| CPC A61N 5/1038 (2013.01) [A61N 5/1031 (2013.01); G16H 20/40 (2018.01); G16H 50/20 (2018.01); A61N 5/1084 (2013.01)] | 28 Claims |
|
1. A system comprising:
a memory; and
one or more processors that, when executing instructions stored in the memory, are configured to perform operations comprising:
receiving dose information representing delivery of a dose distribution to a target corresponding to a first radiotherapy treatment fraction;
determining a type of tumor associated with the target;
selecting a machine learning technique from a plurality of machine learning techniques based on the determined type of tumor associated with the target, a first of the plurality of machine learning techniques trained to estimate a measure of biologically effective dose (BED) associated with a first type of tumor, and a second of the plurality of machine learning techniques trained to estimate a measure of BED associated with a second type of tumor:
applying the selected machine learning technique to the dose information to predict a measure of BED, the selected machine learning technique trained to establish a relationship between a set of prior dose distributions delivered for the determined type of tumor and one or more ground truth BED measurements, the selected machine learning technique trained by:
obtaining a set of training data comprising a set of training dose distributions and a ground truth measure of BED;
applying the selected machine learning technique to the set of training dose distributions to predict a training measure of BED;
computing a deviation between the predicted training measure of BED and the ground truth measure of BED; and
updating one or more parameters of the selected machine learning technique based on the deviation; and
performing an isotoxic planning process based on the predicted measure of BED.
|