Direct medical treatment predictions using artificial intelligence
Inventors
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12051490-B2
Publication Date
2024-07-30
Expiration Date
Abstract
A device is disclosed herein that receives image data corresponding to an anatomy of a patient. The device applies the image data to one or more feature models trained using training data that pairs anatomical images to an anatomical feature label, and receives, as output from the one or more feature models, scores for each of a plurality of anatomical features corresponding to the image data. The device applies the scores as input to a treatment model, the treatment model trained to output a prediction of a measure of efficacy of a particular treatment based on features of the patient's anatomy. The device receives, as output from the treatment model, data representative of the predicted measure of efficacy of the particular treatment.
Core Innovation
The invention is an autonomous treatment determination tool that predicts an efficacy of a particular treatment for a patient. It directly receives image data of an anatomy of a patient and applies the image data to a feature extraction model that includes a first machine learning model trained using training data pairing anatomical images to an anatomical feature label. The feature extraction model outputs scores for each of a plurality of anatomical features corresponding to the patient anatomy.
The invention further applies the plurality of anatomical feature scores as input to a treatment model that includes a second machine learning model different from the first. The second machine learning model is trained to output a prediction of a measure of efficacy of a particular treatment based on features of the patient's anatomy. The tool then receives, as output from the treatment model, data representative of the predicted measure of efficacy of the particular treatment.
In described embodiments, the anatomical feature scores and their structure are consolidated for the treatment model, including using a feature vector that corresponds to the plurality of anatomical features with identification. The tool may include a body part determination and model selection using concordance between candidate feature extraction models and a determined body part. It may also support output of predicted efficacy measures for multiple candidate treatments.
Claims Coverage
The document contains three independent claims that cover a method, a non-transitory computer-readable medium, and a system. Across the independent claims, the coverage includes autonomous prediction of treatment efficacy by combining a first machine learning feature extraction model with a different treatment model.
Two-stage prediction using separate feature extraction and treatment models
receiving image data of an anatomy of a patient; applying the image data to a feature extraction model comprising a first machine learning model trained using training data pairing anatomical images to an anatomical feature label; receiving as output from the feature extraction model scores for each of a plurality of anatomical features corresponding to anatomy of the patient; applying the scores as input to a treatment model comprising a second machine learning model different from the first machine learning model, trained to output a prediction of a measure of efficacy of a particular treatment based on features of the patient's anatomy; receiving as output from the treatment model data representative of the predicted measure of efficacy of the particular treatment
Machine-implemented treatment efficacy determination
a non-transitory computer-readable medium comprising instructions encoded thereon for autonomously determining a treatment for a patient, the instructions when executed causing one or more processors to receive image data of an anatomy of a patient; apply the image data to a feature extraction model; receive as output from the feature extraction model scores for each of a plurality of anatomical features; apply the scores as input to a treatment model comprising a second machine learning model different from the first machine learning model trained to output a prediction of a measure of efficacy of a particular treatment based on features of the patient's anatomy; and receive as output from the treatment model data representative of the predicted measure of efficacy of the particular treatment
System for autonomous prediction of treatment efficacy from anatomical feature scores
memory with instructions encoded thereon for autonomously predicting an efficacy of a treatment for a patient; and one or more processors executing the instructions to receive image data of an anatomy of a patient; apply the image data to a feature extraction model comprising a first machine learning model trained using training data pairing anatomical images to an anatomical feature label; receive as output from the feature extraction model scores for each of a plurality of anatomical features corresponding to anatomy of the patient; apply the scores as input to a treatment model comprising a second machine learning model different from the first machine learning model, trained to output a prediction of a measure of efficacy of a particular treatment based on features of the patient's anatomy; and receive as output from the treatment model data representative of the predicted measure of efficacy of the particular treatment
Collectively, the independent claims cover an autonomous prediction pipeline that extracts anatomical feature scores from patient anatomy image data using a first machine learning model trained with anatomical image-to-feature labels, then uses those scores as input to a second, different machine learning model trained to output data representative of predicted treatment efficacy.
Stated Advantages
reduces clinician subjectivity
reduces validation burden
protection against bias by restricting the treatment model from direct image access
Documented Applications
CPAP for sleep apnea
amoxicillin for ear infection
Interested in licensing this patent?