Related Experiment Video
Updated: Aug 23, 2026

Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
Predicting the recurrence/metastasis of stage II and III breast cancer with lymph node metastasis
Masaya Mukai1, Shinkichi Sato, Tomomi Kimura
1Department of Surgery, Tokai University Oiso Hospital, Oiso, Kanagawa 259-0198, Japan. onc-res@oiso.u-tokai.ac.jp
Abstract:
This study compared prediction of the recurrence of breast cancer by detection of occult neoplastic cells (ONCs) in lymph nodes or by using the criteria for a high risk of recurrence and metastasis of gastric/large bowel cancer. The subjects were 45 patients with stage II and III node-positive breast cancer. Prediction of recurrence by detection of ONCs showed a sensitivity of 78.6% (11/14), a false-negative rate of 21.4% (3/14), a specificity of 96.4% (30/31), a false-positive rate of 3.2% (1/31), and an accuracy of 87.7% in patients with stage II and III node-positive cancer. Prediction of recurrence based on positivity for at least 2 of the high-risk criteria showed a sensitivity of 92.9% (13/14), a false-negative rate of 7.1% (1/14), a specificity of 87.1% (27/31), a false-positive rate of 12.9% (4/31), and an accuracy of 90.0% in patients with stage II and III node-positive cancer. These results suggest that ONCs plus the high-risk criteria are useful for predicting recurrence/metastasis of stage II and III node-positive breast cancer during the early postoperative period with a high sensitivity and accuracy.