Automated Nuclear Morphometry: A Deep Learning Approach for Prognostication in Canine Pulmonary Carcinoma to Enhance Reproducibility

Author:

Glahn Imaine1ORCID,Haghofer Andreas23ORCID,Donovan Taryn A.4ORCID,Degasperi Brigitte5,Bartel Alexander6ORCID,Kreilmeier-Berger Theresa5ORCID,Hyndman Philip S.4ORCID,Janout Hannah23ORCID,Assenmacher Charles-Antoine7ORCID,Bartenschlager Florian8,Bolfa Pompei9ORCID,Dark Michael J.10ORCID,Klang Andrea1,Klopfleisch Robert8ORCID,Merz Sophie11,Richter Barbara1,Schulman F. Yvonne12ORCID,Ganz Jonathan13,Scharinger Josef3ORCID,Aubreville Marc13ORCID,Winkler Stephan M.23,Bertram Christof A.1ORCID

Affiliation:

1. Institute of Pathology, University of Veterinary Medicine Vienna, 1210 Vienna, Austria

2. Bioinformatics Research Group, University of Applied Sciences Upper Austria, 4232 Hagenberg, Austria

3. Department of Computer Science, Johannes Kepler University, 4040 Linz, Austria

4. Department of Anatomic Pathology, The Schwarzman Animal Medical Center, New York, NY 10065, USA

5. University Clinic for Small Animals, University of Veterinary Medicine Vienna, 1210 Vienna, Austria

6. Institute for Veterinary Epidemiology and Biostatistics, Freie Universität Berlin, 14163 Berlin, Germany

7. Comparative Pathology Core, Department of Pathobiology, University of Pennsylvania, Philadelphia, PA 19104, USA

8. Institute of Veterinary Pathology, Freie Universität Berlin, 14163 Berlin, Germany

9. Department of Biomedical Sciences, Ross University School of Veterinary Medicine, Basseterre P.O. Box 334, Saint Kitts and Nevis

10. College of Veterinary Medicine, University of Florida, Gainesville, FL 32611, USA

11. IDEXX Vet Med Labor GmbH, 70806 Kornwestheim, Germany

12. Antech Diagnostics, Mars Petcare Science and Diagnostics, Fountain Valley, CA 92708, USA

13. Department of Computer Science, Technische Hochschule Ingolstadt, 85049 Ingolstadt, Germany

Abstract

The integration of deep learning-based tools into diagnostic workflows is increasingly prevalent due to their efficiency and reproducibility in various settings. We investigated the utility of automated nuclear morphometry for assessing nuclear pleomorphism (NP), a criterion of malignancy in the current grading system in canine pulmonary carcinoma (cPC), and its prognostic implications. We developed a deep learning-based algorithm for evaluating NP (variation in size, i.e., anisokaryosis and/or shape) using a segmentation model. Its performance was evaluated on 46 cPC cases with comprehensive follow-up data regarding its accuracy in nuclear segmentation and its prognostic ability. Its assessment of NP was compared to manual morphometry and established prognostic tests (pathologists’ NP estimates (n = 11), mitotic count, histological grading, and TNM-stage). The standard deviation (SD) of the nuclear area, indicative of anisokaryosis, exhibited good discriminatory ability for tumor-specific survival, with an area under the curve (AUC) of 0.80 and a hazard ratio (HR) of 3.38. The algorithm achieved values comparable to manual morphometry. In contrast, the pathologists’ estimates of anisokaryosis resulted in HR values ranging from 0.86 to 34.8, with slight inter-observer reproducibility (k = 0.204). Other conventional tests had no significant prognostic value in our study cohort. Fully automated morphometry promises a time-efficient and reproducible assessment of NP with a high prognostic value. Further refinement of the algorithm, particularly to address undersegmentation, and application to a larger study population are required.

Publisher

MDPI AG

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3