Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fully Automated CTC Detection, Segmentation and Classification for Multi-Channel IF Imaging

Evan Schwab, Bharat Annaldas, Nisha Ramesh, Anna Lundberg, Vishal Shelke, Xinran Xu, Cole Gilbertson, Jiyun Byun, Ernest T. Lam

TL;DR

This work has developed a fully automated machine learning-based production-level pipeline to efficiently detect, segment and classify CTCs in multi-channel IF images, achieving over 99% sensitivity and 97% specificity on 9,533 cells from 15 mBCa patients.

Abstract

Liquid biopsies (eg., blood draws) offer a less invasive and non-localized alternative to tissue biopsies for monitoring the progression of metastatic breast cancer (mBCa). Immunofluoresence (IF) microscopy is a tool to image and analyze millions of blood cells in a patient sample. By detecting and genetically sequencing circulating tumor cells (CTCs) in the blood, personalized treatment plans are achievable for various cancer subtypes. However, CTCs are rare (about 1 in 2M), making manual CTC detection very difficult. In addition, clinicians rely on quantitative cellular biomarkers to manually classify CTCs. This requires prior tasks of cell detection, segmentation and feature extraction. To assist clinicians, we have developed a fully automated machine learning-based production-level pipeline to efficiently detect, segment and classify CTCs in multi-channel IF images. We achieve over 99% sensitivity and 97% specificity on 9,533 cells from 15 mBCa patients. Our pipeline has been successfully deployed on real mBCa patients, reducing a patient average of 14M detected cells to only 335 CTC candidates for manual review.

Fully Automated CTC Detection, Segmentation and Classification for Multi-Channel IF Imaging

TL;DR

This work has developed a fully automated machine learning-based production-level pipeline to efficiently detect, segment and classify CTCs in multi-channel IF images, achieving over 99% sensitivity and 97% specificity on 9,533 cells from 15 mBCa patients.

Abstract

Liquid biopsies (eg., blood draws) offer a less invasive and non-localized alternative to tissue biopsies for monitoring the progression of metastatic breast cancer (mBCa). Immunofluoresence (IF) microscopy is a tool to image and analyze millions of blood cells in a patient sample. By detecting and genetically sequencing circulating tumor cells (CTCs) in the blood, personalized treatment plans are achievable for various cancer subtypes. However, CTCs are rare (about 1 in 2M), making manual CTC detection very difficult. In addition, clinicians rely on quantitative cellular biomarkers to manually classify CTCs. This requires prior tasks of cell detection, segmentation and feature extraction. To assist clinicians, we have developed a fully automated machine learning-based production-level pipeline to efficiently detect, segment and classify CTCs in multi-channel IF images. We achieve over 99% sensitivity and 97% specificity on 9,533 cells from 15 mBCa patients. Our pipeline has been successfully deployed on real mBCa patients, reducing a patient average of 14M detected cells to only 335 CTC candidates for manual review.
Paper Structure (14 sections, 4 figures, 3 tables)

This paper contains 14 sections, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Background
  3. Methods
  4. Cell Detection
  5. Nuclear Segmentation
  6. Cell Segmentation
  7. Feature Extraction
  8. Experiments
  9. Ground Truth Data
  10. CTC Classifier Training
  11. Results
  12. Conclusion
  13. Acknowledgments.
  14. Disclosure of Interests.

Figures (4)

  • Figure 1: Example user interface of ML classified CTC candidates output by our pipeline and presented to clinicians (bottom row thumbnails). One thumbnail is selected for review and cell and nuclear MFI values reported per channel. The yellow nuclear mask is overlayed on each channel and the color composite.
  • Figure 2: Example CTCs, non-CTCs, and artefacts in 3-channel IF images with a color composite. CTCs have prominent CK signal and low CD45/31. Some Non-CTCs are visually similar which makes manual classification difficult.
  • Figure 3: BRIA Pipeline Overview. After IF image acquisition and slide processing, the main algorithmic steps include Cell Detection, Nuclear and Cell Segmentation, Feature Extraction, and CTC Classification.
  • Figure 4: Top ten features weights for True Positive (CTC) and True Negative (non-CTC/artefact) classes using SHapley Additive exPlanations (SHAP) SHAP averaged over 100 random samples per class in the verification set. (Note: Nuc_IQR_ck is the inter-quartile range (IQR) of pixels in the nucleus (nuc) in CK; STD_ck is the standard deviation of CK; Cell_Coloc_ck_dapi is the co-localization of pixels between CK and DAPI within the cell. See Supp. Table 1 for feature definitions.)