Expert system for medical diagnosis

• US 20050015352A1
• Filed: 10/31/2002
• Published: 01/20/2005
• Est. Priority Date: 11/01/2001
• Status: Active Grant

• Current Assignee: Vivici BV
• Original Assignee: Vivici BV
• Inventors: Verlinden, Stefan Frederic Franciscus, Verlinden, Hendrikus Johannes
• Granted Patent: US 7,590,609 B2
• Time in Patent Office: Days
• US Class Current: 706/45
• CPC Class Codes: G06N 5/042 Backward inferencing
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass

Abstract

An expert system, in particular for medical diagnosis, includes a memory 2 for storing a plurality of hypotheses. The hypotheses are arranged in a first and second disjoint group of hypotheses. The memory stores questions for rejecting hypotheses of the second group. An output 3 is used for supplying questions to a user. A input 4 is used for receiving initial data and answers to questions. A processor 5 is programmed to select questions from the stored questions for those hypotheses from the second group that are possible in dependence on the initial data. The processor also determines from answer(s) received in response to outputting the selected questions whether at least one of the hypotheses of the second group is possible. In response to determining that no hypothesis of the second group is possible, the processor supplies a most likely hypothesis of the first group.

Claims

1. An expert system including:

a memory for storing a plurality of hypotheses, where the hypotheses are arranged in at least a first and second disjoint group of hypotheses; and

for storing questions for rejecting hypotheses of the second group;

an output for supplying questions from the memory;

an input for receiving initial data and answers to outputted questions;

a processor programmed to;

select questions from the stored questions for those hypotheses from the second group that are possible in dependence on the initial data;

determine from at least one answer received in response to outputting the selected questions whether at least one of the hypotheses of the second group is possible; and

in response to determining that no hypothesis of the second group is possible, supply via the output a most likely hypothesis of the first group in dependence on at least the initial data.

2. An expert system as claimed in claim 1, wherein the expert system is used for medical diagnosis;

the initial data including a health complaint; and

the processor being programmed to select the questions in dependence on the health complaint.

3. An expert system as claimed in claim 1, wherein the first group of hypotheses includes hypotheses that may be taken by the system and the second group includes hypotheses that may not be taken by the system.

4. An expert system as claimed in claim 1, wherein the processor is programmed to determine whether at least one of the hypotheses of the second group is possible by comparing received answers to a predetermined answer sequence stored in or derivable from the memory.

5. An expert system as claimed in claim 1, wherein the memory stores for each hypothesis of the second group a respective associated question set for rejecting the hypothesis;

the processor being programmed to select all hypotheses from the second group that are possible in dependence on the initial data and to perform the selection of questions by combining question sets of selected hypotheses of the second group.

6. An expert system as claimed in claim 1, wherein the processor is programmed to supply via the output an indication that no hypothesis can be made in response to determining that at least one hypothesis of the second group is possible.

7. An expert system as claimed in claim 1, wherein the processor is programmed to terminate outputting questions in response to determining that at least one hypothesis of the second group is possible.

8. An expert system as claimed in claim 1, wherein the memory stores for at least one hypothesis of the first group a respective associated question set for confirming the hypothesis;

and wherein the processor is programmed to, in response to determining that no hypothesis of the second group is possible, and the most likely hypothesis of the first group is associated with a question set;

output questions of the associated question set;

determine whether the most likely hypothesis meets a predetermined criterion in dependence on at least one further answer; and

provide an outcome of the determination.

9. An expert system as claimed in claim 1, wherein the memory stores for at least one hypothesis of the second group a respective associated further question set for confirming the hypothesis;

and wherein the processor is programmed to, in response to determining that the hypothesis of the second group is possible;

output questions of the associated question set;

determine whether the hypothesis of the second group meets a predetermined criterion in dependence on at least one further answer; and

provide an outcome of the determination.

10. An expert system as claimed in claim 1, wherein the processor is programmed to update the selection of questions during testing of the questions in dependence on one or more received answers.

11. An expert system as claimed in claim 1, wherein the processor is programmed to perform the determining whether a hypothesis of the second group is possible also in dependence on at least one answer received for earlier questions and/or the initial data.

12. An expert system as claimed in claim 1, wherein the processor is programmed to determine a most likely hypothesis of the first group also in dependence on one or more received answers.

13. An expert system as claimed in claim 1, wherein the processor is programmed to transfer control to a human expert.

14. An expert system as claimed in claim 6, wherein the processor is programmed to transfer the control in response to determining that the selected hypothesis meets the predetermined criterion.

15. An expert system as claimed in claim 1, wherein the expert system is used as a medical diagnosis system;

the expert system including a communication interface for exchanging data regarding a medical record of a patient from a storage system for storing medical records of patients;

the processor being programmed to retrieve answers to questions from the medical record and/or store data in the medical record.

17. An expert system as claimed in claim 12, wherein the processor is programmed to transfer the control in response to determining that the selected hypothesis meets the predetermined criterion.

16. A computer program product operative to cause a processor to perform the steps of:

receiving initial data via an input;

selecting from a memory that stores a plurality of hypotheses, where the hypotheses are arranged in at least a first and second disjoint group of hypotheses, and that stores questions for rejecting hypotheses of the second group, questions for those hypotheses from the second group that are possible in dependence on the initial data;

determining from at least one answer received in response to outputting the selected questions whether at least one of the hypotheses of the second group is possible; and

in response to determining that no hypothesis of the second group is possible, supplying via the output a most likely hypothesis of the first group in dependence on at least the initial data.