Deep learning as a tool for increased accuracy and efficiency of histopathological diagnosis

By Geert J. S. Litjens, C. Sánchez, N. Timofeeva, M. Hermsen, I. Nagtegaal, Iringo Kovacs, Christina A. Hulsbergen - van de Kaa, P. Bult, B. van Ginneken, J. A. van der Laak. 2016

1,025 citations recorded by the source corpus.

Pathologists face a substantial increase in workload and complexity of histopathologic cancer diagnosis due to the advent of personalized medicine. Therefore, diagnostic protocols have to focus equally on efficiency and accuracy. In this paper we introduce ‘deep learning’ as a technique to improve the objectivity and efficiency of histopathologic slide analysis. Through two examples, prostate cancer identification in biopsy specimens and breast cancer metastasis detection in sentinel lymph nodes, we show the potential of this new methodology to reduce the workload for pathologists, while at the same time increasing objectivity of diagnoses. We found that all slides containing prostate cancer and micro- and macro-metastases of breast cancer could be identified automatically while 30–40% of the slides containing benign and normal tissue could be excluded without the use of any additional immunohistochemical markers or human intervention. We conclude that ‘deep learning’ holds great promise to improve the efficacy of prostate cancer diagnosis and breast cancer staging.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Epidemiology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Epidemiology

Citations in the corpus

Improving the accuracy of medical diagnosis with causal machine learning (2020, methodology): “…and using Bayes rule to estimate the posterior, P ð D jE ; θ Þ ¼ P ðEj D ; θ Þ P ð D ; θ Þ P ðE ; θ Þ : Examples of discriminative diagnostic models include neural network and deep learning models 8,10,15,33,34 , whereas generative models are typically Bayesian networks 18,19,21,22,27,35,36 .”
Opportunities and obstacles for deep learning in biology and medicine (2017, background): “Systems to aid in the analysis of histology slides are also promising use cases for deep learning [65].”
Ensemble Bayesian Decision Making with Redundant Deep Perceptual Control Policies (2018, background): “Neural Networks (NNs) are currently one of the most powerful tools for solving difficult problems in decision making such as playing Go [1] or medical diagnosis [2], [3].”
Detecting Cancer Metastases on Gigapixel Pathology Images (2017, methodology): “The Camelyon organizers also trained CNNs on smaller datasets to detect breast cancer in lymph nodes and prostate cancer biopsies [16].”
Leveraging uncertainty information from deep neural networks for disease detection (2016, neutral): “σpred ≈ σ̂pred = 1 T −1 √ T ∑ t=1 (p(y∗|x∗,θ(ω̂t))− μ̂pred)(2) (8)”
Deep Learning-Based Histopathologic Assessment of Kidney Tissue. (2019, background): “10 13 Deep learning techniques allow autonomous learning of increasingly complex structures during the transformation from input (WSI) to desired output ( e.g. , structure detection).”
The impact of pre- and post-image processing techniques on deep learning frameworks: A comprehensive review for digital pathology image analysis (2020, methodology): “Some authors have employed a threshold strategy on the optical density (OD) image and HSV color space [4,40].”
Deep neural network models for computational histopathology: A survey (2019, methodology): “Subsequently, a few authors (Litjens et al., 2016; Vandenberghe et al., 2017) employed a simpler patch-based CNN model for the identification of breast and prostate cancer in WSI, achieving an AUC of 0.”
Visualizing Convolutional Neural Networks to Improve Decision Support for Skin Lesion Classification (2018, methodology): “In medicine, deep learning is used as a tool to assist professionals of various subﬁelds in their diagnoses, such as histopathology [11], oncology [1,4,17], pulmonology [7,15], etc 1 .”
Interpretable classification of Alzheimer’s disease pathologies with a convolutional neural network pipeline (2018, methodology): “We hypothesized deep learning methods could augment neuropathological whole slide image (WSI) analysis 35 .”
Artificial Intelligence-Based Breast Cancer Nodal Metastasis Detection: Insights Into the Black Box for Pathologists. (2018, methodology): “This patch size was selected to encompass several cells and was also used by Litjens et al. 11 The algorithm takes as input a larger square patch of size 299 pixels ( » 75 l m) to provide additional context akin to how a pathologist reviews a slide.”
Weakly Supervised Deep Learning for Whole Slide Lung Cancer Image Analysis (2020, methodology): “The normalized cumulative histogram with percentile analysis [15] and connected component features of WSI probability maps [15], [17], [18] were widely utilized in the second-stage decision fusion model.”
Deep Learning for Brain MRI Segmentation: State of the Art and Future Directions (2017, background): “Deep learning techniques are gaining popularity in many areas of medical image analysis [2], such as computer-aided detection of breast lesions [3], computer-aided diagnosis of breast lesions and pulmonary nodules [4], and in histopathological diagnosis [5].”
Deep learning in histopathology: the path to the clinic (2021, methodology): “Methods that used entire WSIs followed for applications such as breast cancer segmentation 17 , glioma classification 18 , non-alcoholic fatty liver disease 19 , assessment of renal transplant biopsies 20 and prostate cancer detection 21 .”
Self-Recalibrating Surface EMG Pattern Recognition for Neuroprosthesis Control Based on Convolutional Neural Network (2017, background): “, 2015), histopathological diagnosis (Litjens et al., 2016), and protein structure prediction (Wang et al.”