (91.6 vs. 13.1 vs. 2.7), and atypical cells (885.1 vs. 108.9 vs. 30.8) compared to SHGUC and AUC, respectively.
- Top-24 and suspicious cells displayed higher N:C ratio (0.65-0.66) and larger nuclear area (106-113 mm 2 ) than atypical cells (0.57 and 86-91 mm 2 ) across three categories.
- Percentage of cells with all 3 key TPS morphological features across 3 categories:
o Top-24 cells- 84%-95%
o Suspicious cells- 81%-91%
o Atypical cells- 22%-39%
- Critical distinguishing feature between suspicious and atypical was hyperchromasia
o 93% of suspicious cells
o 24%-40% of atypical cells.
- Percentage of top-24 / suspicious cells with all 3 TPS features:
o HGUC (95% / 91%)
o SHGUC (89% / 89%)
o AUC (84% / 81%)
CONCLUSIONS:
HGUC showed the highest numbers of abnormal urothelial cells compared to SHGUC and AUC. N/C ratio and nuclear sizes were larger in top-24 and suspicious cell categories compared to the atypical cell category. Hyperchromasia was the most important distinguishing morphologic feature between suspicious and atypical cells. HGUC had the highest % of top-24 and suspicious cells with all 3 TPS morphologic features, indicating that the AI algorithm correctly selected features predicting HGUC. These findings highlight AI’s potential to enhance accuracy and efficiency in bladder cancer diagnosis.
WHOLE SLIDE IMAGE SCANNER STUDIES
March 2025
USCAP 2025 Poster Presentation
Does Digital Scanner Selection Impact Urine Cytology Whole Slide Imaging? A Comparative Study of Five Commercial Digital Scanning Devices
USCAP 2025 Poster Presentation
Does Digital Scanner Selection Impact Urine Cytology Whole Slide Imaging? A Comparative Study of Five Commercial Digital Scanning Devices
STUDY DESIGN:
- Ten de-identified ThinPrep Urocyte slides with surgical concordance as gold standard/ground truth (4 NHGUC, 2 AUC, 1 SHGUC, 3 HGUC) digitized into WSIs using five digital imagers:
o Leica Aperio AT2
o Hamamatsu S360
o 3DHistech P1000
o Customized Huron WSI model
o Mikroscan SLxCyto
- AI algorithm was applied to the WSIs to detect suspicious cells (high-malignant risk) and atypical urothelial cells (low-malignant risk).
- 3 cytologists (CT) reported diagnosis on slides with 2-week washout period between arms:
o Arm 1- Microscopy only
o Arm 2- AI-assisted WSI review
- Performance Metrics:
o WSI conversion rate
o Comparison of binary (positive vs negative) diagnosis (AUC, SHGUC, and HGUC cases = positive) with ground truth diagnosis for 2 Arms
o Quantitative data analysis
RESULTS:
- All digital imagers achieved a 100% WSI conversion rate.
- In the binary diagnosis, AI-assisted WSI review demonstrated performance comparable to microscopy across all imagers
- Quantitative analysis revealed slight variations in key metrics, including suspicious/atypical cell numbers, nuclear-to-cytoplasmic (N:C) ratio, and nuclear area, among WSIs produced by different imagers
CONCLUSIONS:
First study to assess multiple digital pathology imagers for AI-assisted urine cytology. Results demonstrate that a disease-specific AI algorithm produces consistent diagnostic performance and quantitative analysis across different WSIs, supporting its potential to enhance diagnostic accuracy across imaging platforms.
March 2023
Optimization of the number of Z-stack layers for whole-slide images of urine cytology with different cellularity by an artificial intelligence algorithm
ASC/IAC 2022 Poster Presentation / Abstract Hang JF, Ou YC, Tsao TY et al. VGHTE, Tungs Taichung MetroHarbor Hospital
STUDY DESIGN:
- 14 positive (HGUC) urine cytology specimens (cytospin and ThinPrep) digitized with Leica Aperio AT2 scanner at 4 different scan modes:
- Mode A- Default automatic
- Mode B- Optimized automatic
- Mode C- Manually focused points of adjustment
- Mode D- Multilayer Z-stacking
- Performance metrics:
- Rate of successful scanning (total slides that obtain WSI files from scanner);
- Number of high risk cells detected by AI algorithm
- Coverage rate of high risk cells (detected high risk cells in one Z-layer divided by the total high risk cells detected in 21 Z-layer WSI per slide)
RESULTS:
- Advanced scan modes increase successful slide scanning, but does not increase the average high risk cell detection or coverage rates
- More high risk cells were detected with TP than cytospin but with comparable coverage rates
- More Z-layers increased average high risk cell detection and coverage rates
CONCLUSIONS:
Advanced scan modes improve the success rate of obtaining an optimal WSI and the number and coverage rate of high risk cells increases with multiple Z-layers. More high risk cells were detected on ThinPrep vs cytospin but this did not significantly affect the AI algorithm analytic results.
