Machine learning analysis techniques for clinical and patient data
Systems and methods are disclosed for analyzing data from oncology treatments such as immune checkpoint inhibitor or radiotherapy therapies, including predicting adverse events of the oncology therapies, predicting objective response of the oncology therapies, predicting symptoms from the oncology therapies, and use of such predictions by technological implementations to achieve improved system and medical outcomes. An example technique for generating a predicted treatment outcome includes: receiving patient data for a human subject, which provides patient-reported outcomes collected from the human subject relating to a particular oncology treatment; processing the patient data with a trained artificial intelligence (AI) prediction model, which receives the patient data as input and produces a prediction of a treatment outcome as output; and outputting data to modify a treatment workflow of an oncology treatment for the human subject, based on the prediction of the treatment outcome.
1. A method for generating a predicted treatment outcome of an oncology treatment for a human subject, the method comprising:
receiving patient data for the human subject, the patient data including patient-reported outcomes relating to the oncology treatment that are collected from the human subject;
processing the patient data with a trained artificial intelligence (AI) prediction model, the trained AI prediction model configured to receive the patient data as an input and to produce a prediction of a treatment outcome for the human subject as an output; and
outputting data, based on the prediction of the treatment outcome, to either: (i) modify a treatment workflow of the oncology treatment for the human subject, or (ii) recalculate a predicted dose delivery or efficacy of the oncology treatment for the human subject;
wherein the prediction of the treatment outcome includes a prediction of one or more adverse events, and wherein the prediction of each respective adverse event includes:
a probability of an occurrence of the respective adverse event; and
a timing or a severity of the respective adverse event, if the probability of the occurrence of the respective adverse event exceeds a defined amount.
2. The method of claim 1 , wherein the one or more adverse events are radiotherapy adverse events, wherein the oncology treatment is a radiotherapy treatment, and wherein the data to modify the treatment workflow includes a command to change a plan used for delivering the radiotherapy treatment to the human subject based on the radiotherapy adverse events.
3. The method of claim 2 , wherein a timing, a dosage, or a location of the radiotherapy treatment, to be delivered with the plan, is changed based on the prediction of the radiotherapy adverse events.
4. The method of claim 1 , wherein the prediction of one or more adverse events is based on the processing of the patient data with the trained AI prediction model, using at least one subset of patient data including data related to:
at least one physical characteristic independent of the oncology treatment, provided from among: ethnicity, weight or other physical characteristics, medical history, age, sex, diet, genetics, medication, or fitness;
discretionary clinical treatment plan preferences;
delivery equipment;
patient demographics; or
clinical demographics.
5. The method of claim 1 , wherein the one or more adverse events are immune-related adverse events, wherein the oncology treatment is an immune checkpoint inhibitor therapy, and wherein the data to modify the treatment workflow includes a command to change an amount or a timing of an immunotherapy treatment delivered to the human subject with the immune checkpoint inhibitor therapy.
6. The method of claim 1 , wherein the prediction of the treatment outcome includes a prediction of an objective response rate of the human subject to the oncology treatment, and wherein the prediction of the objective response rate includes a classification of a complete response or an amount of a partial response to the oncology treatment.
7. The method of claim 1 , wherein the patient-reported outcomes are provided from structured data collected in a questionnaire, and wherein the questionnaire provides a series of questions that are customized to the human subject.
8. The method of claim 7 , wherein the patient-reported outcomes are also provided from unstructured data collected in one or more text inputs of the questionnaire.
9. The method of claim 1 , wherein the patient data further includes one or more of:
clinical information of the human subject;
laboratory data from one or more specimens collected from the human subject;
treatment information from prior sessions of the oncology treatment delivered to the human subject;
measurements from one or more wearable devices used by the human subject;
measurements from one or more medical monitoring devices external to the human subject; or
event data from prior occurrence of adverse events by the human subject.
10. The method of claim 9 , further comprising:
verifying performance of the trained AI prediction model after training, and before use with the patient data, based on metrics including one or more of: accuracy, precision and recall, or a correlation coefficient.
11. The method of claim 1 , wherein the trained AI prediction model uses an extreme gradient boosting supervised machine learning algorithm.
12. The method of claim 1 , wherein the trained AI prediction model is trained with training data that is specific to the human subject and a type of the oncology treatment, and wherein the patient data is collected between treatment sessions of the oncology treatment.
13. The method of claim 1 , wherein processing the patient data with the trained AI prediction model includes use of multiple AI prediction models to produce the output, and wherein each of the multiple AI prediction models is customized to a respective symptom or respective outcome associated with the oncology treatment.
14. The method of claim 1 , further comprising:
outputting information related to the treatment outcome to the human subject or a clinician associated with the human subject, based on the prediction of the treatment outcome, wherein the information includes one or more of: an alert, educational content, or a recommendation.
15. A method of dynamically adapting a radiotherapy treatment plan having multiple fractions, based on a predicted treatment outcome of an oncology treatment for a human subject, the method comprising:
developing a treatment workflow for the oncology treatment for the human subject based on clinically determined expected outcomes of such treatment;
generating the predicted treatment outcome of the oncology treatment for the human subject;
receiving intra-fraction patient data for the human subject, the patient data including patient-reported outcomes relating to the oncology treatment that are collected from the human subject;
processing the patient data with a trained artificial intelligence (AI) prediction model, the trained AI prediction model configured to receive the intra-fraction patient data as an input and to produce a prediction of a treatment outcome for the human subject as an output;
comparing the predicted treatment outcome to an expected treatment outcome; and
changing the treatment workflow based on the comparison of the predicted treatment outcome, in response to determining that a difference between the predicted treatment outcome and the expected treatment outcome is outside of a predetermined tolerance.
