Automated medical decision making utilizing bayesian network knowledge domain modeling

• US 6,687,685 B1
• Filed: 04/07/2000
• Issued: 02/03/2004
• Est. Priority Date: 04/07/2000
• Status: Expired due to Fees

• Current Assignee: Dr. Red Duke, Inc.
• Original Assignee: Dr. Red Duke, Inc.
• Inventors: Sadeghi, Sarmad, Barzi, Afsaneh, Zarrin-Khameh, Neda
• Primary Examiner(s): Starks, Jr., Wilbert L.
• Assistant Examiner(s): Booker, Kelvin
• Application Number:
  US09/544,953
• Time in Patent Office: 1,397 Days
• Field of Search: 706/15, 706/45, 706/52, 706/59, 706/924, 706/61, 706/20, 706/60, 706/48, 706/50, 600/300
• US Class Current: 706/15
• CPC Class Codes: G16H 10/20 for electronic clinical trials or questionnaires
• CPC Class Codes: G16H 40/67 for remote operation
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: Y10S 706/924 Medical

Abstract

The present invention relates to a system and method of medical knowledge domain modeling and automated medical decision-making, such as for online, questionnaire-based medical triage. In the present invention, information such as conditions and characteristics related to a diagnosis or disposition level is modeled in a Bayesian Network. The Bayesian Network may comprise instantiable nodes, fault nodes, intermediary nodes, a utility node and a decision node. Using Bayesian inference, the conditional probability of any pair in the network may be determined in real-time. These conditional probabilities are modified upon the input of evidence, which is typically in the form of answers to a dynamic set of questions designed to identify a diagnosis or disposition level for the patient under evaluation.

Claims

1. A method for providing automated medical decision-making comprising:

listing differential diagnoses for each of a plurality of conditions;

listing the characteristics of each differential diagnosis;

establishing a Bayesian Network modeling knowledge pertaining to said conditions, said differential diagnoses, and said characteristics, wherein the presence or absence of at least some of said conditions affects conditional probabilities that a patient under study has at least some of said diagnoses;

generating questions to identify said presence or absence of said conditions;

presenting said patient under study with a subset of said questions for the purpose of generating responses, wherein said subset is selected dynamically by favoring questions which have a significant statistical impact on said conditional probabilities;

inputting said responses to said questions into said Bayesian Network; and

calculating a diagnosis for the patient under study utilizing said Bayesian Network.

2. The method for providing automated medical decision-making of claim 1, wherein said Bayesian Network further comprises a plurality of instantiable nodes for entering evidence into said Bayesian Network.

3. The method for providing automated medical decision-making of claim 2, wherein said Bayesian Network further comprises a plurality of fault nodes to which at least some of said evidence is propagated.

4. The method for providing automated medical decision-making of claim 3, wherein said Bayesian Network further comprises a utility node.

5. The method for providing automated medical decision-making of claim 4, wherein said Bayesian Network further comprises a decision node.

6. The method for providing automated medical decision-making of claim 5, wherein said decision node generates a diagnosis for said patient under study.

7. The method for providing automated medical decision-making of claim 5, wherein said decision node generates a disposition level for said patient under study.

8. The method for providing automated medical decision-making of claim 5, wherein said decision node generates a recommended treatment for said patient under study.

9. The method for providing automated medical decision-making of claim 5, wherein said decision node generates a prognosis for said patient under study.

10. The method for providing automated medical decision-making of claim 5, wherein said Bayesian Network further comprises an intermediary node.

11. The method for providing automated medical decision-making of claim 10, wherein said intermediary node relates to a risk factor associated with a medical diagnosis.

12. The method for providing automated medical decision-making of claim 10, wherein said intermediary node is configured to reduce the number of inputs into said fault node.

13. The method for providing automated medical decision-making of claim 5, wherein said decision node generates a prognosis for said patient under study.

14. A computerized system for providing automated medical triage to a patient under study comprising:

a server including processing means and memory means, said memory means comprising a Bayesian Network, wherein said Bayesian Network comprises a plurality of instantiable nodes, a plurality of fault nodes, a utility node and a decision node;

a remote computer accessible by a patient under study;

a network allowing access by said remote computer to said server;

a scoring mechanism to adjust said probabilities in said diagnosis node responsive to information input by said patient under study into said remote computer.

15. The computerized system for providing automated medical triage of claim 14, further comprising:

medical hardware in direct electrical communication with said remote computer, wherein said medical hardware provides additional information regarding said patient under study, said additional information being transmitted to said server via said network.

16. A method for providing automated medical decision-making comprising:

establishing a Bayesian Network comprising;

a plurality of instantiable nodes for entering evidence;

a plurality of fault nodes in communication with said instantiable nodes for providing results capable of being monitored;

a utility node for providing quantitative information to a decision node having states;

and said decision node, wherein said decision node calculates a utility value for each of said states;

obtaining evidence from a patient-under-study;

entering said evidence into selected ones of said instantiable nodes;

monitoring said fault nodes for,said results;

using said results in said utility node; and

outputting a result in said decision node based upon said quantitative information in said utility node.

17. The method for providing automated medical decision-making of claim 16, wherein said decision node generates a diagnosis for said patient under study.

18. The method for providing automated medical decision-making of claim 16, wherein said decision node generates a disposition level for said patient under study.

19. The method for providing automated medical decision-making of claim 16, wherein said decision node generates a recommended treatment for said patient under study.

20. The method for providing automated medical decision-making of claim 16, wherein said Bayesian Network further comprises an intermediary node.

21. The method for providing automated medical decision-making of claim 20, wherein said intermediary node relates to a risk factor associated with a medical diagnosis.

22. The method for providing automated medical decision-making of claim 20, wherein said intermediary node is configured to reduce the number of inputs into said fault node.

23. The method for providing automated medical decision-making of claim 16, wherein said step of establishing a Bayesian Network further comprises identifying a syndrome and configuring probability tables to recognize said syndrome.