March 2023
SOFTWARE OPTIMIZATION & VALIDATION STUDIES
USCAP Poster Presentation
Optimization of snapshot gallery presentation of artificial intelligence-inferred diagnostic cell images in urine cytology
Optimization of snapshot gallery presentation of artificial intelligence-inferred diagnostic cell images in urine cytology
USCAP 2023 Poster Presentation / Abstract Liu TJ, Yang WL, Tung MC, Hang JF et al. Tungs Taichung MetroHarbor Hospital; VGHTE
STUDY DESIGN:
- 45 HGUC urine cytology slides and 45 negative (NHGUC) slides digitized into WSI and inferred by an AI-algorithm to identify potential cancer cells.
- 12, 24, 36, or 48 abnormal cell thumbnail images were order-ranked by AI and presented in a reviewer gallery
- 2 cytologists reviewed thumbnail images and rendered a diagnosis based on The Paris System for Reporting Urine Cytology 2.0 for each gallery set
- Concordance with microscopy performed for each gallery set, requiring an exact match (HGUC to HGUC, NHGUC to NHGUC)
RESULTS:
- Concordance for HGUC was significantly improved by increasing the number of diagnostic cells in the gallery from 38.9% to 91.1%, but were slightly decreased for NHGUC specimens from 92.2% to 90%.
- The highest concordance rate (95.6%) was established at 36 gallery cell images, but was only slightly decreased with 48 images
CONCLUSIONS:
36 thumbnail cell images provides the highest concordance for HGUC vs NHGUC diagnoses in urine cytology
November 2022
AIxURO URINE CYTOLOGY CLINICAL STUDIES
ASC/IAC Oral Presentation
Validation of a deep-learning-based algorithm for risk stratification and diagnosis of bladder cancer in digital urine cytology
Validation of a deep-learning-based algorithm for risk stratification and diagnosis of bladder cancer in digital urine cytology
ASC/IAC 2022 Oral Presentation / Abstract PLAT 15 Yang WL, Ou YC, Lin YS, Hang JF, Lui TJ. VGHTE, Taichung MetroHarbor Hospital
STUDY DESIGN:
- 104 urine cytology slides (70 NHGUC and 34 positive [AUC, SHGUC, HGUC]), with diagnostic confirmation by an expert panel
- Review of AI (deep learning algorithm)- inferred digital gallery images by 1 cytopathologist (CP) and 2 cytologists (CT), assigning cases to The Paris System (TPS 2.0) diagnostic categories (4 week washout period between Arm 1 and 3)
- Arm 1- Microscopic review of urine cytology glass slide by CP
- Arm 2- WSI review of slide by CP
- Arm 3- AI-assisted review of images by CP and CTs
- Performance Metrics:
- Comparison of diagnostic concordance with sensitivity, specificity, PPV, NPV and accuracy for CP and CTs against initial diagnosis
- Comparison of diagnostic concordance with sensitivity, specificity, PPV, NPV and accuracy between conventional microscopy and AI-assisted image review for the CP
RESULTS:
Microscopic Glass Slide Review versus AI-Assisted Image Review (CP only)
- Sensitivity: 85.2% vs 91.2% (+5.9%)
- Specificity: 100% vs 100%
- PPV: 100% vs 100%
- NPV: 93. vs 95.9% (+2.6%)
- Accuracy: 95.2% vs 97.1% (+1.9%)
AI-Assisted Review (CTA, CTB and CP)
- Sensitivity: 76.5%, 79.4%, 91.2%
- Specificity: 100%, 98.6%, 100%
- PPV: 100%, 96.4%, 100%
- NPV: 89.7%, 90.8%, 95.9%
- Accuracy: 92.3%, 92.3%, 97.1%
CONCLUSIONS:
AI-assisted diagnoses were comparable to expert panel consensus (sensitivity 76.5 – 91.2%) with high specificity (98.6-100%). AI-assisted interpretation showed better sensitivity, NPV, and accuracy than conventional microscopy.
