Computer aided quantification of intratumoral stroma yields an independent prognosticator in rectal cancer

Geessink, Oscar; Baidoshvili, Alexi; Klaase, Joost M.; Bejnordi, Babak Ehteshami; Litjens, Geert; Pelt, Gabi W. van; Mesker, Wilma E.; Nagtegaal, Irıs D.; Ciompi, Francesco; Laak, Jeroen van der

doi:10.1007/s13402-019-00429-z

Cited by 88 publications

(91 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent advances in machine learning have resulted in computer algorithms that are capable of analysing entirely digitized microscopic tissue sections (whole slide images; WSI). It has been shown that such algorithms can, for instance, accurately detect and delineate tumour areas in breast and colon tissue sections and detect mitotic figures in breast cancer [1][2][3]. Next to direct use in research and clinical practice, such algorithms are also of interest to reassess the diagnostic/prognostic value of widely used morphological criteria.…”

Section: Introductionmentioning

confidence: 99%

Deep learning and manual assessment show that the absolute mitotic count does not contain prognostic information in triple negative breast cancer

et al. 2019

Self Cite

View full text Add to dashboard Cite

Purpose The prognostic value of mitotic count for invasive breast cancer is firmly established. As yet, however, limited studies have been aimed at assessing mitotic counts as a prognostic factor for triple negative breast cancers (TNBC). Here, we assessed the prognostic value of absolute mitotic counts for TNBC, using both deep learning and manual procedures. Methods A retrospective TNBC cohort (n = 298) was used. The absolute manual mitotic count was assessed by averaging counts from three independent observers. Deep learning was performed using a convolutional neural network on digitized H&E slides. Multivariable Cox regression models for relapse-free survival and overall survival served as baseline models. These were expanded with dichotomized mitotic counts, attempting every possible cut-off value, and evaluated by means of the c-statistic. Results We found that per 2 mm 2 averaged manual mitotic counts ranged from 1 to 187 (mean 37.6, SD 23.4), whereas automatic counts ranged from 1 to 269 (mean 57.6; SD 42.2). None of the cut-off values improved the models' baseline c-statistic, for both manual and automatic assessments. Conclusions Based on our results we conclude that the level of proliferation, as reflected by mitotic count, does not serve as a prognostic factor for TNBC. Therefore, TNBC patient management based on mitotic count should be discouraged.

show abstract

Section: Introductionmentioning

confidence: 99%

Deep learning and manual assessment show that the absolute mitotic count does not contain prognostic information in triple negative breast cancer

et al. 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…Inspired by previous studies showing that tumor-stroma ratio (TSR) serves as an independent prognostic factor in many types of cancer [7,9,10], we sought to examine deep learning methods in the segmentation of tumor images to differentiate the stroma and non-stroma areas as starting points to potentially initialize more deep downstream analysis relevant to researching the role of TSR in cancer prognosis. As a pilot experiment, we chose to use the U-Net method, which is a convolutional neural network that was developed for biomedical image segmentation [11] and has been a conventional method in the field since its development (reviewed in [35,36]).…”

Section: Segmentation Of Cancer Tissue Images Can Distinguish Betweenmentioning

confidence: 99%

“…With recent advances in deep learning [2], a subfield of artificial intelligence, the remarkable successes in the application of deep learning methods in tumor image segmentation [3] and tumor classification [4] provided promise to state-of-art computer-aided medical image analysis. The ability of deep learning methods to automatically and quantitatively extract image features from digital pathological images and computationally learn the multilevel representations of data for comprehensive image analysis and classification offers the opportunity for better modeling of disease and potentially improved prediction of disease and patient outcome [5,6,7]. As a result, these deep learning methods may help assist and validate diagnosis results from radiologists and pathologists.…”

Section: Introductionmentioning

confidence: 99%

“…Therefore, by identifying the stroma present in cancerous tissue, scientists would be able to learn more about the progression of cancer and start early treatments targeting the tumor stroma. Interestingly, recent studies show that the tumorstroma ratio (TSR) derived from image analysis also potentially serves as an independent prognostic factor in many types of cancer [7,9,10]. However, much of such work was done manually.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A pilot exploratory study of the potentials of deep learning methods in cancer image segmentation and classification

