Purpose
The paper aims to propose “hypothetical enrollment” as an anticipatory, situated and performative methodological approach to appreciate the organizational and epistemological consequences of adopting artificial intelligence (AI) diagnostics into clinical settings. This method provides a methodological contribution to move between the expectations about AI diagnostics and their integration into real-world, clinical settings.
Design/methodology/approach
The validity of this method was tested against an empirical case, the start-up Autism Scope (AS), which applies machine learning models for the early detection of autism spectrum disorder. As part of this pilot study, two interviews were conducted in person with designers from AS and three interviews with pediatric neuropsychiatrists.
Findings
Notwithstanding a generally positive attitude, several organizational and professional challenges emerged thanks to this method, such as the integration of the tool into hospital workflows and the potential effects for professional identity in neuropsychiatry.
Research limitations/implications
Other healthcare stakeholders, such as hospital mangers or policy makers, were not interviewed.
Originality/value
The “hypothetical enrollment” interviews allowed comparing the expectations and implementation strategies devised by AS’ designers with the impediments and challenges highlighted by neuropsychiatrists, that is, potential users who have not been involved in development yet.
The paper aims to propose “hypothetical enrollment” as an anticipatory, situated and performative methodological approach to appreciate the organizational and epistemological consequences of adopting artificial intelligence (AI) diagnostics into clinical settings. This method provides a methodological contribution to move between the expectations about AI diagnostics and their integration into real-world, clinical settings.
Design/methodology/approach
The validity of this method was tested against an empirical case, the start-up Autism Scope (AS), which applies machine learning models for the early detection of autism spectrum disorder. As part of this pilot study, two interviews were conducted in person with designers from AS and three interviews with pediatric neuropsychiatrists.
Findings
Notwithstanding a generally positive attitude, several organizational and professional challenges emerged thanks to this method, such as the integration of the tool into hospital workflows and the potential effects for professional identity in neuropsychiatry.
Research limitations/implications
Other healthcare stakeholders, such as hospital mangers or policy makers, were not interviewed.
Originality/value
The “hypothetical enrollment” interviews allowed comparing the expectations and implementation strategies devised by AS’ designers with the impediments and challenges highlighted by neuropsychiatrists, that is, potential users who have not been involved in development yet.