BMC Medical Informatics and Decision Making

Table 4 Detailed performance for each concept category for BERT_mimic

From: Identify diabetic retinopathy-related clinical concepts and their attributes using transformer-based natural language processing methods

	Strict			Lenient
	Precision	Recall	F1 score	Precision	Recall	F1 score
Lesion	0.9555	0.9576	0.9565	0.9776	0.9743	0.976
Severity	0.9627	0.9317	0.9469	0.9668	0.9357	0.951
Eye part	0.8	0.7059	0.75	0.8	0.7059	0.75
Laterality	0.9339	0.9608	0.9472	0.9439	0.9711	0.9573
Overall	0.9486	0.952	0.9503	0.9642	0.9648	0.9645

Back to article page

ISSN: 1472-6947

Contact us

General enquiries: journalsubmissions@springernature.com