Automated classification of urine biomarkers to diagnose pancreatic cancer using 1-D convolutional neural networks

Table 2 Evaluation of all tested classifiers to diagnose pancreatic cancers using urine biomarkers

Classifier	Pancreas Condition	Recall	Precision	Specificity	F1-score	AUC	Accuracy
MLP Network	Healthy case	0.63	0.76	0.87	0.69	0.80	0.70
	Benign	0.64	0.61	0.72	0.62	0.69
	PDAC	0.84	0.74	0.78	0.79	0.79
Random Forest	Healthy case	0.73	0.83	0.91	0.78	0.89	0.75
	Benign	0.67	0.65	0.77	0.66	0.74
	PDAC	0.87	0.79	0.84	0.82	0.84
1D CNN	Healthy case	0.90	0.97	0.99	0.93	0.99	0.93
	Benign	0.89	0.95	0.97	0.92	0.95
	PDAC	1.00	0.90	0.93	0.95	0.98
1D CNN-LSTM	Healthy case	0.94	1.00	1.00	0.97	0.99	0.97^a
	Benign	0.95	0.95	0.97	0.95	0.96
	PDAC	1.00	0.96	0.97	0.98	0.99

^aBest performance value is indicated in bold. Abbreviations are already defined in the context

ISSN: 1754-1611