Method and system for assisting determination of a medical condition

• US 9,977,958 B2
• Filed: 02/26/2015
• Issued: 05/22/2018
• Est. Priority Date: 02/26/2014
• Status: Active Grant

• Current Assignee: Grain Ip Co.
• Original Assignee: Grain Ip Co.
• Inventors: Van Hoe, Lieven
• Primary Examiner(s): Goradia, Shefali
• Application Number:
  US15/119,849
• Publication Number:
  US 20170220860A1
• Time in Patent Office: 1,181 Days
• CPC Class Codes: G06V 2201/03 Recognition of patterns in medical or anatomical images
• CPC Class Codes: G06V 2201/10 Recognition assisted with metadata
• CPC Class Codes: G06V 30/414 Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text
• CPC Class Codes: G16H 10/20 for electronic clinical trials or questionnaires
• CPC Class Codes: G16H 30/20 for handling medical images, e.g. DICOM, HL7 or PACS
• 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

The present invention relates to a method for assisting a user in determining a medical condition in a subject from one or more medical images by presenting the user with a list of questions, LOQ, about the subject that are to be answered with an affirmative indication (Y), a negative indication (N) or an unknown indication (X) by input by the user. In response to receiving an answer to one question, the method presents the user with a ranked list of conditions, LOC, and also with an updated LOQ wherein at least the answered question has been removed i.e. the LOQ iteratively collapses or reduces in length. An earlier step of selecting a foundation observation, FO, about the subject reduces the initial LOQ and the possible conditions.

Claims

1. A method for assisting a user in determining a medical condition in a subject from one or more medical images, the method comprising the steps:

i) outputting to the user a list of foundation observations, LFO, from a database, wherein the database contains;

a multitude of observations, MO, each corresponding to a description of a medical-image- or subject-related observation,

a set of foundation observations, SFO, corresponding to a predefined subset of the MO, and used to generate the LFO,

a multitude of conditions, MC, each linked to at least one foundation observation,

a multitude of rules, MR,

wherein each rule in the MR links a discrete likelihood of at least one condition with the absence or presence of at least one observation,

ii) receiving an input of a selection of at least one FO from the LFO,

iii) generating;

a set of conditions, SC, from the MC, wherein conditions not linked to the selected FO are excluded,

a set of rules, SR, from the MR, wherein rules that are not linked to the presence of the selected FO are excluded, and

a set of observations, SO, from the MO, wherein observations not linked to rules present in the SR are excluded,

wherein a plurality of conditions in the SC is linked to a plurality of foundation observations in the SO by discrete a priori ratings, APRs, wherein each foundation observation is linked to the corresponding condition by one APR,

iv) generating and outputting a list of conditions LOC from the SC, and a list of questions, LOQ from the SO, wherein each condition in the LOC is initially assigned the corresponding APR,

v) receiving an input as to the status of one question in the LOQ whereby the status is an indication of the presence, absence or unknown status in regard of the subject,

vi) updating and outputting;

the LOC, wherein likelihood(s) for condition(s) linked to the rules satisfied by status received in step (v) are provided, the likelihood(s) replacing the APR initially assigned to the same condition in step (iv) and wherein a satisfied rule indicates one or more specified conditions each at a specified likelihood on the one hand, and a specified absence or presence of one or more specified observations on the other hand, and

the LOQ, wherein at least the question corresponding to the observation status received in step (v) is removed,

vii) repeating steps v) to vi) such that the LOQ iteratively reduces in length optionally until there are no more questions in the LOQ.

2. The method according to claim 1, wherein the number of categories of discrete a priori ratings is between 3 and 5.

3. The method according to claim 1, wherein the number of discrete likelihoods is between 4 and 8.

4. A method for assisting a user in determining a medical condition in a subject from one or more medical images, the method comprising the steps:

i) outputting to the user a list of foundation observations, LFO, from a database, wherein the database contains;

a multitude of observations, MO, each corresponding to a description of a medical-image- or subject-related observation,

a set of foundation observations, SFO, corresponding to a predefined subset of the MO, and used to generate the LFO,

a multitude of conditions, MC, each linked to at least one foundation observation,

a multitude of rules, MR,

wherein each rule in the MR links a discrete likelihood of at least one condition with the absence or presence of at least one observation,

