Urine cell image recognition using a deep-learning model for an automated slide evaluation system.
artificial intelligence
computer-assisted image recognition
deep learning
urine cytology
urothelial carcinoma
Journal
BJU international
ISSN: 1464-410X
Titre abrégé: BJU Int
Pays: England
ID NLM: 100886721
Informations de publication
Date de publication:
08 2022
08 2022
Historique:
revised:
18
05
2021
received:
03
04
2021
accepted:
16
06
2021
pubmed:
19
6
2021
medline:
20
7
2022
entrez:
18
6
2021
Statut:
ppublish
Résumé
To develop a classification system for urine cytology with artificial intelligence (AI) using a convolutional neural network algorithm that classifies urine cell images as negative (benign) or positive (atypical or malignant). We collected 195 urine cytology slides from consecutive patients with a histologically confirmed diagnosis of urothelial cancer (between January 2016 and December 2017). Two certified cytotechnologists independently evaluated and labelled each slide; 4637 cell images with concordant diagnoses were selected, including 3128 benign cells (negative), 398 atypical cells, and 1111 cells that were malignant or suspicious for malignancy (positive). This pathologically confirmed labelled dataset was used to represent the ground truth for AI training/validation/testing. Customized CutMix (CircleCut) and Refined Data Augmentation were used for image processing. The model architecture included EfficientNet B6 and Arcface. We used 80% of the data for training and validation (4:1 ratio) and 20% for testing. Model performance was evaluated with fivefold cross-validation. A receiver-operating characteristic (ROC) analysis was used to evaluate the binary classification model. Bayesian posterior probabilities for the AI performance measure (Y) and cytotechnologist performance measure (X) were compared. The area under the ROC curve was 0.99 (95% confidence interval [CI] 0.98-0.99), the highest accuracy was 95% (95% CI 94-97), sensitivity was 97% (95% CI 95-99), and specificity was 95% (95% CI 93-97). The accuracy of AI surpassed the highest level of cytotechnologists for the binary classification [Pr(Y > X) = 0.95]. AI achieved >90% accuracy for all cell subtypes. In the subgroup analysis based on the clinicopathological characteristics of patients who provided the test cells, the accuracy of AI ranged between 89% and 97%. Our novel AI classification system for urine cytology successfully classified all cell subtypes with an accuracy of higher than 90%, and achieved diagnostic accuracy of malignancy superior to the highest level achieved by cytotechnologists.
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
235-243Informations de copyright
© 2021 The Authors BJU International © 2021 BJU International.
Références
Babjuk M, Burger M, Comperat E, Gontero P, Mostafid A. European association of urology-guidelines-non-muscle-invasive-bladder-cancer-TaT1-CIS-2018. http://uroweb.org/wp-content/uploads/EAU-Guidelines-Non-muscle-invasive-Bladder-Cancer-TaT1-CIS-2018.pdf. Accessed March 25, 2021
Flaig TW, Spiess PE, Agarwal N et al. Bladder cancer, version 3.2020, NCCN clinical practice guidelines in oncology. J Natl Compr Cancer Netw 2020; 18: 329-54
Yafi FA, Brimo F, Steinberg J, Aprikian AG, Tanguay S, Kassouf W. Prospective analysis of sensitivity and specificity of urinary cytology and other urinary biomarkers for bladder cancer. Urol Oncol 2015; 33: e25-31
Blick CGT, Nazir SA, Mallett S et al. Evaluation of diagnostic strategies for bladder cancer using computed tomography (CT) urography, flexible cystoscopy and voided urine cytology: Results for 778 patients from a hospital haematuria clinic. BJU Int 2012; 110: 84-94
Raitanen MP, Aine R, Rintala E et al. Differences between local and review urinary cytology in diagnosis of bladder cancer. an interobserver multicenter analysis. Eur Urol 2002; 41: 284-9