Liu

2019

Preprint

View full text Add to dashboard Cite

Background: Tumor classification and feature quantification from H&E histology images are critical tasks for cancer diagnosis, cancer research, and treatment. However, both tasks involve tedious and time-consuming manual examination of histology images. We explored the possibilities of using deep learning methods to perform segmentation and classification of histology images of cancer tissue for their potential in computeraided tumor diagnosis and other clinical and research applications. Specifically, we tested selected deep learning methods for their performance in the segmentation of stroma and glandular objects in tumor image data and in the classification of tumor images. We automated these tasks to help facilitate downstream tumor image analysis, reduce the labor load of pathologists, and provide them with a second opinion on their analysis. Methods: We modified a patch-based U-Net model and trained it to perform stroma detection and segmentation in cancer tissue. Then the semantic segmentation capabilities of the U-Net model were compared with that of a DeepLabV3+ model. We also explored the possible use of transfer learning to train a patch-based model to classify cancer tissue images as carcinoma and sarcoma and to further classify them as carcinoma subtypes. Results:In spite of the limited dataset available for the pilot study, we found that the unconventional DeepLabV3+ model performed biomedical image segmentation more effectively than U-Net when k-fold cross-validation was utilized, but U-Net still showed promise as an effective and efficient model when we used a customized validation approach. We believe that the DeepLabV3+ model can perform segmentation with even more accuracy if computation resource constraints are removed or if more data is used to augment the result. In terms of tumor classification, our selected models also consistently achieve test accuracies above 80%, with a model trained using transfer learning with VGG-16 network as the feature extractors, or convolutional base performing best. For multi-class tumor subtype classification, we also observed promising test accuracies from our models, and a customized post-processing method provided even higher prediction accuracy on test set images and this method can be further investigated. Conclusions:This pilot exploratory study provided strong evidence for the powerful potentials of deep learning models for segmentation and classification of tumor image data.

show abstract

“…Therefore, by identifying the stroma present in cancerous tissue, scientists would be able to learn more about the progression of cancer and start early treatments targeting the tumor stroma. Interestingly, recent studies show that the tumor-stroma ratio (TSR) derived from image analysis also potentially serves as an independent prognostic factor in many types of cancer [7,9,10]. However, much of such work was done manually.…”

Section: Introductionmentioning

confidence: 99%

A pilot exploratory study of the potentials of deep learning methods in cancer image segmentation and classification

Liu²

2020

Preprint

View full text Add to dashboard Cite

Background Tumor classification and feature quantification from H&E histology images are critical tasks for cancer diagnosis, cancer research, and treatment. However, both tasks involve tedious and time-consuming manual examination of histology images. We explored the usage of deep learning methods in segmentation and classification of histology images of cancer tissue for their potential in computer-aided tumor diagnosis and other clinical and research applications. Specifically, we evaluated performance of selected deep learning methods in stroma and glandular objects segmentation in tumor image data and tumor images classification. We automated these tasks to help facilitate downstream tumor image analysis, reduce the labor load of pathologists, and provide them with a second opinion on their analysis. Methods We modified a patch-based U-Net model and trained it to perform stroma detection and segmentation in cancer tissue. Then the semantic segmentation capabilities of the U-Net model were compared with that of a DeepLabV3+ model. We explored the possible use of transfer learning to train a patch-based model to classify cancer tissue images as carcinoma and sarcoma and to further classify them as carcinoma subtypes. Results In spite of the limited dataset available for the pilot study, we found that the DeepLabV3+ model performed biomedical image segmentation more effectively than U-Net when k-fold cross-validation was utilized, but U-Net still showed promise as an effective and efficient model when we used a customized validation approach. We believe that the DeepLabV3+ model can perform segmentation with even more accuracy if computation resource constraints are removed or if more data is used to augment the result. In terms of tumor classification, our selected models also consistently achieve test accuracies above 80%, with a model trained using transfer learning with VGG-16 network as the feature extractors, or convolutional base performing best. For multi-class tumor subtype classification, we also observed promising test accuracies from our models, and a customized post-processing method provided even higher prediction accuracy on test set images and this method can be further investigated. Conclusions This pilot exploratory study provided strong evidence for the powerful potentials of deep learning models for segmentation and classification of tumor image data.

show abstract

Computer aided quantification of intratumoral stroma yields an independent prognosticator in rectal cancer

Cited by 88 publications

References 26 publications

Deep learning and manual assessment show that the absolute mitotic count does not contain prognostic information in triple negative breast cancer

Deep learning and manual assessment show that the absolute mitotic count does not contain prognostic information in triple negative breast cancer

A pilot exploratory study of the potentials of deep learning methods in cancer image segmentation and classification

A pilot exploratory study of the potentials of deep learning methods in cancer image segmentation and classification

Contact Info

Product

Resources

About