30-05-2023 дата публикации
Номер: CN116186575A
Принадлежит:
The invention discloses a machine learning-based mammary gland sampling data processing method, which comprises the following steps of applying an in-vitro excitation source to a mammary gland part of a patient to be detected, and collecting resistivity data of the patient to be detected; preprocessing the resistivity data to obtain corresponding four-dimensional input data; inputting the input data into a feature extraction network to obtain an output tensor of the feature extraction network; and classifying the output tensor of the feature extraction network by using an Adaboost algorithm, outputting a classified result, and judging whether the resistivity data is abnormal or not according to the output result. The feature extraction network performs processing and feature extraction on the input data, classifies the processed data, judges whether the data is abnormal or not, provides effective data support for judgment of doctors, and improves the accuracy of judgment.
Подробнее