Long T, Layfield LJ, Esebua M, Frazier SR, Giorgadze DT, Schmidt RL. Interobserver reproducibility of The Paris System for Reporting Urinary Cytology. Cytojournal 2017; 14: 17
Layfield LJ, Esebua M, Frazier SR et al. Accuracy and reproducibility of nuclear/cytoplasmic ratio assessments in urinary cytology specimens. Diagn Cytopathol 2017; 45: 107-12
Metter DM, Colgan TJ, Leung ST, Timmons CF, Park JY. Trends in the US and Canadian Pathologist Workforces From 2007 to 2017. JAMA Netw open 2019; 3: e194337
Krizhevsky A, Sutskever I, Hinton GE. ImageNet classification with deep convolutional neural networks. Commun ACM 2017; 60: 84-90
Landau MS, Pantanowitz L. Artificial intelligence in cytopathology: a review of the literature and overview of commercial landscape. J Am Soc Cytopathol 2019; 8: 230-41
Fujisawa Y, Otomo Y, Ogata Y et al. Deep-learning-based, computer-aided classifier developed with a small dataset of clinical images surpasses board-certified dermatologists in skin tumour diagnosis. Br J Dermatol 2019; 180: 373-81
Shkolyar E, Jia X, Chang TC et al. Augmented bladder tumor detection using deep learning. Eur Urol 2019; 76: 714-8
Vaickus LJ, Suriawinata AA, Wei JW, Liu X. Automating the Paris System for urine cytopathology-A hybrid deep-learning and morphometric approach. Cancer Cytopathol 2019; 127: 98-115
Sanghvi AB, Allen EZ, Callenberg KM, Pantanowitz L. Performance of an artificial intelligence algorithm for reporting urine cytopathology. Cancer Cytopathol 2019; 127: 658-66
Awan R, Benes K, Azam A et al. Deep learning based digital cell profiles for risk stratification of urine cytology images. Cytometry Part A 2021. https://doi.org/10.1002/cyto.a.24313. Online ahead of print.
Barkan GA, Wojcik EM, Nayar R et al. The Paris system for reporting urinary cytology: the quest to develop a standardized terminology. Acta Cytol 2016; 60: 185-97
Yun S, Han D, Chun S, Oh SJ, Choe J, CutMix YY. Regularization strategy to train strong classifiers with localizable features. Proc IEEE Int Conf Comput Vis 2019; 6022-31
Zhang H, Cisse M, Dauphin YN, Lopez-Paz D. MixUp: Beyond empirical risk minimization. 6th Int Conf Learn Represent ICLR 2018 - Conf Track Proc 2018; 1-13
He Z, Xie L, Chen X, Zhang Y, Wang Y, Tian Q. Data augmentation revisited: rethinking the distribution gap between clean and augmented data. 2019. http://arxiv.org/abs/1909.09148. Accessed March 25, 2021
Deng J, Guo J, Xue N, Zafeiriou S. ArcFace: Additive angular margin loss for deep face recognition. Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit 2019; 4685
Russakovsky O, Deng J, Su H et al. ImageNet large scale visual recognition challenge. Int J Comput Vis 2015; 115: 211-52
Maier U, Simak R, Neuhold N. The clinical value of urinary cytology: 12 years of experience with 615 patients. J Clin Pathol 1995; 48: 314-7
Miller FS, Nagel LE, Kenny-Moynihan MB. Implementation of the ThinPrep® imaging system in a high-volume metropolitan laboratory. Diagn Cytopathol 2007; 35: 213-7
Eichhorn JH, Buckner L, Buckner SB et al. Internet-based gynecologic telecytology with remote automated image selection: Results of a first-phase developmental trial. Am J Clin Pathol 2008; 129: 686-96
Pantazopoulos D, Karakitsos P, Iokim-Liossi A, Pouliakis A, Botsoli-Stergiou E, Dimopoulos C. Back propagation neural network in the discrimination of benign from malignant lower urinary tract lesions. J Urol 1998; 159: 1619-23
Vriesema JLJ, Van Der Poel HG, Debruyne FMJ, Schalken JA, Kok LP, Boon ME. Neural network-based digitized cell image diagnosis of bladder wash cytology. Diagn Cytopathol 2000; 23: 171-9
Muralidaran C, Dey P, Nijhawan R, Kakkar N. Artificial neural network in diagnosis of urothelial cell carcinoma in urine cytology. Diagn Cytopathol 2015; 43: 443-9
Tan M, Le QV. EfficientNet: Rethinking model scaling for convolutional neural networks. 36th Int Conf Mach Learn ICML 2019, 2019: 10691-700