Research Article

All-Cause Death Prediction Method for CHD Based on Graph Convolutional Networks

Table 1

Basic information of postoperative patients obtained through follow-up from 2016 to 2021.

ParametersBefore processing (n = 5764)After processing (n = 2702)

Age (years)65 (57–72)65 (57–72)
Sex (M/F)4703 (81.6%)2207 (81.7%)
BMI24.4 (23.0–25.6)24.0 (22.7–26.0)
Smoking1618/5337 (30.3%)1171/2679 (43.3%)
Diabetes1600/5301 (30.2%)794/2650 (30.0%)
Hypertension2852/5332 (53.5%)1141/2683 (42.5%)
History of renal insufficiency63/5309 (1.2%)26/2673 (1.0%)
Heart rate69.7 (62.0–76.0)68 (61.0–76.0)
E ′wave rate0.06 (0.05–0.14)0.06 (0.04–0.07)
Left ventricular ejection fraction59.6 (56.1–66.5)63.3 (55.1–68.6)
Left ventricular mass index110.7 (99.0–113.8)103.4 (86.5–121.0)
Total cholesterol4.27 (3.43–4.80)3.95 (3.24–4.85)
Low-density lipoprotein2.74 (1.98–3.27)2.48 (1.85–3.28)
Triglycerides1.57 (1.10–2.54)1.43 (1.03–2.00)
NT-proBNP351 (77–866)129 (37–538)
Apolipoprotein A1.21 (1.05–1.40)1.19 (1.04–1.33)
Statins4104/5265 (77.9%)2099/2670 (78.6%)
Spironolactone621/4796 (12.9%)252/2626 (9.6%)
Aspirin5113/5194 (98.4%)2617/2664 (98.2%)