Table 2.
Diagnostic performance of combined positron emission tomography (PET)/diffusion-weighted imaging (DWI) MRI for the detection of head and neck squamous cell carcinoma recurrence for a cut off of 3 in the patient-by-patient analysis (N = 74) and in the lesion-by-lesion analysis (N = 105)
| Combined PET/DWIMRI patient-by-patient analysis | Combined PET/DWIMRI lesion-by-lesion analysis | |
|---|---|---|
| Total number of patients/lesions | 74 | 105 |
| True positive, n | 37 | 40 |
| True negative, n | 33 | 58 |
| False positive, n | 3 | 4 |
| False negative, n | 1 | 3 |
| Sensitivity, % (95% CI) |
37/38, 97.4% (86.2–99.9) |
40/43, 93.0% (85.4–100.0) |
| Specificity, % (95% CI) |
33/36, 91.7% (77.5–98.2) |
58/62, 93.5% (87.5–99.6) |
| Positive predictive value, % (95% CI) | 37/40, 92.5% (79.6–98.4) |
40/44, 90.9% (82.6–99.2) |
| Negative predictive value, % (95% CI) | 33/34, 97.1% (84.7–99.9) |
58/61, 95.1% (90.1–100.0) |
| Accuracy, % (95% CI) |
70/74, 94.6% (86.7–98.5) |
98/105, 93.3% (89.0–97.7) |
Results of prospective assessments (without ADC or SUV thresholds) are shown. Results are reported with percentages (%) and 95% confidence intervals (95% CI)
ADC apparent diffusion coefficient, SUV standardised uptake value