November 2022
WHOLE SLIDE IMAGE SCANNER STUDIES
ASC/IAC Poster Presentation
Optimization of digital urine cytology via focus and Z-layer scan settings of the whole-slide imager
Optimization of digital urine cytology via focus and Z-layer scan settings of the whole-slide imager
ASC/IAC 2022 Poster Presentation / Abstract Hang JF, Ou YC, Tsao TY et al. VGHTE, Tungs Taichung MetroHarbor Hospital
STUDY DESIGN:
- 14 positive (HGUC) urine cytology specimens (cytospin and ThinPrep) digitized with Leica Aperio AT2 scanner at 4 different scan modes:
- Mode A- Default automatic
- Mode B- Optimized automatic
- Mode C- Manually focused points of adjustment
- Mode D- Multilayer Z-stacking
- Performance metrics:
- Rate of successful scanning (total slides that obtain WSI files from scanner);
- Number of high risk cells detected by AI algorithm
- Coverage rate of high risk cells (detected high risk cells in one Z-layer divided by the total high risk cells detected in 21 Z-layer WSI per slide)
RESULTS:
- Advanced scan modes increase successful slide scanning, but does not increase the average high risk cell detection or coverage rates
- More high risk cells were detected with TP than cytospin but with comparable coverage rates
- More Z-layers increased average high risk cell detection and coverage rates
CONCLUSIONS:
Advanced scan modes improve the success rate of obtaining an optimal WSI and the number and coverage rate of high risk cells increases with multiple Z-layers. More high risk cells were detected on ThinPrep vs cytospin but this did not significantly affect the AI algorithm analytic results.
June 2022
AIxURO URINE CYTOLOGY CLINICAL STUDIES
Academic Article: Cancer Cytopathology
Evaluation of an artificial intelligence algorithm for assisting the Paris System in reporting urinary cytology: A pilot study
Evaluation of an artificial intelligence algorithm for assisting the Paris System in reporting urinary cytology: A pilot study
Ou YC, Tsoa TY, Chang MC, et al. Cancer Cytopathology 2022; 130:872-880. Tung University Clinical Study
STUDY DESIGN:
- 131 urine cytology slides digitized and analyzed with AI-assisted algorithm for detecting and ranking abnormal urothelial cells using criteria from The Paris System (TPS2.0); ground truth diagnosis established by an expert panel
- Two arm study with 1 cytopathologist (CP) in both arms and 2 cytologists (CT) in the AI-assisted Arm 2
- Arm 1: CP review of urine cytology glass slide with diagnosis (4 week washout period)
- Arm 2: CP and CT review of AI-assisted inferred WSI of urine cytology slide with quantitative data statistics
- Performance Metrics: Comparison of research TPS diagnosis with expert diagnosis and calculation of sensitivity, specificity, PPV, NPV and accuracy
RESULTS:
- Conventional Microscopy vs AI-Assisted Interpretation
- Sensitivity: 87% vs 92.3% (+5.1)
- Specificity: 100% vs 100%
- PPV: 100% vs 100%
- NPV: 94.8% vs 96.8% (+2.0)
- AI-Assisted Microscopy, CTA vs CTB vs CP interpretation
- Sensitivity: 79.55 VS 82.1% VS 92.3%
- Specificity: 100% vs 98.9% vs 100%
- PPV: 100% vs 97% vs 100%
- NPV: 92% vs 92.9% vs 96.8%
- ĸ (95% CI): 0.845 vs 0.847 vs 0.994
CONCLUSIONS:
Ai-assisted interpretation was comparable to the expert consensus diagnosis, with superior sensitivity and NPV. AI-assisted interpretations showed near perfect agreement with the expert consensus diagnosis (ĸ = 0.944) and the microscopic diagnosis (ĸ = 0.862).
May 2021
AI SOFTWARE DEVELOPMENT
AACR Poster Presentation
Monitoring of stem cell differentiation to mature hepatocytes with a machine learning-based AI model
Monitoring of stem cell differentiation to mature hepatocytes with a machine learning-based AI model
AACR 2021 Poster Presentation / Abstract 185 Cancer Res (2021) 81 (13_Supplement): 185. https://doi.org/10.1158/1538-7445.AM2021-185 Yang WL, Huo Z, Chen SY, Zhu D, Liu TJ, Lee DF.
STUDY DESIGN:
- Develop a machine-learning-based artificial intelligence (AI) model to assist monitoring morphologic changes in human embryonic stem cells (hESC) in color, using bright field microscopy images
- Pilot Study: Train the model to estimate degree of stem cell differentiation at the Hepatic Progenitor Cell (HPC stage), the critical checkpoint for hepatocyte differentiation, based on cellular morphologic features
- Initial training set: Expert annotated images of 341 successful HPC differentiations and 366 failed HPC differentiations
- Cross-validation set: Images of 86 successful and 51 failed HPC results
- Test set: Images of 64 successful and 29 failed HPC results
- Failed differentiation = no differentiation or differentiation into non-hepatocyte tissue types
- Performance Metrics: Accuracy and F1 scores of test set
RESULTS:
The AI model showed excellent performance compared with the conventional method of determining degree of hepatocyte differentiation
- Accuracy = 0.978
- F1 score = 0.975
CONCLUSIONS:
AI-assisted models have the potential to improve the detection of degrees of hepatocyte differentiation, thereby improving the efficiency of a manual process that is very time-intensive.
