Related Experiment Video
Updated: Jul 16, 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
A logistic regression model to predict nodal malignancy among cases with lymphadenopahty
G M Malik1, M A Abolfotouh, S Jastania
1Department of Medicine, Family and Community Medicine, Surgery, and Pathology, College of Medicine, King Saud University, and Departments of Surgery and Medicine, Asir Central Hospital, Abha, Saudi Arabia.
Background:
Peripheral lymphadenopathy can be caused by benign and malignant diseases. In this logistic regression model, we attempted to identify the clinical findings predicting high probability of nodal malignancy.
Materials And Methods:
Two hundred and twenty cases diagnosed by peripheral lymph node biopsy were studied. Of these, 164 had benign lymph node pathology, while the other 56 were malignant. The patientsâ medical charts were reviewed and a logistic regression model used to identify physical signs and simple investigations that will predict nodal malignancy.
Results:
Hard lymph node consistency, negative Mantoux test and positive abdominal ultrasound (showing hepatosplenomegaly with or without lymphadenopathy) were found to correlate with a higher probability of nodal malignancy.
Conclusion:
A logistic regression model is proposed to calculate the probability of lymph node malignancy at different ages and both sexes in relation to lymph node consistency, Mantoux test and abdominal ultrasound results.
