Table 3 Classification scores (> 0.97) for LOSO CV and other CV schemes.
From: Video-based formative and summative assessment of surgical tasks using deep learning
Model | Method | Suturing | Needle passing | Knot tying | Mean |
|---|---|---|---|---|---|
Khalid et al.25 [2020] | Autoencoder | 0.970 | 0.970 | 0.970 | 0.970 |
Nguyen et al.41 [2019] | CNN-LSTM | 0.984 | 0.984 | 0.948 | 0.972 |
CNN (adjusted LOSO) | 1.0 | 1.0 | 0.921 | 0.974 | |
Funke et al.7 [2019] | 3DCNN + TSN | 1.0 | 0.964 | 0.958 | 0.974 |
Soleymani et al.38* [2021] | CNN + FFT (tenfold) | N/A | N/A | N/A | 0.973 |
VBA-Net | DAE + classifier | 1.0 | 1.0 | 0.926 | 0.975 |
Castro et al.10* [2019] | CNN (adjusted LOSO) | 0.984 | 0.989 | 0.989 | 0.987 |
Zia and Essa34 [2018] | ApEn** | 1.0 | 1.0 | 0.999 | 0.999 |
