ECG beat classification on MIT-BIH Arrhythmia Database AAMI labeling convention (intersubject) benchmark leaderboard · SOTA2 Research

ECG beat classification on MIT-BIH Arrhythmia Database AAMI labeling convention (intersubject)

96.4Overall AccuracyNiu et al.

Evaluation Results

Method
Niu et al.2020.06	96.4	—	76.5	76.6	76.6	—	85.7	94.1	89.7
Huang et al.2020.06	93.8	95.1	91.1	42.2	57.7	99	93.9	90.9	92.4
Subject-specific Contrastive Loss2020.06	93.2	96	42.8	50.1	43.8	98.9	96.6	88.7	92.4
End-to-end2020.06	91.9	96.3	44.5	52.2	46.7	98.4	96.4	83.4	89.2
Subject-invariant Self-Supervised Learning2020.06	91.8	96.2	35.5	48.8	40.8	98	95.3	79.2	86.3
Base SSL2020.06	91.7	95.6	30.2	39.2	33.3	98.5	96.8	83.7	89.6
Subject-invariant Self-Supervised Learning2020.06	91.3	95.8	26.5	40.1	31.6	98.1	96.2	80.4	87.2
End-to-end2020.06	91	95.9	25.8	41.8	31.4	97.8	95.6	78.9	85.8
Subject-invariant Self-Supervised Learning2020.06	90.4	95.7	35.2	42.2	37.3	98.1	96.4	80	87.2
Subject-invariant Self-Supervised Learning2020.06	89.4	96.3	38.8	48.9	42.5	97.8	95.8	77.4	85.4
Luo et al.2020.06	89.3	96.2	15.4	47.3	23.2	95.5	60.4	66.8	63.4