METHOD FOR PREDICTION OF LOCAL POSTOPERATIVE RECURRENCES IN PATIENTS WITH BREAST CANCER Russian patent published in 2015 - IPC A61B5/00 

FIELD: medicine.

SUBSTANCE: degree of malignancy and tumour growth form are described, as well as metastatic lymph nodes are detected. If observing 2-3 metastatic lymph nodes, the value f is derived by a regression equation of formula: f_rec = -2.686+2.681·x₁+2.3·x₂+2.172·x_3, wherein -2.686 is an intercept regression coefficient; x₁ is the tumour growth form: 0 is a unicentric growth, and 1 is a multicentric growth, 2.681 is the regression coefficient of the above sign; x₂ is the degree of tumour malignancy: 0 is the low/moderate degree, 1 is the high degree of malignancy, 2.3 is the regression coefficient of the above sign; x₃ is the status assessment of regional lymph nodes N: 1 - 2 to 3 involved lymph nodes, 2.172 is the regression coefficient of the above sign. That is followed by calculating a probability (P) of disease progression by formula: $$P_{\text{recurrence}}=\frac{1}{1+e^{-f_{rec}}},$$ wherein e is a mathematic constant equal to 2.72. If P<0.5, a low probability of local recurrences is predicted; and if P>0.5, a high probability of local recurrences is predicted.

EFFECT: method enables increasing the prediction accuracy by taking into account the complex of prognostically significant sings.