Related Experiment Video
Updated: May 30, 2026
07:51
The Application of 1% Methylene Blue Dye As a Single Technique in Breast Cancer Sentinel Node Biopsy
Published on: June 1, 2019
Development and evaluation of a robust algorithm for computer-assisted detection of sentinel lymph node
Gina M Clarke1, Chris Peressotti, Claire M B Holloway
1Imaging Research, Sunnybrook Health Sciences Centre, Department of Surgical Oncology, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada. gina.clarke@sunnybrook.ca
Histopathology
|July 21, 2011
Summary
A new computer-assisted detection (CAD) algorithm enhances breast sentinel lymph node evaluation. This tool accurately identifies micrometastases, improving detection of occult cancer spread.
Area of Science:
- Digital Pathology
- Oncology
- Medical Imaging Analysis
Background:
- Increasing the sectioning rate of breast sentinel lymph nodes is crucial for detecting micrometastases.
- Serial sectioning requires efficient methods for analyzing digitized lymph node sections.
Purpose of the Study:
- To develop and evaluate a computer-assisted detection (CAD) algorithm for analyzing digitized breast sentinel lymph node sections.
- To assess the algorithm's accuracy in detecting micrometastases and other malignant foci.
Main Methods:
- Developed a CAD algorithm utilizing K-means clustering to segment image pixels into distinct categories.
- Employed four filters to refine the tumor cluster and remove false-positive pixels.
- Validated the algorithm on 43 positive and 59 negative lymph node sections, comparing results to pathologist diagnoses.
Main Results:
- Achieved 100% sensitivity and specificity in identifying the largest focus (micrometastasis) in positive sections.
- Detected isolated tumor cells (ITCs) in a previously negative slide and identified all 259 malignant foci.
- Demonstrated high sensitivity for micrometastases (89.5%) and larger metastases (100%), detecting nine additional previously unidentified metastases.
Conclusions:
- The developed CAD algorithm is highly effective for sentinel lymph node evaluation.
- The algorithm shows significant potential for enhancing the detection of occult micrometastases in breast cancer diagnosis.