ii) receiving an input of a selection of at least one FO from the LFO,

iii) generating;

a set of conditions, SC, from the MC, wherein conditions not linked to the selected FO are excluded,

a set of rules, SR, from the MR, wherein rules that are not linked to the presence of the selected FO are excluded, and

a set of observations, SO, from the MO, wherein observations not linked to rules present in the SR are excluded,

wherein a plurality of conditions in the SC is linked to a plurality of foundation observations in the SO by discrete a priori ratings, APRs, wherein each foundation observation is linked to the corresponding condition by one APR,

iv) generating and outputting a list of conditions LOC from the SC, and a list of questions, LOQ from the SO,

v) receiving an input as to the status of one question in the LOQ whereby the status is an indication of the presence, absence or unknown status in regard of the subject,

vi) updating and outputting;

the LOC, wherein likelihood(s) for condition(s) linked to the rules satisfied by status received in step (v) are provided, the likelihood(s) replacing the APR initially assigned to the same condition in step (iv), and

the LOQ, wherein at least the question corresponding to the observation status received in step (v) is removed,

vii) repeating steps v) to vi) such that the LOQ iteratively reduces in length optionally until there are no more questions in the LOQ,

wherein each condition in the LOC is initially assigned the corresponding APR, and wherein the conditions in the LOC are ordered according to likelihood.

5. The method according to claim 4, wherein the number of categories of discrete a priori ratings is between 3 and 5.

6. The method according to claim 4, wherein the number of discrete likelihoods is between 4 and 8.

7. A method for assisting a user in determining a medical condition in a subject from one or more medical images, the method comprising the steps:

i) outputting to the user a list of foundation observations, LFO, from a database, wherein the database contains;

a multitude of observations, MO, each corresponding to a description of a medical-image- or subject-related observation,

a set of foundation observations, SFO, corresponding to a predefined subset of the MO, and used to generate the LFO, wherein the observations in the SFO are pre-defined patterns in the one or more medical images,

a multitude of conditions, MC, each linked to at least one foundation observation,

a multitude of rules, MR,

wherein each rule in the MR links a discrete likelihood of at least one condition with the absence or presence of at least one observation,

ii) receiving an input of a selection of at least one FO from the LFO,

iii) generating;

a set of conditions, SC, from the MC, wherein conditions not linked to the selected FO are excluded,

a set of rules, SR, from the MR, wherein rules that are not linked to the presence of the selected FO are excluded, and

a set of observations, SO, from the MO, wherein observations not linked to rules present in the SR are excluded,

wherein a plurality of conditions in the SC is linked to a plurality of foundation observations in the SO by discrete a priori ratings, APRs, wherein each foundation observation is linked to the corresponding condition by one APR,

iv) generating and outputting a list of conditions LOC from the SC, and a list of questions, LOQ from the SO, wherein each condition in the LOC is initially assigned the corresponding APR,

v) receiving an input as to the status of one question in the LOQ whereby the status is an indication of the presence, absence or unknown status in regard of the subject,

vi) updating and outputting;

the LOC, wherein likelihood(s) for condition(s) linked to the rules satisfied by status received in step (v) are provided, the likelihood(s) replacing the APR initially assigned to the same condition in step (iv), and

the LOQ, wherein at least the question corresponding to the observation status received in step (v) is removed,

vii) repeating steps v) to vi) such that the LOQ iteratively reduces in length optionally until there are no more questions in the LOQ.

8. The method according to claim 7, wherein the number of categories of discrete a priori ratings is between 3 and 5.

9. The method according to claim 7, wherein the number of discrete likelihoods is between 4 and 8.

10. A method for assisting a user in determining a medical condition in a subject from one or more medical images, the method comprising the steps:

i) outputting to the user a list of foundation observations, LFO, from a database, wherein the database contains;

a multitude of observations, MO, each corresponding to a description of a medical-image- or subject-related observation,

a set of foundation observations, SFO, corresponding to a predefined subset of the MO, and used to generate the LFO,

a multitude of conditions, MC, each linked to at least one foundation observation,

a multitude of rules, MR,

wherein each rule in the MR links a discrete likelihood of at least one condition with the absence or presence of at least one observation,

ii) receiving an input of a selection of at least one FO from the LFO,

iii) generating;

