14 June 2022: Clinical Research
Fully Automatic Knee Joint Segmentation and Quantitative Analysis for Osteoarthritis from Magnetic Resonance (MR) Images Using a Deep Learning Model
Xiongfeng Tang 1ABCEF , Deming Guo 1BF , Aie Liu 2ACDE , Dijia Wu 2ADE , Jianhua Liu 3BCD , Nannan Xu 3BDF , Yanguo Qin 1ACDG*DOI: 10.12659/MSM.936733
Med Sci Monit 2022; 28:e936733
Table 1 Dataset demographic breakdown.
| Datasets | Model training set | Model test set | Clinical test set |
|---|---|---|---|
| Patients (n) | n=500 | n=137 | n=206 |
| Age (years) | 46 (32–56) | 43 (32–54) | 53 (40–62) |
| Male | 40.6 (27.75–52) | 36 (29.5–49.5) | 44 (33–57) |
| Female | 49 (33.75–59) | 52 (40–60.25) | 56 (48–65) |
| Sex | |||
| Male (%) | 246 (49.2) | 66 (48.2) | 72 (35) |
| Femal (%) | 254 (50.8) | 71 (51.8) | 134 (65) |
| Magnetic strength | |||
| 1.5T (%) | 283 (56.6) | 70 (51.1) | 136 (66) |
| 3.0T (%) | 217 (43.4) | 67 (48.9) | 70 (34) |
| Side | |||
| Left (%) | 247 (49.4) | 75 (54.7) | 99 (48) |
| Right (%) | 253 (50.6) | 62 (45.3) | 107 (52) |
| Cohort (%) | OA: n=167 (33.4) | OA: n=46 (33.5) | OA: n=82 (40.2) |
| ACL/MI: n=174 (36.8) | ACL/MI: n=53 (38.7) | Control: n=124 (60.8) | |
| Control: n=159 (31.8) | Control: n=38 (27.7) | ||
| Typical parameters | GE Optima MR430s 1.5T: Filed of view, 160×160 mm;Dimensions 512×512×20; voxel spacing 0.35×0.35×4.5 mm; slice thickness, 3.5 mm; spacing between slices, 4.5 mm; Repetition Time, 2000 msec; Echo Time, 36.0 msec; fiip angle, 90. GE Discovery MR750 3.0T: Filed of view, 160×160 mm; Dimensions 512×512×20; voxel spacing 0.35×0.35×4.5; slice thickness, 3.5 mm; spacing between slices, 4.5 mm; Repetition Time, 2600 msec; Echo Time, 34.0 msec; fiip angle, 90 | ||
| Unless otherwise specified, data in parentheses are percentages. ACL – anterior cruciate ligament injury; MI – meniscus injury; OA – osteoarthritis. The data of the OA and control group in model testing and clinical test set are added together to make nonparametric test for quantitative biomarkers identification. | |||