Predicting diagnosis of a patient

• US 8,996,428 B2
• Filed: 01/17/2012
• Issued: 03/31/2015
• Est. Priority Date: 01/17/2012
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Baras, Dorit, Farkash, Ariel, Vitkin, Edward
• Primary Examiner(s): Chaki, Kakali
• Assistant Examiner(s): Seck, Ababacar
• Application Number:
  US13/351,243
• Publication Number:
  US 20130185231A1
• Time in Patent Office: 1,169 Days
• Field of Search: None
• US Class Current: 706/12
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders

Abstract

Method, system, and computer program product are provided for predicting diagnosis of a patient performed by a computerized device. The method may include: modeling data from a group of successfully diagnosed patients, wherein the data is modeled as treatment paths of patients including referrals to medical practitioners; and predicting diagnosis for a current patient by comparing a treatment path of the current patient with the modeled treatment paths of successfully diagnosed patients, including calculating a probability of a given diagnosis from the modeled treatment paths. The method may include: defining a set of medical entities including medical practitioners to which a patient has been referred; and gathering treatment paths of successfully diagnosed patients, wherein the treatment path links medical entities in a directional route. Predicting diagnosis for a current patient may use the modeled data to calculate the probability of each model instance for each diagnosis and choosing the model instance of the diagnosis that maximizes the treatment path probability.

Claims

1. A computer-implemented method for predicting diagnosis of a patient performed by a computerized device having a processor, comprising:

modeling data from a group of patients successfully diagnosed as having one or more medical conditions, wherein the data is modeled as treatment paths of patients including referrals to medical practitioners; and

predicting diagnosis of one or more medical conditions for a current patient by comparing a treatment path of the current patient with the modeled treatment paths of successfully diagnosed patients, including calculating a probability of a given diagnosis of one or more medical conditions from the modeled treatment paths, wherein the calculating is in accordance with;

2. The method as claimed in claim 1, wherein modeling data includes:

defining a set of medical entities including medical practitioners to which a patient has been referred; and

gathering treatment paths of successfully diagnosed patients, wherein the treatment path links medical entities in a directional route.

3. The method as claimed in claim 2, wherein the medical entities include one or more of the group of:

medical practitioners visited;

medical practitioners referred to;

tests carried out or recommended;

medications prescribed;

therapies carried out or recommended;

diagnosis.

4. The method as claimed in claim 2, wherein the medical entities are defined by manual or automated clustering techniques.

5. The method as claimed in claim 1, wherein modeling data models the treatment paths of successfully diagnosed patients using a different Markov model instance for each possible diagnosis.

6. The method as claimed in claim 5, wherein the Markov model instances include transition probabilities between medical entities learned using probability estimation techniques.

7. The method as claimed in claim 1, wherein modeling data is carried out off-line and regularly updated.

8. The method as claimed in claim 1, wherein predicting diagnosis for a current patient uses the modeled data to calculate the probability of each model instance for each diagnosis and choosing the model instance of the diagnosis that maximizes the treatment path probability.

9. The method as claimed in claim 6, wherein predicting diagnosis includes using a Bayesian classifier to predict a given diagnosis given the Markov model instances and a current path.

10. The method as claimed in claim 1, including sorting the possible diagnosis by the probability of each possible diagnosis.

11. The method as claimed in claim 1, including:

given two or more patients and multiple diagnosis models;

for each patient calculating the probability vectors of paths for one or more diagnoses; and

comparing the probability vectors for the patients to determine similarities.

12. A computer program product for predicting diagnosis of a patient, the computer program product comprising a non-transitory computer readable storage medium having computer readable program code embodied therewith, the computer readable program code executable by at least one processor to:

model data from a group of patients successfully diagnosed as having one or more medical conditions, wherein the data is modeled as treatment paths of patients including referrals to medical practitioners; and

predict diagnosis of one or more medical conditions for a current patient by comparing a treatment path of the current patient with the modeled treatment paths of successfully diagnosed patients, including calculating a probability of a given diagnosis of one or more medical conditions from the modeled treatment paths, wherein the calculating is in accordance with;

13. A system for predicting diagnosis of a patient, comprising:

a processor;

a modeling component for modeling data from a group of patients successfully diagnosed as having one or more medical conditions, wherein the data is modeled as treatment paths of patients including referrals to medical practitioners; and

a patient analysis component for predicting diagnosis of one or more medical conditions for a current patient by comparing a treatment path of the current patient with the modeled treatment paths of successfully diagnosed patients, including calculating a probability of a given diagnosis of one or more medical conditions from the modeled treatment paths, wherein the calculating is in accordance with;

14. The system as claimed in claim 13, wherein the modeling component includes:

an entity defining component for defining a set of medical entities including medical practitioners to which a patient has been referred; and

a path gathering component for gathering treatment paths of successfully diagnosed patients, wherein the treatment path links medical entities in a directional route.

15. The system as claimed in claim 14, wherein the entity defining component is for defining medical entities by manual or automated clustering techniques.

16. The system as claimed in claim 13, wherein the modeling component includes:

a model building component for modeling the treatment paths of successfully diagnosed patients using a different Markov model instance for each diagnosis.

17. The system as claimed in claim 16, wherein the Markov model instances include transition probabilities between medical entities learned using probability estimation techniques.

18. The system as claimed in claim 13, wherein the modeling component operates off-line and regularly updated.

19. The system as claimed in claim 13, wherein the patient analysis component includes:

a current patient receiving component for receiving a current patient treatment path;

a probability calculating component for using the modeled data to calculate the probability of each model instance for each possible diagnosis; and

a selection component for choosing the model instance of the diagnosis that maximizes the treatment path probability.

20. The system as claimed in claim 16, wherein the probability calculating component includes a Bayesian classifier to predict a given diagnosis given the Markov model instances and a current path.

21. The system as claimed in claim 19, wherein the selection component includes a ranking component for sorting the possible diagnosis by the probability of each possible diagnosis.

22. The system as claimed in claim 13, including a patient comparing component for comparing two or more patients given multiple diagnosis models, including:

for each patient calculating the probability vectors of paths for one or more diagnoses; and

comparing the probability vectors for the patients to determine similarities.