Bayesian diagnosis

From Opasnet
Jump to: navigation, search

Bayesian diagnosis is a method for making clinical inferences about a patient's disease by using available information and Bayes' rule to update the diagnosis.


What is such a method for making clinical diagnosis that fulfils the following criteria:

  • It is based on the principles of open assessment.
  • It uses Bayesian updating.
  • The output are probabilities of different diseases, given information about a patient and his/her symptoms.
  • It offers a practical interface for a medical doctor or a patient to search for information.
  • It offers a feedback form where a user can upload and store patient information to the system, and later retrieve it.
  • It utilises national statistics of mortality and hospital discharge registries, and others.


To protect individual information, the system must not contains names, social security numbers, addresses or other information that can be used for identification. Individuals in the system are given an id number, which contains nothing of the previous information.

An assessor of the nugget must be able to access the data collected for the nugget. Therefore, he/she must have a password for the whole object. In addition, a person who has uploaded individual information, must have a password for the particular row in the database, but not other rows. Should there be two different passwords? The user password may relate to information in several objects, and therefore it should not be object-specific.

--> There should be two different passwords.

This closely relates to Opasnet base.



See also