16. The method of claim 15 , wherein the patient-reported outcomes are utilized as an input for changing the treatment workflow.
17. A method of monitoring efficacy of a treatment plan of a radiotherapy treatment for a human subject comprising:
processing the treatment plan with a trained artificial intelligence (AI) prediction model, the trained AI prediction model configured to receive data for the treatment plan as an input and to produce an adverse effect prediction report of potential adverse patient-reported outputs associated with the treatment plan indicative of ineffective treatment for the human subject as an output;
receiving patient data for the human subject, the patient data including patient-reported outcomes relating to the treatment plan that are collected from the human subject; and
monitoring for predicted adverse patient-reported outputs and, where identified, outputting data indicative that the treatment plan may require adjustment.
18. The method of claim 17 , wherein the output data is used to determine whether delivery efficacy of the treatment plan is outside of acceptable parameters.
19. The method of claim 17 , wherein the output data includes data to modify a treatment workflow of the radiotherapy treatment for the human subject, based on a prediction of one or more outcomes of the radiotherapy treatment.
20. A non-transitory computer-readable storage medium comprising computer-readable instructions for generating a predicted treatment outcome of an oncology treatment for a human subject, wherein the instructions, when executed, cause a computing machine to perform operations comprising:
receiving patient data for the human subject, the patient data including patient-reported outcomes relating to the oncology treatment that are collected from the human subject;
processing the patient data with a trained artificial intelligence (AI) prediction model, the trained AI prediction model configured to receive the patient data as an input and to produce a prediction of a treatment outcome for the human subject as an output; and
outputting data to modify a treatment workflow of the oncology treatment for the human subject, based on the prediction of the treatment outcome;
wherein the prediction of the treatment outcome includes a prediction of one or more adverse events, and wherein the prediction of each respective adverse event includes:
a probability of an occurrence of the respective adverse event; and
a timing or a severity of the respective adverse event, if the probability of the occurrence of the respective adverse event exceeds a defined amount.
21. The computer-readable storage medium of claim 20 , wherein the one or more adverse events are radiotherapy adverse events, wherein the oncology treatment is a radiotherapy treatment, and wherein the data to modify the treatment workflow includes a command to change a plan used for delivering the radiotherapy treatment to the human subject based on the radiotherapy adverse events.
22. The computer-readable storage medium of claim 21 , wherein a timing, a dosage, or a location of the radiotherapy treatment, to be delivered with the plan, is changed based on the prediction of the radiotherapy adverse events.
23. The computer-readable storage medium of claim 20 , wherein the one or more adverse events are immune-related adverse events, wherein the oncology treatment is an immune checkpoint inhibitor therapy, and wherein the data to modify the treatment workflow includes a command to change an amount or a timing of an immunotherapy treatment delivered to the human subject with the immune checkpoint inhibitor therapy.
24. The computer-readable storage medium of claim 20 , wherein the prediction of the treatment outcome includes a prediction of an objective response rate of the human subject to the oncology treatment, and wherein the prediction of the objective response rate includes a classification of a complete response or an amount of a partial response to the oncology treatment.
25. The computer-readable storage medium of claim 20 , wherein the patient-reported outcomes are provided from structured data collected in a questionnaire, and wherein the questionnaire provides a series of questions that is customized to the human subject.
26. The computer-readable storage medium of claim 25 , wherein the patient-reported outcomes are also provided from unstructured data collected in one or more text inputs of the questionnaire.
27. The computer-readable storage medium of claim 20 , wherein the patient data further includes one or more of:
clinical information of the human subject;
laboratory data from one or more specimens collected from the human subject;
treatment information from prior sessions of the oncology treatment delivered to the human subject;
measurements from one or more wearable devices used by the human subject;
measurements from one or more medical monitoring devices external to the human subject; or
event data from prior occurrence of adverse events by the human subject.
28. The computer-readable storage medium of claim 20 , wherein the trained AI prediction model uses an extreme gradient boosting supervised machine learning algorithm.
29. The computer-readable storage medium of claim 28 , wherein the instructions further cause the computing machine to perform operations comprising:
verifying performance of the trained AI prediction model after training, and before use with the patient data, based on metrics including one or more of: accuracy, precision and recall, or a correlation coefficient.
30. The computer-readable storage medium of claim 20 , wherein the trained AI prediction model is trained with training data that is specific to the human subject and a type of the oncology treatment, and wherein the patient data is collected between treatment sessions of the oncology treatment.
31. The computer-readable storage medium of claim 20 , wherein processing the patient data with the trained Al prediction model includes use of multiple AI prediction models to produce the output, and wherein each of the multiple AI prediction models is customized to a respective symptom or respective outcome associated with the oncology treatment.
32. The computer-readable storage medium of claim 20 , wherein the instructions further cause the computing machine to perform operations comprising:
outputting information related to the treatment outcome to the human subject or a clinician associated with the human subject, based on the prediction of the treatment outcome, wherein the information includes one or more of:
an alert, educational content, or a recommendation.