Method for predicting the type of tumors, electronic apparatus, and computer storage medium

Abstract

The invention relates to a method for predicting the type of tumors, an electronic apparatus, and a computer storage medium. The method includes following steps: 1) acquiring the feature information of to-be-detected tumors; 2) acquiring the comparing result information between the genome detected sequence and reference genome sequence of a to-be-detection sample of the to-be-detected tumors; 3) based on the comparing result information, generating mutation type data related to the various preset mutation types; 4) based on the feature information and the mutation type data, generating the input data for inputting a prediction model; 5) through the prediction model, extracting a feature value of the input data, and predicting the type of the to-be-detection tumors on the basis of the extracted feature value, wherein the prediction model is generated by training machine learning models of a plurality of training samples. The disclosure is used for improving the prediction accuracy on the type of tumors in primary sites.

Description

technical field [0001] The present disclosure relates generally to biological information processing, and in particular, to methods, electronic devices, and computer storage media for predicting tumor types. Background technique [0002] The diagnosis of the primary site of cancer is the main basis for guiding clinical treatment. Traditional methods for predicting tumor types are mainly based on histology, such as immunohistochemical-based assessment and high-quality cross-sectional imaging of tumor tissue. The clinical treatment of cancer is closely related to the site of origin, histopathological subtype and stage of the tumor. However, traditional histology-based approaches to predict tumor type are challenging in many tumor cases, especially for those presenting metastatic, poorly differentiated tumors, where it is sometimes difficult to clearly and accurately determine the primary site. Unclear or incorrect classification of tumor types may negatively impact treatment...

AI Technical Summary

Problems solved by technology

However, traditional histology-based approaches to predict tumor type are challenging in many tumor cases, especially those presenting metastatic, poorly differentiated tumors, where it is sometimes difficult to clearly and accurately determine the primary site

Unclear or incorrect tumor type classification may negatively impact treatment options and outcomes

[0003] In summary, the traditional methods for predicting tumor types have the disadvantage that it is difficult to clearly and accurately determine the type of tumor at the primary site

Images