Abstract
In this chapter, we explore the potential applications of machine learning to brain disorders. Specifically, we illustrate why the use of machine learning in brain disorders is attracting so much interest among researchers and clinicians by highlighting three key applications: prediction of illness onset, assistance with diagnosis, and prediction of longitudinal outcomes. After illustrating these applications, we discuss the challenges that need to be overcome for a successful translational implementation of machine learning in everyday psychiatric and neurologic care. In particular, we identify three main pitfalls in the absence of biomarkers, the unreliability of clinical diagnosis, and the heterogeneity of the patients. In the final part of the chapter, we consider the requirements a machine learning algorithm needs to fulfill to be eligible for clinical use and discuss potential future directions.
Original language | English |
---|---|
Title of host publication | Machine Learning |
Subtitle of host publication | Methods and Applications to Brain Disorders |
Publisher | Elsevier |
Pages | 45-65 |
Number of pages | 21 |
ISBN (Electronic) | 9780128157398 |
DOIs | |
Publication status | Published - 1 Jan 2019 |
Keywords
- Biomarkers
- Clinical translation
- Diagnosis
- Diagnostic reliability
- Heterogeneity
- Machine learning
- Neuroimaging
- Neurological disorders
- Prediction
- Psychiatric disorders