Deep learning of circulating tumour cells

Zeune, Leonie L.; Boink, Yoeri E.; Dalum, Guus van; Nanou, Afroditi; Wit, Sanne de; Andree, Kiki; Swennenhuis, Joost F.; Gils, Stephan A. van; Terstappen, Leonardus Wendelinus Mathias Marie; Brüne, Christoph

doi:10.1038/s42256-020-0153-x

Cited by 65 publications

(53 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The F-score provides a measure of the overall performance of the model by considering the equal importance of precision and recall. As a comparison, in a recent study 47 , deep learning networks have shown the ability to unlock the hidden information in fluorescent images. The networks could classify fluorescent images of single cells including CTCs with a very high accuracy (96%).…”

Section: Resultsmentioning

confidence: 99%

Label-free detection of rare circulating tumor cells by image analysis and machine learning

Shen

Zhou

Qin

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Detection and characterization of rare circulating tumor cells (ctcs) in patients' blood is important for the diagnosis and monitoring of cancer. The traditional way of counting CTCs via fluorescent images requires a series of tedious experimental procedures and often impacts the viability of cells. Here we present a method for label-free detection of ctcs from patient blood samples, by taking advantage of data analysis of bright field microscopy images. The approach uses the convolutional neural network, a powerful image classification and machine learning algorithm to perform label-free classification of cells detected in microscopic images of patient blood samples containing white blood cells and ctcs. it requires minimal data pre-processing and has an easy experimental setup. through our experiments, we show that our method can achieve high accuracy on the identification of rare CTCs without the need for advanced devices or expert users, thus providing a faster and simpler way for counting and identifying ctcs. With more data becoming available in the future, the machine learning model can be further improved and can serve as an accurate and easy-to-use tool for ctc analysis. Circulating tumor cells (CTCs) found in peripheral blood are originated from solid tumors. They are cells shed by a primary tumor into the vasculature, circulating through bloodstream of cancer patients, and colonizing at distant sites which may form metastatic tumors 1. CTCs are an important biomarker for early tumor diagnosis and early evaluation of disease recurrence and metastatic spread in various types of cancer 2-6. Early detection of CTCs provides high chances for patients to survive before severe cancer growth occurs 7. The CTC count is also an important prognostic factor for patients with metastatic cancer 8-12. For example, a study has shown that the number of CTCs is an independent predictor of survival in patients for breast cancer and prostate cancer 8-10 , and the changes of the CTC count predict the survival in patients for lung cancer 12. However, the identification of the CTCs population is a challenging problem. Various approaches to identifying and isolating CTCs including antibody-based methods and physical-characteristics-based methods have been developed 13-19. This task is difficult because of the low concentration of CTCs existing in a patient's peripheral blood-a few CTCs out of 10 billion blood cells 20,21 , as well as heterogeneity in the characteristics of CTCs 22,23. For example, the mechanism of CTCs maintaining metastatic potential during circulating is not well understood 24 ; CTCs derived from some patients allow a cell line to be established, but CTCs from some others lose the capability of proliferation after a few hours of blood drawing 13. Therefore, the incapability to draw a large volume of blood from patients leads to the need for improvements of CTC isolation methods so that CTCs can be detected in small sample volumes. Further, the inconsistency in the viability of CTCs hinders further explorati...

show abstract

Section: Resultsmentioning

confidence: 99%

Label-free detection of rare circulating tumor cells by image analysis and machine learning

Shen

Zhou

Qin

et al. 2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…Therefore, for the current analysis we restricted that part of our analysis to only 10% of the nucleated events which were randomly selected. Experience obtained with the deep learning based classification algorithm we developed showed very promising results and the new segmentation method also showed that it is often superior to the more traditional segmentation approach currently used in ACCEPT [ 22 ]. Thus, in the future, we would like to incorporate a semantic deep learning segmentation into the ACCEPT toolbox to automatically detect and classify CTC, tumor derived extra cellular vesicles, leukocytes, and other cell populations for which reagents are added.…”

