METHOD FOR PREDICTING EFFICIENCY OF NEOADJUVANT CHEMIO-RADIATION THERAPY IN PATIENTS WITH RECTAL CANCER DURING PRIMARY MRI STUDIES Russian patent published in 2021 - IPC A61B6/00 G06T7/10 A61N5/00 

FIELD: medicine.

SUBSTANCE: invention relates to medicine, namely to oncology, and can be used to predict the effectiveness of neoadjuvant chemoradiation therapy (NDCT) in patients with rectal cancer during primary MRI examination. Texture analysis of the MRI image of the primary tumor in T2-WI mode is performed. For analysis, MRI images obtained in the axial plane at the level of the center of the tumor using an FSE pulse sequence with high spatial resolution are used: TR (sequence repetition period) - 4020 ms, TE (echo appearance time) - 97 ms, slice thickness/step - 3.0/0.3 mm, FoV (field of view size) - 250 mm, MTX (matrix size/pixels) - 286×512. Next, the selected images are loaded into the Mazda ver.4.6 program and the image segmentation is carried out. The texture parameters are determined by the GLCM gray level co-occurrence matrix method at a distance of 2 pixels and in four different directions - 0°, 45°, 90°, 135°. The obtained values are averaged and, using a computer program, the parameters of texture analysis are automatically calculated, which are used to predict the effectiveness of NHLT based on the scoring system. AngScMom - if it is greater than or equal to 0.0016, then 1 point is assigned, if less than 0.0016 - 0 points. SumofSqs - if less than or equal to 101.88, then 1 point is assigned, if more than 101.88 - 0 points. SumVarnc - if less than or equal to 277.51, then 1 point is assigned, if more than 277.51 - 0 points. SumEntrp - if less than or equal to 1.81, then 1 point is assigned, if more than 1.81 - 0 points. Entropy - if less than or equal to 2.88, then 1 point is assigned, if more than 2.88 - 0 points. If the sum of the received points is greater than or equal to 3, then a good response to NDCT is predicted. If the sum of the received points is less than 3, then no response to NDCT is predicted.

EFFECT: invention provides prediction of the effectiveness of NDCT due to quantitative data of 5 parameters of texture analysis (AngScMom, SumVarnc, SumofSqs, Entropy, SumEntrp) T2-WI magnetic resonance imaging of the tumor during the primary MRI study.

1 cl, 11 dwg, 4 tbl, 4 ex