Skip to main content
Fig. 2 | Respiratory Research

Fig. 2

From: The large language model diagnoses tuberculous pleural effusion in pleural effusion patients through clinical feature landscapes

Fig. 2

Clinical characteristic landscape of TPE. A PLS-DA score plot showing the separation between non-TPE and TPE samples based on blood routine and biochemical markers. The percentage of variance explained by each principal component (PC1 and PC2) is indicated in the figure. The distribution of the TPE group and non-TPE group was distinguishable along PC1 (16.76%) and PC2 (16.37%). B Heatmap displaying the expression levels of various biochemical and hematological variables in non-TPE and TPE samples. Samples are clustered into two major groups, with variable names listed on the right side. Red indicates values greater than the corresponding cutoff point (1), and green indicates values smaller than the corresponding cutoff point (0). The variables include: (1) Liver function: Alkaline phosphatase (ALP), Total bile acids (TBA); (2) Electrolytes: Osmolality, Sodium (Na); (3) Tumor markers: Carcinoembryonic antigen (CEA); (4) Hematological analysis: Red cell distribution width-CV (RDW-CV), Neutrophil count, Neutrophil percentage, Lymphocyte count, Monocyte percentage; (5) Hospitalization: Length of hospital stay after pleural biopsy; (6) Lipid profile: Triglycerides; (7) PE biochemistry: Adenosine deaminase (ADA), Albumin, Total protein; (8) Age. The figure illustrates the distribution differences of these variables between the TPE and non-TPE groups. C The bar chart displays differences in the levels of ADA, monocytes, and total protein. ADA (PE biochemistry): The proportion differences between the TPE and non-TPE groups when ADA values are either below or above the cutoff point (25.85). Monocyte proportion (blood cell analysis): The proportion differences between the TPE and non-TPE groups when the monocyte proportion is either below or above the cutoff point (6.95). Total protein (PE biochemistry): The proportion differences between the TPE and non-TPE groups when total protein values are either below or above the cutoff point (47.35). D ROC curves for distinguishing non-TPE and TPE using pleural fluid ADA levels, blood routine monocyte percentage, and total protein in pleural fluid as biomarkers. The Area Under the Curve (AUC) and 95% CI for each biomarker are provided. These were used to evaluate the diagnostic ability of ADA (PE biochemistry), monocyte proportion (blood cell analysis), and total protein (PE biochemistry) in distinguishing between the TPE and non-TPE groups

Back to article page