Section: Discussionmentioning

confidence: 99%

Classification of Cells in CTC-Enriched Samples by Advanced Image Analysis

Wit

Zeune

Hiltermann

et al. 2018

Cancers

Self Cite

View full text Add to dashboard Cite

In the CellSearch® system, blood is immunomagnetically enriched for epithelial cell adhesion molecule (EpCAM) expression and cells are stained with the nucleic acid dye 4′6-diamidino-2-phenylindole (DAPI), Cytokeratin-PE (CK), and CD45-APC. Only DAPI+/CK+ objects are presented to the operator to identify circulating tumor cells (CTC) and the identity of all other cells and potential undetected CTC remains unrevealed. Here, we used the open source imaging program Automatic CTC Classification, Enumeration and PhenoTyping (ACCEPT) to analyze all DAPI+ nuclei in EpCAM-enriched blood samples obtained from 192 metastatic non-small cell lung cancer (NSCLC) patients and 162 controls. Significantly larger numbers of nuclei were detected in 300 patient samples with an average and standard deviation of 73,570 ± 74,948, as compared to 359 control samples with an average and standard deviation of 4191 ± 4463 (p < 0.001). In patients, only 18% ± 21% and in controls 23% ± 15% of the nuclei were identified as leukocytes or CTC. Adding CD16-PerCP for granulocyte staining, the use of an LED as the light source for CD45-APC excitation and plasma membrane staining obtained with wheat germ agglutinin significantly improved the classification of EpCAM-enriched cells, resulting in the identification of 94% ± 5% of the cells. However, especially in patients, the origin of the unidentified cells remains unknown. Further studies are needed to determine if undetected EpCAM+/DAPI+/CK-/CD45- CTC is present among these cells.

show abstract

“…The integration of label‐free imaging, AI, and microfluidics is a subject of great interest in the scientific community to solve open biomedical questions 21,28‐31 . In particular, the authors of the present review strongly believe that such integration would represent a keystone for the identification of CTCs (Figure 1).…”

Section: Introductionmentioning

confidence: 87%

“…The latter refers to neural networks, that is, architectures that can operate directly on the input image (rather than its descriptors) and can learn, from a wide dataset, its most appropriate representation with several levels of abstraction 151‐154 . Deep learning approaches to image segmentation and classification have been demonstrated to be robust with a huge generalization power 15,19,20,28,29,154‐168 . Cell segmentation in microfluidic streams is obtainable using pretrained networks (eg, Mask R‐CNN and Faster R‐CNN) 151,152,167 .…”

Section: Deep Learning‐assisted Imaging For Cell Identificationmentioning

confidence: 99%

“…AggNet first introduced data aggregation to learn image annotations from no experts to refine the training process 164 . Recently, an auto‐encoder feed by fluorescent images of blood samples enriched for CTCs has been proved to be 96% accurate in their identification in blood streams 28 . However, labeling the sample to obtain a contrast agent makes the dataset strictly dependent on the preparation protocol and the specific dye/marker employed.…”

Section: Deep Learning‐assisted Imaging For Cell Identificationmentioning

confidence: 99%

See 1 more Smart Citation

Perspectives on liquid biopsy for label‐free detection of “circulating tumor cells” through intelligent lab‐on‐chips

Miccio

Cimmino

Kurelac

et al. 2020

VIEW

View full text Add to dashboard Cite

Circulating tumor cells (CTCs) are rare tumor cells released from primary, metastatic, or recurrent tumors in the peripheral blood of cancer patients. CTCs isolation from peripheral blood and their molecular characterization represent a new marker in cancer screening, a diagnostic tool called “liquid biopsy” (LB). Compared to traditional tissue biopsy that is invasive and does not reveal tumor heterogeneity, LB is noninvasive and reflects in “real‐time” tumor dynamism and drug sensitivity. In the frame of LB, a new paradigm based on single‐cell and label‐free analysis based on morphological analysis is emerging. Here, we review the latest research developments in this emerging vision of LB. In particular, we survey and discuss recent improvements in microfluidics, imaging label‐free diagnosis and cell classification by artificial intelligence and how to combine them to realize an intelligent platform based on lab‐on‐chip technology. This prospect appears to open up promising and intriguing new scenarios for cancer management through single‐cell analysis that will revolutionize the future of early cancer diagnosis and therapeutic choice with disruptive impact on the society.

show abstract

Deep learning of circulating tumour cells

Abstract: Composition of training and validation sets. The original dataset used to train and validate our networks was obtained through the automated processing of 499 patient samples with ACCEPT and

Cited by 65 publications

References 44 publications

Label-free detection of rare circulating tumor cells by image analysis and machine learning

Label-free detection of rare circulating tumor cells by image analysis and machine learning

Classification of Cells in CTC-Enriched Samples by Advanced Image Analysis

Perspectives on liquid biopsy for label‐free detection of “circulating tumor cells” through intelligent lab‐on‐chips

Contact Info

Product

Resources

About