Diagnostic apparatus

• US 7,440,928 B2
• Filed: 08/15/2002
• Issued: 10/21/2008
• Est. Priority Date: 08/15/2001
• Status: Expired due to Fees

• Current Assignee: Metacause Solutions, Ltd.
• Original Assignee: Metacause Solutions, Ltd.
• Inventors: Lewis, Roland W., Ransing, Rajesh S., Ransing, Meghana R.
• Primary Examiner(s): Vincent; David
• Assistant Examiner(s): Fernandez Rivas; Omar F
• Application Number:
  US10/484,673
• Publication Number:
  US 20040249777A1
• Time in Patent Office: 2,259 Days
• Field of Search: 706/1, 706/14, 706/18, 706/20, 706/21, 706 45- 48, 706/50, 706/52, 706/61, 706/903, 706/906, 706/914, 706/12, 700/95, 700/96, 700108-110, 700/115, 700/116, 714/25, 714/26, 714/48
• US Class Current: 706/14
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass

Abstract

Apparatus and method for determining a likely cause or the likelihood of the occurrence of a cause of one or more effects, in which training data relating to previously identified relationships between one or more causes and one or more effects is used to learn the cause and effect relationship. A number of primary and secondary reference points are chosen in the input space created by belief values representing the strength of effect. A function representing the cause and effects relationship and a weight value is associated with each reference point. Weight values associated with primary reference points are considered as independent variables (primary weight values) and other weight values, which are associated with secondary reference points (secondary weight values), depend on one or more primary weight values. Belief value in the occurrence of likely causes of one or more given effects can be determined using this method or apparatus.

Claims

1. A computer-implemented diagnostic system for determining a likelihood of occurrence of a cause of one or more effects occurring in respect of a subject or process, said one or more effects being input to the system, the system being arranged to receive and/or access training data in the form of discrete values relating to previously-identified relationships between one or more effects and said cause, the system comprising:

one or more input nodes arranged to receive respective input signals, the or each input signal being representative of an effect and its relative strength, wherein the number of effects defines the number of dimensional axes of a function representative of said previously-identified relationships;

define an input space on the basis of said dimensional axes and select a number of reference points within said input space, a predefined number of said reference points being designated as primary reference points and the rest of said reference points being designated as secondary reference points, said predefined number being dependent on the number of effects associated with said cause in said training data, wherein each reference point has a predetermined weight value assigned thereto, a weight value being representative of a belief value which quantifies the extent of occurrence of said cause given an effect, the weight values assigned to said primary reference points being independent variables (“

primary weight”

) and the weight values assigned to said secondary reference points being dependent on one or more of said primary weights (“

secondary weight”

), and wherein, in combination, said primary and secondary weights define a multi-dimensional decision hyper-surface representative of said training data;

determine, using said decision hyper-surface, a likelihood of occurrence of said cause given said one or more effects input to the system by;

receiving input data representative of one or more effects occurring in respect of said subject or process and the relative strength thereof;

mapping said input data onto said decision hyper-surface; and

determining from said decision hyper-surface a belief value in said cause given said input data and outputting data representative of said determined likelihood of occurrence of said cause to a recipient mechanism for diagnosing a cause of the occurrence of said one or more effects in respect of said subject or process.

3. A system or method according to claim 1 or claim 2, wherein equidistant reference points are selected along each dimension corresponding to each single effect-cause relationship.

4. A system or method according to claim 1 or claim 2, wherein non-equidistant reference points are selected along one or more of said dimensions, and wherein said non-equidistant reference points are mapped onto a set of equidistant reference points.

5. A system or method according to claim 1 or claim 2, wherein primary weight values are assigned to reference points on dimensional axes of said function(s).

6. A system or method according to claim 5, wherein secondary weight values are assigned to reference points in the input space defined by the said dimensions.

7. A system or method according to claim 1 or claim 2, wherein a p-dimensional Lagrange Interpolation polynomial is constructed in respect of each reference point, where p is the number of effects associated with said cause.

8. A system or method according to claim 7, wherein a said Lagrange Interpolation polynomials are combined to construct a multi-dimensional hyper-surface that is representative of said decision hyper-surface.

9. A system or method according to claim 1 or claim 2, wherein a secondary weight value associated with a reference point is a linear combination of those primary weight values, which are associated with primary reference points corresponding to the co-ordinates of the said reference point.

10. A system or method according to claim 1 or claim 2, wherein a secondary weight value associated with a reference point is a constant multiplied by the average of primary weight values, which are associated with primary reference points corresponding to the co-ordinates of the said reference point.

