METHOD OF PREDICTION OF PROBABILITY OF CERVICAL CANCER DETECTION BASED ON ROUTINE BLOOD PARAMETERS Russian patent published in 2023 - IPC A61B5/00 G01N33/49 G01N33/574 G01N33/68 G01N33/70 G01N33/72 G01N33/84 G01N33/86 G01N33/90 

FIELD: medicine; gynecology; oncology.

SUBSTANCE: invention can be used to predict the probability of detecting cervical cancer (CC). The determination of indicators in the blood serum is carried out. The following indicators are determined: Age, Fibrinogen, Platelet aggregation with ADP, APTT, ESR, Hemoglobin, Erythrocytes, Leukocytes, Average volume of erythrocyte, Alkaline phosphatase, CRP, Serum iron, Transferrin saturation with iron, Creatinine, AST, Transferrin. Taking into account the obtained data, the prognostic coefficient is calculated by the formula: VLOOKUP = 1/(1 + e^-x), where: PRT is the probability of predicting the result — the presence of cervical cancer, e is the base of the natural logarithm, x = +0.663× (Age) + 3,786× (Fibrinogen) + 0.904× (APTT) + 0.368× (Saturation of transferrin with iron) + 0.254× (ESR) + 0.180× (Alkaline Phosphatase) + 0.137× (CRP) - 0.116× (Serum iron) — 0.155× (creatinine) — 0.195× (Cf. erythrocyte volume) — 0.195× (Platelet aggregation with ADP) — 0.231× (AST) — 0.378× (Transferrin) — 0.964× (Leukocytes) — 4.462× (Erythrocytes) — 4.001, where: Age, years; Fibrinogen, g/l; APTT is activated partial thrombin time, s; Transferrin saturation with iron, %; ESR according to Westergren, mm/h; Alkaline phosphatase, u/l; CRP is C-reactive protein, mg/l; Serum iron, µmol/l; Creatinine, mmol/l; Ср volume of erythrocyte is average volume of erythrocyte, fl; Platelet aggregation with ADP, %; AST, aspartate aminotransferase, U/L; Transferrin, g/l; Leukocytes, 10⁹/l; Erythrocytes, 10¹²/l. If the CDF is greater than 0.50, a high probability of detecting cervical cancer is predicted. If the CDF is greater than 0.27 and less than or equal to 0.50, the average probability of detecting cervical cancer is predicted. If the CDF is less than or equal to 0.27, cervical cancer is unlikely.

EFFECT: method provides the possibility of predicting the probability of cervical cancer detection in women and reducing the risk of errors at any stage of preclinical diagnosis of cervical cancer with negative or doubtful results of other types of examination in patients with existing visual and/or cytological pathology of the cervix by determining a number of laboratory parameters, taking into account the age of the patient and the use of a multivariate binary logistic regression machine learning algorithm.

1 cl, 1 dwg, 6 tbl, 5 ex