March 2021
AI SOFTWARE DEVELOPMENT
AACR Virtual Special Conference on AI Poster Presentation
A deep learning model assists urine cytology reporting with computational estimates of the nuclear/cytoplasmic ratios of the urothelial cells based on the Paris System
A deep learning model assists urine cytology reporting with computational estimates of the nuclear/cytoplasmic ratios of the urothelial cells based on the Paris System
AACR Virtual Special Conference on AI 2021 Poster Presentation / Abstract PO-069 Yang WL, Li CB, Ou YC, Lin YS, Tsao TY, ChaAnnong MC, Hang JF, Liu TJ VGHTE, Taichung MetroHarbor Hospital
STUDY DESIGN:
- Descriptive study using deidentified urine cytology slides digitized into WSI
- Artificial intelligence model training on slides with “active learning” to improve results
- Annotated WSI initially used to train computational model, then expert review of results with feedback to the model to learn.
- Sequence was repeated until satisfactory results were achieved
RESULTS:
- AI deep-learning model was able to differentiate nucleus from cytoplasm to calculate N/C ratio using whole slide images (WSI)
- The model correctly provided statistical data (N/C ratio and nuclear size) on cells and successfully categorized them as atypical (NHGUC or AUC) or suspicious (SHGUC or HGUC) cells
CONCLUSIONS:
AI-assistance for interpreting urine cytology using The Paris System for Reporting Urine Cytology has the potential to enhance abnormal cell detection and diagnostic concordance.
June 2020
AI SOFTWARE DEVELOPMENT
AACR Special Conference on Advances in Liquid Biopsies
Artificial intelligence assists automation and high performance of circulating tumor cells enumeration and circulating tumor microemboli characterization in fluorescence microscopy images
Artificial intelligence assists automation and high performance of circulating tumor cells enumeration and circulating tumor microemboli characterization in fluorescence microscopy images
AACR Special Conference on Advances in Liquid Biopsies 2020 Poster Presentation / Abstract B14 Yang WL, Chen JY, Chen S, Liu TJ, Chang YC.
STUDY DESIGN:
- Development of an automated deep-learning AI model for circulating tumor cell (CTC) analysis and enumeration
- Fluorescent microscopy CTC images (CK+/CD45-/DAPI+) collected from blood samples of non-small cell lung carcinoma patient by CMx CTC capture platform
- AI model developed with active learning implemented to train after expert image annotation on 20 slides and validation with 4 extra images
- 18 new test images studied for performance
RESULTS:
- AI model predicted 34% more total CTC than current methods (1775 vs 1328)
- AI model recovered 45% more total CTCs absent from original human annotation (2507 vs 1732 events)
- AI model produced 90% time savings over conventional methods of enumeration (< 20 min vs approximately 4 hours)
- The model correctly characterized features of circulating tumor microthrombi (CTM), including CTC clusters and CTC-associated immune cells
CONCLUSIONS:
An AI model trained to detect and enumerate circulating tumor cells in nonsmall cell lung cancer patients outperformed semiautomated methods, with higher sensitivity and significantly reduced review time (less than 20 minutes) for CTC enumeration in lung cancer specimens.
June 2020
AI SOFTWARE DEVELOPMENT
AACR Poster Presentation
Applying machine learning for urine cytology—computational urothelial carcinoma analysis and diagnosis
Applying machine learning for urine cytology—computational urothelial carcinoma analysis and diagnosis
American Association for Cancer Research (AACR) 2020 Poster Presentation / Abstract B13 Yang WL, Lee CM, Wu ML, Peng YC, Liu TJ, Liu YH.
STUDY DESIGN:
- Development of a deep-learning based image analysis model for cell classification and enumeration in urine cytology
- De-identified whole slide images (WSI) digitized and “active learning” approach used to train the model
- 3 sub-images (3335 cells) annotated by 3 domain experts for initial training
- Cells classified into 7 categories: High grade urothelial carcinoma (HGUC), cluster HGUC, atypical neoplastic cell, atypical reactive cell, inflammatory cell, epithelial cell, and unidentified cell, with expert feedback to the model
- Pilot study after training involved 2 sub-images from 5 digital slides (10 total sub-images)
RESULTS:
- Ai model successfully learned the morphologies of all 6 cell types and was able to quantify total cell counts in each class
CONCLUSIONS:
An artificial intelligence model that enumerates and classifies abnormal urothelial cells may improve urine cytology throughput, accuracy and reproducibility.
Upcoming Events
No items found.
Closed event
American Society of Cytopathology (ASC) 72nd Annual Scientific Meeting
07
-
10
Nov 2024
Hyatt Regency, Orlando, Florida
November
07
-
10
,
2024
Latest news
November 2024
American Society of Cytopathology (ASC)
Dr. Barbara Crothers is the recipient of the ASC’s highest award- 2024 Papanicolaou AwardApril 2024
