Feasibility of fully automated classification of whole slide images based on deep learning

Cho, Kyung-Ok; Lee, Sung Hak; Jang, Hyun‐Jong

doi:10.4196/kjpp.2020.24.1.89

Cited by 18 publications

(25 citation statements)

References 29 publications

(41 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…16 In our previous study for a normal/tumor classifier in stomach cancer, we tried to solve the class imbalance problem with a modified loss function. 17 However, we concluded that it is not easy to obtain a balanced performance with hugely imbalanced training data for DL-based tissue classifiers. Thus, we decided to limit the number of MSS patients to 1.5 times that of the MSI-H patients and randomly selected 126 MSS patients for our study (Table S1).…”

Section: Patient Cohortmentioning

confidence: 88%

Feasibility of deep learning‐based fully automated classification of microsatellite instability in tissue slides of colorectal cancer

Lee

Song

Jang

2021

Intl Journal of Cancer

Self Cite

View full text Add to dashboard Cite

High levels of microsatellite instability (MSI-H) occurs in about 15% of sporadic colorectal cancer (CRC) and is an important predictive marker for response to immune checkpoint inhibitors. To test the feasibility of a deep learning (DL)-based classifier as a screening tool for MSI status, we built a fully automated DL-based MSI classifier using pathology whole-slide images (WSIs) of CRCs. On small image patches of The Cancer Genome Atlas (TCGA) CRC WSI dataset, tissue/non-tissue, normal/ tumor and MSS/MSI-H classifiers were applied sequentially for the fully automated prediction of the MSI status. The classifiers were also tested on an independent cohort. Furthermore, to test how the expansion of the training data affects the performance of the DL-based classifier, additional classifier trained on both TCGA and external datasets was tested. The areas under the receiver operating characteristic curves were 0.892 and 0.972 for the TCGA and external datasets, respectively, by a classifier trained on both datasets. The performance of the DL-based classifier was much better than that of previously reported histomorphology-based methods. We speculated that about 40% of CRC slides could be screened for MSI status without molecular testing by the DL-based classifier. These results demonstrated that the DL-based method has potential as a screening tool to discriminate molecular alteration in tissue slides.

show abstract

Section: Patient Cohortmentioning

confidence: 88%

Feasibility of deep learning‐based fully automated classification of microsatellite instability in tissue slides of colorectal cancer

Lee

Song

Jang

2021

Intl Journal of Cancer

Self Cite

View full text Add to dashboard Cite

show abstract

“…Deep learning did not perform optimally when there was a huge imbalance between classes[ 19 ]. In a previous study, we failed to obtain the balanced performance in tissue classification tasks unless the dataset itself was forced to have similar numbers between the classes[ 20 ]. Thus, we limited the difference in patient numbers between the mutation group and wild-type group by less than 1.4 fold through a random sampling.…”

Section: Methodsmentioning

confidence: 99%

Prediction of clinically actionable genetic alterations from colorectal cancer histopathology images using deep learning

Jang

Lee

Kang

et al. 2020

WJG

Self Cite

View full text Add to dashboard Cite

BACKGROUND Identifying genetic mutations in cancer patients have been increasingly important because distinctive mutational patterns can be very informative to determine the optimal therapeutic strategy. Recent studies have shown that deep learning-based molecular cancer subtyping can be performed directly from the standard hematoxylin and eosin (H&E) sections in diverse tumors including colorectal cancers (CRCs). Since H&E-stained tissue slides are ubiquitously available, mutation prediction with the pathology images from cancers can be a time- and cost-effective complementary method for personalized treatment. AIM To predict the frequently occurring actionable mutations from the H&E-stained CRC whole-slide images (WSIs) with deep learning-based classifiers. METHODS A total of 629 CRC patients from The Cancer Genome Atlas (TCGA-COAD and TCGA-READ) and 142 CRC patients from Seoul St. Mary Hospital (SMH) were included. Based on the mutation frequency in TCGA and SMH datasets, we chose APC , KRAS , PIK3CA , SMAD4 , and TP53 genes for the study. The classifiers were trained with 360 × 360 pixel patches of tissue images. The receiver operating characteristic (ROC) curves and area under the curves (AUCs) for all the classifiers were presented. RESULTS The AUCs for ROC curves ranged from 0.693 to 0.809 for the TCGA frozen WSIs and from 0.645 to 0.783 for the TCGA formalin-fixed paraffin-embedded WSIs. The prediction performance can be enhanced with the expansion of datasets. When the classifiers were trained with both TCGA and SMH data, the prediction performance was improved. CONCLUSION APC , KRAS , PIK3CA , SMAD4 , and TP53 mutations can be predicted from H&E pathology images using deep learning-based classifiers, demonstrating the potential for deep learning-based mutation prediction in the CRC tissue slides.

show abstract

“…Unlike the tissue/non-tissue classifier, the Inception-v3 model was used without any parameter changes to train the normal/tumor classifiers for the 360 × 360-pixel patches at 20× magnification. We adopted the inception-v3 model because it was superior for the normal/tumor discrimination task than other CNN architectures in our previous study [13]. Deep neural networks were implemented using the TensorFlow deep learning library (http://tensorflow.org).…”

Section: Normal/tumor Classifiers For the Tcga Tissue Datasetsmentioning

confidence: 99%

“…In contrast to other classic machine learning approaches, deep learning can learn relevant features for given tasks directly from raw input datasets, and thus eliminate the necessity of domain-specific feature extraction processes [10]. Combined with huge digital WSI datasets, deep learning has been rapidly adopted for the pathologic diagnosis tasks such as Gleason grading of prostate cancer [4], detection of invasive ductal carcinoma in breast cancer [11], detection of metastasis for breast cancer [12], and detection of tumor tissues in gastric cancer [13].…”

Section: Introductionmentioning

confidence: 99%

Generalizability of Deep Learning System for the Pathologic Diagnosis of Various Cancers

2021

Self Cite

View full text Add to dashboard Cite

The deep learning (DL)-based approaches in tumor pathology help to overcome the limitations of subjective visual examination from pathologists and improve diagnostic accuracy and objectivity. However, it is unclear how a DL system trained to discriminate normal/tumor tissues in a specific cancer could perform on other tumor types. Herein, we cross-validated the DL-based normal/tumor classifiers separately trained on the tissue slides of cancers from bladder, lung, colon and rectum, stomach, bile duct, and liver. Furthermore, we compared the differences between the classifiers trained on the frozen or formalin-fixed paraffin-embedded (FFPE) tissues. The Area under the curve (AUC) for the receiver operating characteristic (ROC) curve ranged from 0.982 to 0.999 when the tissues were analyzed by the classifiers trained on the same tissue preparation modalities and cancer types. However, the AUCs could drop to 0.476 and 0.439 when the classifiers trained for different tissue modalities and cancer types were applied. Overall, the optimal performance could be achieved only when the tissue slides were analyzed by the classifiers trained on the same preparation modalities and cancer types.

show abstract

Feasibility of fully automated classification of whole slide images based on deep learning

Cited by 18 publications

References 29 publications

Feasibility of deep learning‐based fully automated classification of microsatellite instability in tissue slides of colorectal cancer

Feasibility of deep learning‐based fully automated classification of microsatellite instability in tissue slides of colorectal cancer

Prediction of clinically actionable genetic alterations from colorectal cancer histopathology images using deep learning

Generalizability of Deep Learning System for the Pathologic Diagnosis of Various Cancers

Contact Info

Product

Resources

About