Author: Nina Hallowell, Michael Parker, Christoffer Nellåker
Publisher: Genetics in Medicine
Publication Year: 2018
Summary: The following article discusses how computational phenotyping (using machine learning algorithms [MLAs] to analyze photographic images) has improved healthcare experience for rare disease patients and facilitated the research for clinical geneticists. Although there are many benefits and beneficiaries of computational phenotyping, its use raises a number of ethical and legal issues. Some of these pertain to the use of personal data, which in this case is even more challenging as it involves the usage of data from children. This paper provides 3 other ethical issues which are relevant to computational phenotyping: 1). Data-induced discrimination, 2). The management of incidental findings, and 3). The commodification of (phenotypic) datasets. All apply to the use of MLAs in general and their use in other healthcare contexts, and will become more relevant for those working in genetics research and clinical practice as computational phenotyping tools are increasingly deployed.