Lymph Node Morbidity Diagnosis Using Multiclass Machine Learning Models

Lymphography, considered a corner stone in prognosis and diagnosis of lymphatic disorders continues to be a gold standard of reference in spite of the advancements in health technologies. However, analyzing the lymphatic characteristics implicitly curtails the diagnostic accuracy of few dreaded cancers such as lymphoma, malign lymph's etc., Thus to provide objective diagnosis computer aided diagnostic tools (CAD) play a prominent role. In this research, the role of robust machine learning classifiers in classifying lymphatic characteristics is proposed. The highest accuracy obtained by considering the prominent lymph characteristics is 85%. A good balance between specificity and sensitivity was obtained. The proposed system can be employed in a clinical scenario particularly in regions with poor medical infrastructures.