Digital/computational phenotyping: What are the differences in the science and the ethics?

Area of Science:

• Computational science
• Digital health
• Behavioral science

Background:

• Digital phenotyping originated in mental health research but has expanded to diverse scientific fields.
• Algorithmic phenotyping technologies are being developed in projects like RADAR-AD and The Minerva Initiative at the University of Oxford's Big Data Institute.

Purpose of the Study:

• To describe and analyze digital biomarkers and computational phenotyping concepts.
• To link these concepts to broader digital phenotyping research.
• To compare and contrast epistemological and ethical implications of these technologies.

Main Methods:

• Conceptual analysis of digital phenotyping, digital biomarkers, and computational phenotyping.
• Examination of two case studies: The RADAR-AD project and The Minerva Initiative.
• Application of the concept of 'affordance' to analyze technological functionalities and implications.

Main Results:

• The phenotyping paradigm in the studied projects assumes 'objectivity,' interpreted differently based on the role of digital tools.
• The concept of 'affordance' highlights the relationship between technological functionalities and their potential uses and social impacts.
• The term 'digital phenotyping' is used with diverse meanings, necessitating careful distinctions.

Conclusions:

• Distinguishing between different meanings and applications of digital phenotyping is crucial as its usage diversifies.
• Understanding the 'affordance' of digital tools is key to navigating the ethical and social implications of computational phenotyping.
• The assumption of objectivity in digital phenotyping warrants critical examination in various disciplinary contexts.