Local Binary Convolutional Neural Networks' Long Short-Term Memory Model for Human Embryos' Anomaly Detection

Table 10

Results and discussion based on different types of approach and methodology.

Authors	Year	Device	Input image	Results	Models
Kragh et al. [26]	2019	Time-lapse incubator	Video: from days 1 to 5	Quality analysis: AUC = 0.96	Approach of CNN-LSTM
Khosravi et al. [23]	2019	Time-lapse incubator	Blastocyst	Quality analysis: AUC = 0.98	Google’s Inception model
Tran et al. [22]	2019	Time-lapse incubator	Blastocyst transfer	Abnormality analysis: AUC = 0.93	Deep learning approach
Dirvanauskas et al. [27]	2019	Time-lapse incubator	Video: from days 1 to 6	Abnormality analysis: AUC = 0.98	Two-classifier vote-based CNN
Lee et al. [25]	2021	Time-lapse incubator	Video: from days 1 to 5	Abnormality analysis: AUC = 0.74	Deep learning approach
Sawada et al. [19]	2021	Time-lapse incubator	Blastocyst	Abnormality analysis: AUC = 0.93	LSTM with attention map
Payá et al. [40]	2022	Time-lapse incubator	Video: from days 1 to 4	Abnormality analysis: AUC = 0.94	A supervised contrastive learning framework
Our proposed method	2022	Time-lapse incubator	Video: from days 1 to 5	Abnormality analysis: AUC = 0.98	LBCNN-LSTM