Machine Learning Analysis of Image Data Based on Detailed MR Image Reports for Nasopharyngeal Carcinoma Prognosis

Table 1

Patient characteristics.

Characteristics	n (%)	OS (%)	DMFS (%)	LRFS (%)
Total	792 (100%)	89.0	88.1	90.2
Gender
Male	576 (72.7%)	91.7	91.7	93.1
Female	216 (27.3%)	97.3	96.5	97.1
Age
≥45	397 (50.1%)	92.3	92.8	94.4
<45	395 (49.9%)	96.7	95.3	95.7
Histology
WHO I	5 (0.6%)	99.9	99.9	99.9
WHO II	41 (5.2%)	99.1	99.1	99.4
WHO III	746 (94.2%)	90.0	89.1	90.9
T stage (a)
T1	204 (25.8%)	99.0	98.5	98.2
T2	97 (12.2%)	98.6	98.6	99.1
T3	296 (37.4%)	96.6	95.7	96.5
T4	195 (24.6%)	94.8	95.3	96.3
N-stage (a)
N0	182 (23.0%)	98.7	98.6	98.9
N1	438 (55.3%)	94.4	94.3	93.8
N2	113 (14.3%)	97.7	96.8	98.4
N3	59 (7.4%)	98.1	98.4	99.1
Clinical stage (a)
1	73 (9.2%)	99.9	99.7	99.6
2	175 (22.1%)	98.9	98.5	98.2
3	303 (38.3%)	96.6	95.6	96.6
4	241 (20.4%)	93.7	94.3	95.7

Abbreviations: WHO, World Health Organization; (a) according to the 8th AJCC/UICC.