a set of conditions, SC, from the MC, wherein conditions not linked to the selected FO are excluded,

a set of rules, SR, from the MR, wherein rules that are not linked to the presence of the selected FO are excluded, and

a set of observations, SO, from the MO, wherein observations not linked to rules present in the SR are excluded,

wherein a plurality of conditions in the SC is linked to a plurality of foundation observations in the SO by discrete a priori ratings, APRs, wherein each foundation observation is linked to the corresponding condition by one APR,

iv) generating and outputting a list of conditions LOC from the SC, and a list of questions, LOQ from the SO, wherein each condition in the LOC is initially assigned the corresponding APR,

v) receiving an input as to the status of one question in the LOQ whereby the status is an indication of the presence, absence or unknown status in regard of the subject,

vi) updating and outputting;

the LOC, wherein likelihood(s) for condition(s) linked to the rules satisfied by status received in step (v) are provided, the likelihood(s) replacing the APR initially assigned to the same condition in step (iv), and

the LOQ, wherein at least the question corresponding to the observation status received in step (v) is removed,

vii) repeating steps v) to vi) such that the LOQ iteratively reduces in length optionally until there are no more questions in the LOQ,

wherein two or more observations in the SO constitute a subset of observations, SSO, such that the presence or absence of one observation of the SSO is indicative of the presence or absence of the remaining observations in the SSO.

11. Method according to claim 10, wherein the updated LOQ in step vi) is devoid of all questions enquiring as to the status of the observations in the SSO, when the answer received in step ii) is an indication of the status of one observation of the SSO.

12. The method according to claim 10, wherein the number of categories of discrete a priori ratings is between 3 and 5.

13. The method according to claim 10, wherein the number of discrete likelihoods is between 4 and 8.

14. A method for assisting a user in determining a medical condition in a subject from one or more medical images, the method comprising the steps:

i) outputting to the user a list of foundation observations, LFO, from a database, wherein the database contains;

a multitude of observations, MO, each corresponding to a description of a medical-image- or subject-related observation,

a set of foundation observations, SFO, corresponding to a predefined subset of the MO, and used to generate the LFO,

a multitude of conditions, MC, each linked to at least one foundation observation,

a multitude of rules, MR,

wherein each rule in the MR links a discrete likelihood of at least one condition with the absence or presence of at least one observation,

ii) receiving an input of a selection of at least one FO from the LFO,

iii) generating;

a set of conditions, SC, from the MC, wherein conditions not linked to the selected FO are excluded,

a set of rules, SR, from the MR, wherein rules that are not linked to the presence of the selected FO are excluded, and

a set of observations, SO, from the MO, wherein observations not linked to rules present in the SR are excluded,

wherein a plurality of conditions in the SC is linked to a plurality of foundation observations in the SO by discrete a priori ratings, APRs, wherein each foundation observation is linked to the corresponding condition by one APR,

iv) generating and outputting a list of conditions LOC from the SC, and a list of questions, LOQ from the SO, wherein each condition in the LOC is initially assigned the corresponding APR,

v) receiving an input as to the status of one question in the LOQ whereby the status is an indication of the presence, absence or unknown status in regard of the subject,

vi) updating and outputting;

the LOC, wherein likelihood(s) for condition(s) linked to the rules satisfied by status received in step (v) are provided, the likelihood(s) replacing the APR initially assigned to the same condition in step (iv), and

the LOQ, wherein at least the question corresponding to the observation status received in step (v) is removed,

vii) repeating steps v) to vi) such that the LOQ iteratively reduces in length optionally until there are no more questions in the LOQ,

wherein two or more observations in the SO constitute a linked set of observations, LSO, associated to a status of a linking observation, and the LSO is used to generate a corresponding linked set of questions LSOQ enquiring as to the status of the LSO, such that an indication of a status of the linking observation initiates outputting the LSOQ enquiring as to the status observations of the LSO.

