Table 5 The top-ranked features with the corresponding coefficient in the model and the UMLS concept preferred name for the CUIs
From: Using natural language processing and machine learning to identify breast cancer local recurrence
Feature | Coefficient | UMLS Concept Preferred Name |
|---|---|---|
C0278493 | 0.66 | ‘Recurrent breast cancer’ |
{C0007124; C0222600; C0222600} | 0.46 | {‘Noninfiltrating Intraductal Carcinoma’; ‘Right breast’; ‘Right breast’} |
C0920420 | 0.43 | ‘Cancer recurrence’ |
C1458156 | 0.41 | ‘Recurrent Malignant Neoplasm’ |
C2945760 | 0.40 | ‘Recurrent’ |
C0235653 | −0.36 | ‘Malignant neoplasm of female breast’ |
C0277556 | 0.36 | ‘Recurrent disease’ |
C1512083 | − 0.35 | ‘Ductal’ |
{C0007124; C0205090; C0262512} | 0.32 | {‘Noninfiltrating Intraductal Carcinoma’; Right; ‘History of present illness’} |
C4042789 | 0.30 | ‘Right-Sided Breast Neoplasms’ |