Files

Download

Download Full Text (2.1 MB)

Abstract

Axillary Lymph Node involvement is the most important prognostic factor in the assessment of Estrogen Receptor positive (ER+) breast cancer (BCa) patients. Patients with evidence of metastatic disease in axillary lymph nodes positive (LN+) patients have poorer survival prospects and higher likelihood of metastasis. Standard clinical treatment for LN+ patients includes adjuvant chemotherapy, though not all LN+ patients will benefit from it. Although the absence of axillary lymph node involvement at initial diagnosis typically indicates lower risk, 30% of these LN-, ER+ BCa patients will ultimately die from breast cancer metastasis, even with optimal treatment. There is no standard criteria for identifying high-risk LN- tumors that require adjuvant treatment. The ability to prognosticate risk of recurrence and mortality with respect to LN status would enable physicians to develop more appropriate treatment plans for their patients. This study uses the machine learning approach, Multiple Instance Learning (MIL), to identify the prognostic ability of computer extracted feature of cancer nuclei on H&E images for predicting short-term (

Symposium Date

Fall 12-1-2012

Keywords

machine learning, pathology, cancer

Creative Commons License

Creative Commons Attribution-NonCommercial 4.0 International License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License

Nuclear Shape of ER+ Breast Tumor Whole Slide Images Predict Recurrence and Survival

Share

COinS