System for assessing and mitigating potential spread of infectious disease among dialysis patients
A method for mitigating the spread of infectious diseases among dialysis patients is provided. The method comprises: receiving, by a prediction system and from a medical facility, individual treatment data indicating dialysis treatment information associated with a patient undergoing dialysis treatment; receiving, by the prediction system and from a blood testing laboratory, individual lab data indicating blood analysis information associated with the patient; determining, by the prediction system, disease analysis results for the patient based on inputting the individual treatment data and the individual lab data into a disease prediction machine learning (ML) model, wherein the disease analysis results indicate a likelihood of the patient being infected with a contagious disease; and providing, by the prediction system and to the medical facility, instructions indicating one or more responsive actions based on the disease analysis results.
1. A method, comprising:
training, by a prediction system, a first disease prediction machine learning (ML) model based on geographical disease data indicating new reported cases of a contagious disease within a first geographical area;
receiving, by the prediction system and from a medical facility, individual treatment data indicating dialysis treatment information associated with a patient undergoing an initial dialysis treatment, wherein the medical facility is associated with the first geographical area;
receiving, by the prediction system and from a blood testing laboratory, individual lab data indicating blood analysis information associated with the patient;
determining, by the prediction system, disease analysis results for the patient based on inputting the individual treatment data and the individual lab data into the first disease prediction ML model, wherein the disease analysis results indicate a likelihood of the patient being infected with the contagious disease;
providing, by the prediction system and to the medical facility, instructions indicating one or more responsive actions based on the disease analysis results, wherein the one or more responsive actions comprise lowering an ultrafiltration rate for the patient to use for a subsequent dialysis treatment;
performing, by a dialysis machine located at the medical facility, the subsequent dialysis treatment for the patient using the lowered ultrafiltration rate;
receiving, by the prediction system and from the medical facility, feedback information indicating a plurality of inaccuracy counts associated with the disease analysis results; and
based on comparing the plurality of inaccuracy counts with one or more thresholds, re-training, by the prediction system, the first disease prediction ML model.
2. The method of claim 1 , further comprising:
receiving, by the prediction system, group treatment data indicating dialysis treatment information associated with a plurality of patients undergoing the initial dialysis treatment; and
receiving, by the prediction system, group lab data indicating blood analysis information associated with the plurality of patients undergoing the initial dialysis treatment,
wherein training the first disease prediction ML model is further based on the group treatment data and the group lab data.
3. The method of claim 2 , further comprising:
receiving, by the prediction system, group physician data indicating clinical or treatment notes associated with the plurality of patients undergoing the initial dialysis treatment,
wherein training the first disease prediction ML model is further based on the group physician data.
4. The method of claim 3 , further comprising:
receiving, by the prediction system and from the medical facility, individual physician data indicating clinical or treatment notes associated with the patient undergoing the initial dialysis treatment, and
wherein determining the disease analysis results is further based on inputting the individual physician data into the first disease prediction ML model.
5. The method of claim 2 , further comprising:
obtaining, by the prediction system, group patient data indicating patient demographics and history associated with the plurality of patients undergoing the initial dialysis treatment,
wherein training the first disease prediction ML model is further based on the group patient data.
6. The method of claim 5 , further comprising:
receiving, by the prediction system, individual patient data indicating clinical or treatment notes associated with the patient undergoing the initial dialysis treatment, and
wherein determining the disease analysis results is further based on inputting the individual patient data into the first disease prediction ML model.
7. The method of claim 2 , wherein the group treatment data is associated with the medical facility and the geographical disease data further indicates new reported cases of the contagious disease within a second geographical area, and wherein the method further comprises:
training, by the prediction system and based on the geographical disease data, a second disease prediction ML model for a second medical facility that is within the second geographical area;
determining that the individual treatment data is from the medical facility that is associated with the first geographical area; and
selecting to use the first disease prediction ML model for the patient rather than the second disease prediction ML model based on determining that the individual treatment data is from the medical facility that is associated with the first geographical area.
8. The method of claim 7 , wherein the first disease prediction ML model is a first eXtreme Gradient Boosting (XGBoost) model comprising a plurality of first decision trees and the second disease prediction ML model is a second XGBoost model comprising a plurality of second decision trees that are different from the plurality of first decision trees.
9. The method of claim 8 , wherein training the first disease prediction ML model comprises generating the plurality of first decision trees, wherein training the second disease prediction ML model comprises generating the plurality of second decision trees, and wherein a set of decision trees from the plurality of second decision trees are different from decision trees within the plurality of first decision trees based on the geographical disease data indicating different newly reported cases of the contagious disease within the first geographical area and the second geographical area.
