Reinforcement learning based tumor treatment auxiliary decision-making method

Abstract

The invention provides a tumor treatment auxiliary decision-making method based on reinforcement learning, and the method comprises the steps: constructing and training a patient state transition model and a treatment decision-making model, wherein the patient state transition model carries out the modeling of the state change of a tumor patient through a recurrent neural network, using a diagnosis and treatment scheme and examination indexes obtained in actual diagnosis and treatment as data sources, and simulating the state change of the tumor patient under a certain diagnosis and treatment scheme. The treatment decision model carrying out modeling on the process of treating a tumor patient through reinforcement learning, the patient state transfer model is used as a data source, a deep Q network (DQN) is constructed to serve as a strategy network, an excitation function is constructed on the basis of RECIST solid tumor curative effect evaluation standards and TNM tumor imageological staging standards, and iterative training is carried out. According to the method, the workload of tumor specialists can be reduced to a certain extent, the treatment means predicted by the model can serve as decision reference to reduce decision errors of doctors, and the doctors are assisted to specify feasible treatment schemes.

Description

technical field [0001] The invention belongs to the medical field, and relates to a computer-aided decision-making method for medical use, in particular to an auxiliary decision-making method for tumor treatment based on reinforcement learning. Background technique [0002] Reinforcement learning is an important branch of machine learning, which is used to describe and solve the problem of agents learning strategies to maximize rewards or achieve specific goals in the process of interacting with the environment. The basic principle is: if a certain behavioral strategy of the agent leads to a positive reward signal in the environment, then the tendency of the agent to produce this behavioral strategy will be strengthened in the future. The agent's goal is to discover the optimal policy at each discrete state to maximize the sum of expected discounted rewards. For example, the situation of Go is the state, and what the agent needs to learn is how to play each situation. [0...

AI Technical Summary

Problems solved by technology

[0004] The treatment of tumors is closely related to the life safety of patients, and oncologists who make decisions on tumor treatment are under heavy workload and pressure, and are prone to misjudgment or missed judgment of the disease, thus making inappropriate treatment decisions

Images