Education‎ > ‎MSc theses‎ > ‎Open topics‎ > ‎

MS2017-03: Data driven diagnosis in Psychiatry

posted Aug 30, 2017, 2:41 AM by Marco Spruit   [ updated Aug 30, 2017, 2:41 AM ]
At the psychiatry department of the UMCU, our data science team works on bringing the results of data analysis to the daily practice of the psychiatry work floor. Over the past two years, we have been working on creating an environment that allows working with patient data, and an infrastructure that makes the diverse types of patient data that are gathered available for analysis - partially with the help of MBI students. Currently, we are looking for a motivated MBIstudent with an interest in data analysis for the following problem:

Diagnosis in the Psychiatry domain is currently done using the standard classification of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). This is a set of standard guidelines that determines what disorder a patient has, based on its symptoms. The DSM-5 is currently criticised because of its rigidity and tendency to stereotype patients into certain categories. From practice we know that diagnoses are not usually so clear, for example because of multiple diagnoses that are applicable, or symptoms that do not clearly belong to a certain diagnosis.

We have several initial questions regarding this problem:
  • How exactly is the DSM-5 currently used in the psychiatry care process?
  • Is there a basis for putting the same 'label' on a certain group of people, i.e. are they actually similar based on what data we have about them? Or does it work for some diagnoses, and not for others?
  • If not, what is a better way to label patients then the DSM-5? 
  • How can we validate this better labelling? 
  • More conceptually, what exactly is the sense of a diagnosis - is it even needed?
To answer these questions, you will first have to become acquainted with the psychiatry domain and the patient data we have gathered, and be familiar with relevant data analysis techniques (such as clustering or classification), preferably in Python or R. Then we will devise a plan to answer these questions, or other questions we come up with during the initial phase. We can offer you a chance to work with actual patient data in a challenging environment - the actual work with the patient data has to be conducted within the UMCU (at the Uithof). For the rest of the project, you are free to work where and when it suits you.

For more information, get in touch with Vincent Menger or Marco Spruit.