10. The method of claim 1 , wherein the first disease prediction ML model is a deep learning model.
11. The method of claim 1 , wherein the one or more responsive actions further comprise initiating a treatment regimen for the patient, allocating personal protective equipment (PPE) for the medical facility, or adjusting patient scheduling to re-assign the patient to an isolation shift for the subsequent dialysis treatment.
12. A prediction system, comprising:
one or more processors; and
a non-transitory computer-readable medium having processor-executable instructions stored thereon, wherein the processor-executable instructions, when executed by the one or more processors, facilitate:
training a first disease prediction machine learning (ML) model based on geographical disease data indicating new reported cases of a contagious disease within a first geographical area;
receiving, from a medical facility, individual treatment data indicating dialysis treatment information associated with a patient undergoing initial dialysis treatment, wherein the medical facility is associated with the first geographical area;
receiving, from a blood testing laboratory, individual lab data indicating blood analysis information associated with the patient;
determining disease analysis results for the patient based on inputting the individual treatment data and the individual lab data into the first disease prediction ML model, wherein the disease analysis results indicate a likelihood of the patient being infected with the contagious disease;
providing, to the medical facility, instructions indicating one or more responsive actions based on the disease analysis results, wherein the one or more responsive actions comprise lowering an ultrafiltration rate for the patient to use for a subsequent dialysis treatment, wherein a dialysis machine located at the medical facility performs the subsequent dialysis treatment for the patient using the lowered ultrafiltration rate;
receiving, from the medical facility, feedback information indicating a plurality of inaccuracy counts associated with the disease analysis results; and
based on comparing the plurality of inaccuracy counts with one or more thresholds, re-training the first disease prediction ML model.
13. The prediction system of claim 12 , wherein the processor-executable instructions, when executed by the one or more processors, further facilitate:
receiving group treatment data indicating dialysis treatment information associated with a plurality of patients undergoing the initial dialysis treatment; and
receiving group lab data indicating blood analysis information associated with the plurality of patients undergoing the initial dialysis treatment,
wherein training the first disease prediction ML model is further based on the group treatment data and the group lab data.
14. The prediction system of claim 13 , wherein the processor-executable instructions, when executed by the one or more processors, further facilitate:
receiving group physician data indicating clinical or treatment notes associated with the plurality of patients undergoing the initial dialysis treatment,
wherein training the first disease prediction ML model is further based on the group physician data.
15. The prediction system of claim 14 , wherein the processor-executable instructions, when executed by the one or more processors, further facilitate:
receiving, from the medical facility, individual physician data indicating clinical or treatment notes associated with the patient undergoing the initial dialysis treatment, and
wherein determining the disease analysis results is further based on inputting the individual physician data into the first disease prediction ML model.
16. The prediction system of claim 13 , wherein the processor-executable instructions, when executed by the one or more processors, further facilitate:
obtaining group patient data indicating patient demographics and history associated with the plurality of patients undergoing the initial dialysis treatment,
wherein training the first disease prediction ML model is further based on the group patient data.
17. The prediction system of claim 16 , wherein the processor-executable instructions, when executed by the one or more processors, further facilitate:
receiving individual patient data indicating clinical or treatment notes associated with the patient undergoing the initial dialysis treatment, and
wherein determining the disease analysis results is further based on inputting the individual patient data into the first disease prediction ML model.
18. A non-transitory computer-readable medium having processor-executable instructions stored thereon, wherein the processor-executable instructions, when executed, facilitate:
training a first disease prediction machine learning (ML) model based on geographical disease data indicating new reported cases of a contagious disease within a first geographical area;
receiving, from a medical facility, individual treatment data indicating dialysis treatment information associated with a patient undergoing initial dialysis treatment, wherein the medical facility is associated with the first geographical area;
receiving, from a blood testing laboratory, individual lab data indicating blood analysis information associated with the patient;
determining disease analysis results for the patient based on inputting the individual treatment data and the individual lab data into the first disease prediction ML model, wherein the disease analysis results indicate a likelihood of the patient being infected with the contagious disease;
providing, to the medical facility, instructions indicating one or more responsive actions based on the disease analysis results, wherein the one or more responsive actions comprise lowering an ultrafiltration rate for the patient to use for a subsequent dialysis treatment, wherein a dialysis machine located at the medical facility performs the subsequent dialysis treatment for the patient using the lowered ultrafiltration rate;
receiving, from the medical facility, feedback information indicating a plurality of inaccuracy counts associated with the disease analysis results; and
based on comparing the plurality of inaccuracy counts with one or more thresholds, re-training, by the prediction system, the first disease prediction ML model.