P-1 Gastric cancer: Artificial intelligence, complex system analysis and simulation for best management

We examined factors significantly affecting gastric cancer (EC) patients (GCP) survival. analyzed data of 799 consecutive GCP (T1-4N0-2M0) (age=57.1±9.4 years; tumor size=5.4±3.1 cm) radically operated (R0) and monitored in 1975-2022 (m=558, f=241; total gastrectomies=173, distal gastrectomies=461; proximal gastrectomies=165; combined gastrectomies=247 with resection esophagus, pancreas, liver, duodenum, diaphragm, colon transversum, splenectomy, etc; only surgery-S=624, adjuvant chemoimmunotherapy-AT=175 (5-FU + thymalin/taktivin); T1=238, T2=220, T3=184, T4=157; N0=437, N1=109, N2=253, M0=799; G1=222, G2=164, G3=413. Variables selected for prognosis study were input levels 45 blood parameters, sex, age, TNMG, cell type, size. Survival curves estimated by the Kaplan-Meier method. Differences between groups evaluated using a log-rank test. Multivariate Cox modeling, discriminant analysis, clustering, SEPATH, Monte Carlo, bootstrap neural networks computing used to determine any significant dependence. Overall life span (LS) was 2128.9±2300.3 days cumulative 5-year survival (5YS) reached 58.4%, 10 years – 51.9%, 20 39%, 30 27.2%. 318 lived more than 5 (LS=4304.5±2290.6 days), 169 (LS=5919.5±2020 days). 290 died because GC (LS=651±347.2 modeling displayed that G CP depended on CRF: healthy cells/CC, erythrocytes/CC, monocytes/CC, phase transition (PT) terms synergetics early—invasive cancer; PT N0--N12, G1-3, hemorrhage time, ESS, AT, prothrombin index, residual nitrogen. Neural networks, genetic algorithm selection simulation revealed relationships 5YS (rank=1); N0--N12 (2); cells/CC (3), erythrocytes/CC (4), thrombocytes/CC (5), monocytes/CC (6), segmented neutrophils/CC (7), leucocytes/CC (8), lymphocytes/CC (9), stick (10), eosinophils/CC (11). Correct prediction 100% (area under ROC curve=1.0; error=0.0). after radical procedures on: 1) “early-invasive cancer”; 2) N0--N12; 3) Cell Ratio Factors; 4) circuit; 5) biochemical factors; 6) hemostasis system; 7) AT; 8) characteristics; 9) localization; 10) anthropometric data; 11) surgery type; 12) dynamics. Best management are: screening early detection GC; availability experienced abdominal surgeons complexity procedures; aggressive en block adequate lymph node dissection completeness; precise prediction; chemoimmunotherapy unfavorable prognosis.