Skip to main content

Table 3 Performance comparison for concept extraction

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

LSTM_general

0.9492

0.9186

0.9337

0.9630

0.9320

0.9472

LSTM_mimic

0.9464

0.8682

0.9056

0.9609

0.8810

0.9192

BERT_general

0.8885

0.9575

0.9217

0.9067

0.9739

0.9391

BERT_mimic

0.9486

0.952

0.9503

0.9642

0.9648

0.9645

RoBERTa_general

0.9248

0.9636

0.9438

0.9353

0.9739

0.9542

RoBERTa_mimic

0.9391

0.9551

0.947

0.9498

0.9654

0.9575

  1. *Best F1 scores are highlighted in bold
Back to article page

BMC Medical Informatics and Decision Making

ISSN: 1472-6947

Contact us