Image feature and smart regressive model-based method for optimizing intensity modulated radiotherapy plan

Abstract

The invention discloses an image feature and smart regressive model-based method for optimizing an intensity modulated radiotherapy plan, which comprises the steps of: firstly, establishing a database of an expert-level intensity modulated radiotherapy plan, and storing a record of each intensity modulated radiotherapy case in the database; secondly, establishing a smart regressive model of medical image feature-based intensity modulated radiotherapy plan optimization parameters by supervised machine learning according to relevance information between the original medical image characteristics in the database of the expert-level intensity modulated radiotherapy plan and the intensity modulated radiotherapy plan optimization parameters; carrying out image analysis on a medical image of a specific intensity modulated radiotherapy object, and carrying out characteristic extraction; inputting the image characteristics into the smart regressive model, and outputting the intensity modulated radiotherapy plan optimization parameters; and setting the intensity modulated radiotherapy plan optimization parameters on a therapy plane system (TPS) according to the results, and operating a plan optimization function of the TPS to obtain an optimal therapy plane of intensity modulated radiotherapy.

Description

technical field [0001] The invention relates to a method for optimizing a tumor radiotherapy plan, which belongs to the technical field of tumor radiotherapy, in particular to a method for optimizing an intensity-modulated radiotherapy plan based on image features and an intelligent regression model. Background technique [0002] Intensity-modulated radiation therapy is a precise radiation therapy technology that makes full use of medical linear accelerator beams in different directions and intensities to irradiate the target tumor with high doses while minimizing radiation damage to surrounding vital organs. This technology can fully kill tumor tissue cells and protect surrounding important organs, and has produced huge application value in the field of tumor radiation therapy. It is called "one of the most important technological advances in the field of radiation therapy since the emergence of medical accelerators." ". The basic principle of intensity-modulated radiation...

AI Technical Summary

Problems solved by technology

[0009] 2. The efficiency of high-quality plan design is low

This exploratory process is clinically particularly time-consuming

For some tumors such as the head and neck, this kind of adjustment takes a week to complete, which greatly increases the burden on the dosimetrist

[0011] 3. The training cost of IMRT plan design technology is high

That is, 4 years of expensive training in exchange for a 20% reduction in the risk of death

Therefore, we have good reason to believe that it is quite critical to train dosimetrists in IMRT plan design in community hospitals or county and city hospitals, but the cost of training is indeed high

Images