15. The method according to claim 14, wherein the number of categories of discrete a priori ratings is between 3 and 5.

16. The method according to claim 14, wherein the number of discrete likelihoods is between 4 and 8.

17. A method for assisting a user in determining a medical condition in a subject from one or more medical images, the method comprising the steps:

i) outputting to the user a list of foundation observations, LFO, from a database, wherein the database contains;

a multitude of observations, MO, each corresponding to a description of a medical-image- or subject-related observation,

a set of foundation observations, SFO, corresponding to a predefined subset of the MO, and used to generate the LFO,

a multitude of conditions, MC, each linked to at least one foundation observation,

a multitude of rules, MR,

wherein each rule in the MR links a discrete likelihood of at least one condition with the absence or presence of at least one observation,

ii) receiving an input of a selection of at least one FO from the LFO,

iii) generating;

a set of conditions, SC, from the MC, wherein conditions not linked to the selected FO are excluded,

a set of rules, SR, from the MR, wherein rules that are not linked to the presence of the selected FO are excluded, and

a set of observations, SO, from the MO, wherein observations not linked to rules present in the SR are excluded,

wherein a plurality of conditions in the SC is linked to a plurality of foundation observations in the SO by discrete a priori ratings, APRs, wherein each foundation observation is linked to the corresponding condition by one APR (FIG. 7 A2),

iv) generating and outputting a list of conditions LOC from the SC, and a list of questions, LOQ from the SO, wherein each condition in the LOC is initially assigned the corresponding APR,

v) receiving an input as to the status of one question in the LOQ whereby the status is an indication of the presence, absence or unknown status in regard of the subject,

vi) updating and outputting;

the LOC, wherein likelihood(s) for condition(s) linked to the rules satisfied by status received in step (v) are provided, the likelihood(s) replacing the APR initially assigned to the same condition in step (iv), and

the LOQ, wherein at least the question corresponding to the observation status received in step (v) is removed,

vii) repeating steps v) to vi) such that the LOQ iteratively reduces in length optionally until there are no more questions in the LOQ,

wherein step ii) further comprises presenting the user with a series of choices, arranged in increasing specificity, to guide the user towards a selection of an indication of the FO present in the patient.

18. The method according to claim 17, wherein the number of categories of discrete a priori ratings is between 3 and 5.

19. The method according to claim 17, wherein the number of discrete likelihoods is between 4 and 8.

20. A method for assisting a user in determining a medical condition in a subject from one or more medical images, the method comprising the steps:

i) outputting to the user a list of foundation observations, LFO, from a database, wherein the database contains;

a multitude of observations, MO, each corresponding to a description of a medical-image- or subject-related observation,

a set of foundation observations, SFO, corresponding to a predefined subset of the MO, and used to generate the LFO,

a multitude of conditions, MC, each linked to at least one foundation observation,

a multitude of rules, MR,

wherein each rule in the MR links a discrete likelihood of at least one condition with the absence or presence of at least one observation,

ii) receiving an input of a selection of at least one FO from the LFO,

iii) generating;

a set of conditions, SC, from the MC, wherein conditions not linked to the selected FO are excluded,

a set of rules, SR, from the MR, wherein rules that are not linked to the presence of the selected FO are excluded, and

a set of observations, SO, from the MO, wherein observations not linked to rules present in the SR are excluded,

wherein a plurality of conditions in the SC is linked to a plurality of foundation observations in the SO by discrete a priori ratings, APRs, wherein each foundation observation is linked to the corresponding condition by one APR,

iv) generating and outputting a list of conditions LOC from the SC, and a list of questions, LOQ from the SO, wherein each condition in the LOC is initially assigned the corresponding APR,

v) receiving an input as to the status of one question in the LOQ whereby the status is an indication of the presence, absence or unknown status in regard of the subject,

vi) updating and outputting;

the LOC, wherein likelihood(s) for condition(s) linked to the rules satisfied by status received in step (v) are provided, the likelihood(s) replacing the APR initially assigned to the same condition in step (iv), and

the LOQ, wherein at least the question corresponding to the observation status received in step (v) is removed,

vii) repeating steps v) to vi) such that the LOQ iteratively reduces in length optionally until there are no more questions in the LOQ,

wherein

the medical image-related observations comprise the recognition of a morphological pattern from a medical image and/or

the patient-related observations comprise age, gender, symptoms, medical history, laboratory results, or a combination thereof.

21. The method according to claim 20, wherein the number of categories of discrete a priori ratings is between 3 and 5.

22. The method according to claim 20, wherein the number of discrete likelihoods is between 4 and 8.