Generating artificial patient data

From Master Projects
Jump to: navigation, search

About Generating artificial patient data


Access to large volumes of patient data is crucial for many research projects in medical applications of AI. However, it is often difficult to get access to such data, because of privacy restrictions and organisational barriers. In such cases, artificially generated patient data is a plausible replacement for real patient data, assuming that the artificial data is sufficiently similar to real data. We have built an artificial patient data generator, that uses knowledge about probability distributions of symptoms, diseases and patient characteristics to generate large cohorts of realistic patient data. In ths project, you will extend the functionality of the patient-data generator to cover wider application scenarios, and to generate realistic patient cohorts with psychiatric conditions (Smart Ward). Your work will contribute to ongoing research projects in Amsterdam and Beijing.

Supervisors: Zhisheng Huang and Annette ten Teije