Fig. 4From: Identifying tuberculous pleural effusion using artificial intelligence machine learning algorithmsThe impacts of the first 12 features on the accuracy of the RF model. pfADA: pleural fluid adenosine deaminase; pfCEA: pleural fluid carcinoembryonic antigen; WBC: white blood cells; M: monocyte; L: lymphocyte; N: neutrophil; pfLDH: pleural fluid lactate dehydrogenaseBack to article page