Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma

BackgroundAn accurate pathological diagnosis of hepatocellular carcinoma (HCC), one of the malignant tumors with the highest mortality rate, is time-consuming and heavily reliant on the experience of a pathologist. In this report, we proposed a deep learning model that required minimal noise reducti...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Shi Feng, Xiaotian Yu, Wenjie Liang, Xuejie Li, Weixiang Zhong, Wanwan Hu, Han Zhang, Zunlei Feng, Mingli Song, Jing Zhang, Xiuming Zhang
Formato:	article
Lenguaje:	EN
Publicado:	Frontiers Media S.A. 2021
Materias:	HCC classification pathological images deep learning whole-slide image noisy annotation Neoplasms. Tumors. Oncology. Including cancer and carcinogens RC254-282
Acceso en línea:	https://doaj.org/article/ad8d7d83bec544e19696d05b87852c76
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ad8d7d83bec544e19696d05b87852c76
record_format	dspace
spelling	oai:doaj.org-article:ad8d7d83bec544e19696d05b87852c762021-12-01T21:24:10ZDevelopment of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma2234-943X10.3389/fonc.2021.762733https://doaj.org/article/ad8d7d83bec544e19696d05b87852c762021-12-01T00:00:00Zhttps://www.frontiersin.org/articles/10.3389/fonc.2021.762733/fullhttps://doaj.org/toc/2234-943XBackgroundAn accurate pathological diagnosis of hepatocellular carcinoma (HCC), one of the malignant tumors with the highest mortality rate, is time-consuming and heavily reliant on the experience of a pathologist. In this report, we proposed a deep learning model that required minimal noise reduction or manual annotation by an experienced pathologist for HCC diagnosis and classification.MethodsWe collected a whole-slide image of hematoxylin and eosin-stained pathological slides from 592 HCC patients at the First Affiliated Hospital, College of Medicine, Zhejiang University between 2015 and 2020. We propose a noise-specific deep learning model. The model was trained initially with 137 cases cropped into multiple-scaled datasets. Patch screening and dynamic label smoothing strategies are adopted to handle the histopathological liver image with noise annotation from the perspective of input and output. The model was then tested in an independent cohort of 455 cases with comparable tumor types and differentiations.ResultsExhaustive experiments demonstrated that our two-step method achieved 87.81% pixel-level accuracy and 98.77% slide-level accuracy in the test dataset. Furthermore, the generalization performance of our model was also verified using The Cancer Genome Atlas dataset, which contains 157 HCC pathological slides, and achieved an accuracy of 87.90%.ConclusionsThe noise-specific histopathological classification model of HCC based on deep learning is effective for the dataset with noisy annotation, and it significantly improved the pixel-level accuracy of the regular convolutional neural network (CNN) model. Moreover, the model also has an advantage in detecting well-differentiated HCC and microvascular invasion.Shi FengXiaotian YuWenjie LiangXuejie LiWeixiang ZhongWanwan HuHan ZhangZunlei FengMingli SongJing ZhangXiuming ZhangFrontiers Media S.A.articleHCC classificationpathological imagesdeep learningwhole-slide imagenoisy annotationNeoplasms. Tumors. Oncology. Including cancer and carcinogensRC254-282ENFrontiers in Oncology, Vol 11 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	HCC classification pathological images deep learning whole-slide image noisy annotation Neoplasms. Tumors. Oncology. Including cancer and carcinogens RC254-282
spellingShingle	HCC classification pathological images deep learning whole-slide image noisy annotation Neoplasms. Tumors. Oncology. Including cancer and carcinogens RC254-282 Shi Feng Xiaotian Yu Wenjie Liang Xuejie Li Weixiang Zhong Wanwan Hu Han Zhang Zunlei Feng Mingli Song Jing Zhang Xiuming Zhang Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma
description	BackgroundAn accurate pathological diagnosis of hepatocellular carcinoma (HCC), one of the malignant tumors with the highest mortality rate, is time-consuming and heavily reliant on the experience of a pathologist. In this report, we proposed a deep learning model that required minimal noise reduction or manual annotation by an experienced pathologist for HCC diagnosis and classification.MethodsWe collected a whole-slide image of hematoxylin and eosin-stained pathological slides from 592 HCC patients at the First Affiliated Hospital, College of Medicine, Zhejiang University between 2015 and 2020. We propose a noise-specific deep learning model. The model was trained initially with 137 cases cropped into multiple-scaled datasets. Patch screening and dynamic label smoothing strategies are adopted to handle the histopathological liver image with noise annotation from the perspective of input and output. The model was then tested in an independent cohort of 455 cases with comparable tumor types and differentiations.ResultsExhaustive experiments demonstrated that our two-step method achieved 87.81% pixel-level accuracy and 98.77% slide-level accuracy in the test dataset. Furthermore, the generalization performance of our model was also verified using The Cancer Genome Atlas dataset, which contains 157 HCC pathological slides, and achieved an accuracy of 87.90%.ConclusionsThe noise-specific histopathological classification model of HCC based on deep learning is effective for the dataset with noisy annotation, and it significantly improved the pixel-level accuracy of the regular convolutional neural network (CNN) model. Moreover, the model also has an advantage in detecting well-differentiated HCC and microvascular invasion.
format	article
author	Shi Feng Xiaotian Yu Wenjie Liang Xuejie Li Weixiang Zhong Wanwan Hu Han Zhang Zunlei Feng Mingli Song Jing Zhang Xiuming Zhang
author_facet	Shi Feng Xiaotian Yu Wenjie Liang Xuejie Li Weixiang Zhong Wanwan Hu Han Zhang Zunlei Feng Mingli Song Jing Zhang Xiuming Zhang
author_sort	Shi Feng
title	Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma
title_short	Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma
title_full	Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma
title_fullStr	Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma
title_full_unstemmed	Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma
title_sort	development of a deep learning model to assist with diagnosis of hepatocellular carcinoma
publisher	Frontiers Media S.A.
publishDate	2021
url	https://doaj.org/article/ad8d7d83bec544e19696d05b87852c76
work_keys_str_mv	AT shifeng developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT xiaotianyu developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT wenjieliang developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT xuejieli developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT weixiangzhong developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT wanwanhu developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT hanzhang developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT zunleifeng developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT minglisong developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT jingzhang developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma AT xiumingzhang developmentofadeeplearningmodeltoassistwithdiagnosisofhepatocellularcarcinoma
_version_	1718404590780547072

Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma

Ejemplares similares