11. A system or method according to claim 1 or claim 2, wherein the training data is made up of one or more training files, the or each file comprising an input vector storing belief values representing strengths of all associated effects and its corresponding desired output vector storing belief values which quantify the extent of occurrence of the corresponding cause.

12. A system or method according to claim 1 or claim 2 wherein optimal values of primary weights and coefficients used in linear combination are determined, during a training or learning process, using an optimization technique involving minimizing the error in the predicted belief value and the corresponding known belief value as given in the training data.

2. A computer implemented diagnostic method for determining a likelihood of occurrence of a cause of one or more effects occurring in respect of a subject or process, the method comprising:

receiving and/or accessing training data in the form of discrete values relating to previously-identified relationships between one or more effects and said cause;

providing one or more input nodes arranged to receive respective input signals, the or each input signal being representative of an effect and its relative strength, wherein the number of effects defines the number of dimensional axes of a function representative of said previously-identified relationships;

defining an input space on the basis of said dimensional axes and means for selecting a number of reference points within said input space, a predefined number of said reference points being designated as primary reference points and the rest of said reference points being designated as secondary reference points, said predefined number of primary reference points being dependent on the number of effects associated with said cause in said training data, wherein each reference point has a predetermined weight value assigned thereto, a weight value being representative of a belief value which quantifies the extent of occurrence of said cause given an effect, the weight values assigned to said primary reference points being independent variables (“

primary weight”

) and the weight values assigned to said secondary reference points being dependent on one or more of said primary weights (“

secondary weight”

), and wherein, in combination, said primary and secondary weights define a multi-dimensional decision hyper-surface representative of said training data;

determining, using said decision hyper-surface, a likelihood of occurrence of said cause given said one or more effects by;

receiving input data representative of one or more effects occurring in respect of said system or process and the relative strength thereof;

mapping said input data onto said decision hyper-surface; and

determining from said decision hyper-surface a belief value in said cause given said input data;

and

outputting data representative of said determined likelihood of occurrence of said cause to a recipient mechanism for diagnosing a cause of the occurrence of said one or more effects in respect of said subject or process.

13. A computer-implemented diagnostic system for determining a likelihood of occurrence of a cause of one or more effects occurring in respect of a subject or process, said one or more effects being input to the system, the system being arranged to receive and/or access training data in the form of discrete values relating to previously-identified relationships between one or more effects and said cause, the system comprising:

one or more input nodes arranged to receive respective input signals, the or each input signal being representative of an effect and its relative strength, wherein the number of effects defines the number of dimensional axes of a function representative of said previously-identified relationships;

define an input space on the basis of said dimensional axes and select a number of reference points within said input space, a predefined number of said reference points being designated as primary reference points and the rest of said reference points being designated as secondary reference points, said predefined number of primary reference points dependent on the number of effects associated with said cause in said training data, wherein each reference point has a predetermined weight value assigned thereto, a weight value being representative of a belief value which quantifies the extent of occurrence of said cause given an effect, the weight values assigned to said primary reference points being independent variables (“

primary weight”

) and each weight value (“

secondary weight”

) assigned to respective secondary reference points is expressed as a linear combination of one or more of said primary weights, and wherein, in combination, said primary and secondary weights define a multi-dimensional decision hyper-surface representative of said training data;

determine, using said decision hyper-surface, a likelihood of occurrence of said cause given said one or more effects input to the system by;

receiving input data representative of one or more effects occurring in respect of said subject or process and the relative strength thereof;

mapping said input data onto said decision hyper-surface; and

determining from said decision hyper-surface a belief value in said cause given said input data;

and

output data representative of said determined likelihood of occurrence of said cause to a recipient mechanism for diagnosing a cause of the occurrence of said one or more effects in respect of said subject or process.

14. A computer-implemented diagnostic method for determining a likelihood of occurrence of a cause of one or more effects occurring in respect of a subject or process, the method comprising:

receiving and/or accessing training data in the form of discrete values relating to previously-identified relationships between one or more effects and said cause;

providing one or more input nodes arranged to receive respective input signals, the or each input signal being representative of an effect and its relative strength, wherein the number of effects defines the number of dimensional axes of a function representative of said previously-identified relationships;

defining an input space on the basis of said dimensional axes and select a number of reference points within said input space, a predefined number of said reference points being designated as primary reference points and the rest of said reference points being designated as secondary reference points, said predefined number of primary reference points dependent on the number of effects associated with said cause in said training data, wherein each reference point has a predetermined weight value assigned thereto, a weight value being representative of a belief value which quantifies the extent of occurrence of said cause given an effect, the weight values assigned to said primary reference points being independent variables (“

primary weight”

) and each weight value (“

secondary weight”

) assigned to respective secondary reference points is expressed as a linear combination of one or more of said primary weights, and wherein, in combination, said primary and secondary weights define a multi-dimensional decision hyper-surface representative of said training data;

determining, using said decision hyper-surface, a likelihood of occurrence of said cause given said one or more given effects by;

receiving input data representative of one or more effects occurring in respect of said subject or process and the relative strength thereof;

mapping said input data onto said decision hyper-surface; and

determining from said decision hyper-surface a belief value in said cause given said input data;

and

outputting data representative of said determined likelihood of occurrence of said cause to a recipient mechanism for diagnosing a cause of the occurrence of said one or more effects in respect of said subject or process.

