Skip to main content

Table 5 The prediction performance of radiomics for NAC response in osteosarcoma patients

From: An updated systematic review of radiomics in osteosarcoma: utilizing CLAIM to adapt the increasing trend of deep learning application in radiomics

Clinical question MRI-driven radiomics prediction model for NAC response in osteosarcoma patients
Number of studies 4
Good responder/sample size 44/115
Pooled analysis
 DOR (95%CI) 28.83 (10.27–80.95)
 p value for DOR p < 0.001
 Sensitivity (95% CI) 0.84 (0.70–0.92)
 Specificity (95% CI) 0.85 (0.74–0.91)
 PLR (95% CI) 5.43 (3.11–9.49)
 NLR (95% CI) 0.19 (0.09–0.37)
 AUC (95% CI) 0.91 (0.88–0.93)
Heterogeneity
 Higgins I2 test I2 = 42.04%
 Cochran’s Q test Q = 5.18, p = 0.160
Publication bias
 Egger’s test p = 0.035
 Begg’s test p = 0.089
 Deeks’ test p = 0.069
Trim and fill method
 Number of missing studies 2
 Adjusted DOR (95%CI) 20.53 (7.80–54.06)
 p value for adjusted DOR p < 0.001
 Level of Evidence Weak
  1. AUC area under curve, CI confidential interval, DOR diagnostic odds ratio, NAC neoadjuvant chemotherapy, NLR negative likelihood ratio, n/a not applicable, PLR positive likelihood ratio