Method of generating MLC-based intensity modulated treatment plan model for predicting delivery errors using machine learning techniques and method of predicting delivery errors using it

摘要

A method for generating a propagation error prediction model for an intensity-modulated radiation therapy plan of a multi-leaf collimator (MLC) using mechanical learning and a propagation error predicting method using the same. A method for generating a propagation error prediction model for an intensity-modulated radiation therapy plan of an MLC using mechanical learning comprises the steps of receiving at least one treatment plan MLC data, driving a linear accelerator based on the received treatment plan MLC data, Calculating an error between the treatment planning MLC data and the transfer MLC data, and generating a transmission error prediction model based on the calculated error, wherein the transmission error prediction method using the transmission error prediction model Comprising the steps of receiving a treatment plan MLC data of a patient, predicting predictive MLC data using the delivery error prediction model based on the received treatment plan MLC data, and estimating the predicted MLC data based on the predicted MLC data And judging whether or not the dose to be irradiated to the patient is acceptable at the control point.;