15. A computer-implemented diagnostic system for determining a likelihood of occurrence of a cause of one or more effects occurring in respect of a subject or process, said one or more effects being input to the system, the system being arranged to receive and/or access training data in the form of discrete values relating to previously-identified relationships between one or more effects and said cause, the system comprising:

one or more input nodes arranged to receive respective input signals, the or each input signal being representative of an effect and its relative strength, wherein each relative strength value is normalized between 0 and 1 or −

1 and 1 and wherein the number of effects defines the number of dimensional axes of a function representative of said previously-identified relationships;

define an input space on the basis of said dimensional axes and select a number of reference points within said input space, a predefined number of said reference points being designated as primary reference points and the rest of said reference points being designated as secondary reference points, said predefined number of primary reference points dependent on the number of effects associated with said cause in said training data, wherein each reference point has a predetermined weight value assigned thereto, a weight value being representative of a belief value which quantifies the extent of occurrence of said cause given an effect, the weight values assigned to said primary reference points being independent variables (“

primary weight”

) and the weight values assigned to said secondary reference points being dependent on one or more of said primary weights (“

secondary weight”

), and wherein, in combination, said primary and secondary weights define a multi-dimensional decision hyper-surface representative of said training data, said weight values being normalized between 0 and 1 or −

1 and 1;

determine, using said decision hyper-surface, a likelihood of occurrence of said cause given said one or more effects input to the system by;

receiving input data representative of one or more effects occurring in respect of said subject or process and the relative strength thereof;

mapping said input data onto said decision hyper-surface; and

determining from said decision hyper-surface a belief value in said cause given said input data, said belief value being normalized between 0 and 1 or −

1 and 1;

and

output data representative of said determined likelihood of occurrence of said cause to a recipient mechanism for diagnosing a cause of the occurrence of said one or more effects in respect of said subject or process.

16. A computer-implemented diagnostic method for determining a likelihood of occurrence of a cause of one or more effects occurring in respect of a subject or process, the method comprising:

receiving and/or accessing training data in the form of discrete values relating to previously-identified relationships between one or more effects and said cause;

providing one or more input nodes arranged to receive respective input signals, the or each input signal being representative of an effect and its relative strength, wherein each relative strength value is normalized between and 1 or −

1 and 1 and wherein the number of effects defines the number of dimensional axes of a function representative of said previously-identified relationships;

defining an input space on the basis of said dimensional axes and select a number of reference points within said input space, a predefined number of said reference points being designated as primary reference points and the rest of said reference points being designated as secondary reference points, said predefined number of primary reference points being dependent on the number of effects associated with said cause in said training data, wherein each reference point has a predetermined weight value assigned thereto, a weight value being representative of a belief value which quantifies the extent of occurrence of said cause given an effect, the weight values assigned to said primary reference points being independent variables (“

primary weight”

) and the weight values assigned to said secondary reference points being dependent on one or more of said primary weights (“

secondary weight”

), and wherein, in combination, said primary and secondary weights define a multi-dimensional decision hyper-surface representative of said training data, said weight values being normalized between 0 and 1 or −

1 and 1;

determining, using said decision hyper-surface, a likelihood of occurrence of said cause given said one or more effects by;

receiving input data representative of one or more effects occurring in respect of said subject or process and the relative strength thereof;

mapping said input data onto said decision hyper-surface; and

determining from said decision hyper-surface a belief value in said cause given said input data, said belief value being normalized between 0 and 1 or −

1 and 1;

and

outputting data representative of said determined likelihood of occurrence of said cause to a recipient mechanism for diagnosing a cause of the occurrence of said one or more effects in respect of said subject or process.

