Table 4 Other work relevant to automated segmentation of knee structures [19,20,21,22,23]. DSC = Dice Score Coefficient, OAI = Osteoarthritis Initiative, DESS = Double Echo Steady State
Study | Sequences | Subjects | Neural network | Results (DSC ± Standard deviation) | Comment |
|---|---|---|---|---|---|
F. Liu et al. [19] | T2 FS FSE PD FSE T2 Mapping | 175 | 2D encoder-decoder VGG16 | Femur (0.96 ± 0.03) Tibia (0.95 ± 0.03) Femoral cartilage (0.81 ± 0.04) Tibial cartilage (0.82 ± 0.04) | |
Z. Zhou et al. [20] | PD FS FSE | 20 | 2D encoder-decoder VGG16 | Femur (0.970 ± 0.010) Femoral cartilage (0.806 ± 0.062) Tibia (0.962 ± 0.015) Tibial cartilage (0.801 ± 0.052) Patella (0.898 ± 0.033) Patellar cartilage (0.807 ± 0.101) Meniscus (0.831 ± 0.031) Quadriceps and patellar tendon (0.815 ± 0.029) Muscle (0.932 ± 0.024) Joint effusion and Baker’s cyst (0.736 ± 0.069) Infrapatellar fat pad (0.882 ± 0.040) Other non-specified tissues (0.913 ± 0.017) | All subjects had various degrees of Osteoarthritis |
A. Tack et al. [21] | DESS | 88 | 3D U-Net | Medial menisci (83.8) Lateral menisci (88.9) | OAI Imorphics dataset |
F. Ambellan et al. [22] | DESS | 88 | 3D U-Net | Femoral cartilage (0.89 ± 2.41) Medial tibial cartilage (86.1 ± 5.33) Lateral tibial cartilage (90.4 ± 2.42) | OAI Imorphics dataset |
DESS | 507 | 3D U-Net | Femoral bone (98.5 ± 3.02) Femoral cartilage (89.9 ± 3.25) Tibial bone (98.5 ± 3.25) Tibial cartilage (85.6 ± 4.54) | OAI ZIB dataset | |
E. Panfilov et al. [23] | DESS | 88 | U-Net | Femoral cartilage (0.907 ± 0.019) Tibial cartilage (0.897 ± 0.028) Patellar cartilage (0.871 ± 0.046) Meniscus (0.863 ± 0.034) | OAI Imorphics dataset |
DESS | 44 | U-Net | Femoral cartilage (0.827 ± 0.024) Tibial cartilage (0.816 ± 0.029) |