17. A computer-implemented diagnostic system for determining a likelihood of occurrence of an effect of one or more causes occurring in respect of a subject or process, said one or more causes being input to the system, the system being arranged and configured to receive and/or access training data in the form of discrete values relating to previously-identified relationships between one or more causes and said effect, the system comprising:

one or more input nodes arranged to receive respective input signals, the or each input signal being representative of a cause and its relative strength, wherein the number of causes defines the number of dimensional axes of a function representative of said previously-identified relationships;

define an input space on the basis of said dimensional axes and means for selecting a number of reference points within said input space, a predefined number of said reference points being designated as primary reference points and the rest of said reference points being designated as secondary reference points, said predefined number of primary reference points being dependent on the number of causes associated with said effect in said training data, wherein each reference point has a predetermined weight value assigned thereto, a weight value being representative of a belief value which quantifies the extent of occurrence of in said effect given a cause, the weight values assigned to said primary reference points being independent variables (“

primary weight”

) and the weight values assigned to said secondary reference points being dependent on one or more of said primary weights (“

secondary weight”

), and wherein, in combination, said primary and secondary weights define a multi-dimensional decision hyper-surface representative of said training data;

determine, using said decision hyper-surface, a likelihood of occurrence of said effect given said one or more causes input to the system by;

receiving input data representative of one or more causes occurring in respect of said subject or process and the relative strength thereof;

mapping said input data onto said decision hyper-surface; and

determining from said decision hyper-surface a belief value in said effect given said input data;

and

output data representative of said determined likelihood of occurrence of said effect to a recipient mechanism for diagnosing a cause of the occurrence of said one or more causes in respect of said subject or process.

18. A computer-implemented diagnostic method for determining a likelihood of occurrence of an effect of one or more causes occurring in respect of a subject or process, the method comprising receiving and/or accessing training data in the form of discrete values relating to previously-identified relationships between one or more effects and said cause:

providing one or more input nodes arranged to receive respective input signals, the or each input signal being representative of a cause and its relative strength, wherein the number of causes defines the number of dimensional axes of a function representative of said previously-identified relationships;

defining an input space on the basis of said dimensional axes and selecting a number of reference points within said input space, a predefined number of said reference points being designated as primary reference points and the rest of said reference points being designated as secondary reference points, said predefined number of primary reference points being dependent on the number of causes associated with said effect in said training data, wherein each reference point has a predetermined weight value assigned thereto, a weight value being representative of a belief value which quantifies the extent of occurrence of said effect given a cause, the weight values assigned to said primary reference points being independent variables (“

primary weight”

) and the weight values assigned to said secondary reference points being dependent on one or more of said primary weights (“

secondary weight”

), and wherein, in combination, said primary and secondary weights define a multi-dimensional decision hyper-surface representative of said training data;

determining, using said decision hyper-surface, a likelihood of occurrence of said effect given said one or more causes by;

receiving input data representative of one or more causes occurring in the subject or process and the relative strength thereof;

mapping said input data onto said decision hyper-surface; and

determining from said decision hyper-surface a belief value in said cause given said input data;

and

outputting data representative of said determined likelihood of occurrence of said effect to a recipient mechanism for diagnosing an effect of the occurrence of said one or more causes in respect of said subject or process.

19. A computer-implemented diagnostic system for determining a likelihood of occurrence of a medical condition (cause) of one or more manifestations of the medical condition occurring in respect of a subject or process, said one or more manifestations being input to the system, the system being arranged to receive and/or access training data in the form of discrete values relating to previously-identified relationships between one or more manifestations and said cause, the system comprising:

one or more input nodes arranged to receive respective input signals, the or each input signal being representative of a manifestation and its relative strength, wherein the number of manifestations defines the number of dimensional axes of a function representative of said previously-identified relationships;

define an input space on the basis of said dimensional axes and select a number of reference points within said input space, a predefined number of said reference points being designated as primary reference points and the rest of said reference points being designated as secondary reference points, said predefined number being dependent on the number of manifestations associated with said cause in said training data, wherein each reference point has a predetermined weight value assigned thereto, a weight value being representative of a belief value which quantifies the extent of occurrence of said cause given a manifestation, the weight values assigned to said primary reference points being independent variables (“

primary weight”

) and the weight values assigned to said secondary reference points being dependent on one or more of said primary weights (“

secondary weight”

), and wherein, in combination, said primary and secondary weights define a multi-dimensional decision hyper-surface representative of said training data;

determine, using said decision hyper-surface, a likelihood of occurrence of said cause given said one or more manifestations input to the system by;

receiving input data representative of one or more manifestations occurring in respect of said subject or process and the relative strength thereof;

mapping said input data onto said decision hyper-surface; and

determining from said decision hyper-surface a belief value in said cause given said input data;

and

outputting data representative of said determined likelihood of occurrence of said cause to a recipient mechanism for diagnosing a cause of the occurrence of said one or more manifestations in respect of said subject or process.