Social Complexity and Fairness in Synthetic Medical Data

Medical research is increasingly using big data and powerful computers. But one problem with this is that the powerful computer methods we have now for dealing with big data make it easy to figure out who is who in a data set, which is especially bad for the privacy issues related to sensitive medical data. A solution is to use machine learning to generate synthetic data from the raw data, that is, to make a fake data set that still represents important elements of the data.

While this is good in theory, early results from this process indicates that machine learning generated datasets may over-represent majority elements and diminish representation of minority elements. When applied to medical data, this would mean that synthetic datasets probably have an over-representation of ‘standard’ patients, i.e. white, middle class, 35-yr old men, despite decades of regulation and research practice that has tried to include other patients and bodies in medical research.

This project, focused around a 2 year postdoc, and as a collaboration between WASP-HS & WASP researchers will develop fairness metrics to evaluate the production of synthetic data from a specific medical dataset, with the hypothesis that intersectionality can contribute to better data. Additionally, we will closely examine existing synthetic medical data to see if there are lessons social science can take from it to inform theoretical work about intersectional power dynamics in society.

Colleagues in this project are: Francis Lee, Gabriel Eilertsen, Saghi Hajisharif, and Tahereh Dehdarirad

The research is funded by  WASP-HS.

Representative and equitable synthetic data: ML algorithms and working practices.

This 12-month, 3-partner project aims to develop a combined sociotechnical tool and process to ensure that Swedish industry engages with synthetic data that is fair, equitable and representative of edge cases. Our work both develops technical solutions for representative synthetic data and explores the market to discover what alignment needs Swedish industry has for ML and synthetic data. We are integrating understandings of intersectional hallucinations in this work. Here is a film made to represent the project at the AI exhibition at the House of Sweden in Washington DC.

The research is funded by Vinnova

The ethics and social consequences of AI & caring robots. Learning trust, empathy and accountability

Our fascination with robots is old. So are our misgivings. Science fiction has warned us of the day they will take over for more than a century. Social theorists have long been predicting the consequences of robots’ entrance into the workplace. Luddites have been warning of their impact on our lives and our relationships. And more nuanced examinations have probed the way we think of ourselves when we think of (and with) them. Yet, for many of us, robots in that stereotypical, personified form, as a unit we interact/intra-act with on an emotional level, have stayed in the realm of science fiction. We may have a robotic vacuum cleaner at home. We may have even given that vacuum cleaner a name. But an autonomous housekeeper robot who is part of the family (à la Rosie the robot maid in The Jetsons)? Not yet.

This is about to change. Robots are starting to enter our daily life. We and our children are going to be expected to interact with robots as they perform different kinds of care for and with us at different life stages. What will that do to how we – and how the robots – think of care? And how are we going to produce accountability, trust and empathy in the relational intra-actions we have, together?

In the project I am working with Katherine Harrison, Senior lecturer, Tema, Linköping University,  Ginevra Castellano, Senior lecturer, Social Robotics Lab, Uppsala University, and Dominika Lisy, Tema, Linköping University, Maria Arnelid, Tema, Linköping University, and Isabel García Velázquez, Tema, Linköping University.

The research is funded by the Marianne and Marcus Wallenberg Foundation and WASP-HS.

A Constant Torment. Tracing the Discursive Contours of the Aging Prostate

This project employs Medical Sociology, History of Medicine, STS & Medicine, and Feminist Science Studies in collaborative work around an evocative object, the aging prostate. It combines empirical rigor with an analytical interest in the practices surrounding medical technologies and their roles in how we conceive of and challenge gendered subject positions and bodily knowledges.

Each of the projects within this study traces the contours and textures of discourses that produce the aging prostate in different incarnations. Some examine cultural and historical constructions of the prostate as a node of pain, discomfort, and angst, using interviews and archive material. Others explore the medical discourses (including material semiotic practices), which enact the prostate as a discrete anatomical object to be physically examined and surgically removed, using observations, interviews, and analysis of medical discourses. And some projects delineate how the prostate is known and invoked in its absence, after surgical removal, using post-structuralist analysis of interview material.

Colleagues in this project were: Jelmer Brüggeman, Carina Danemalm-Jägervall, Jenny Gleisner, Sonja Jerak-Zuiderent, Alma Persson, Maria Björkman, Oscar Maldonado Castaneda, and Elin Björk.

The research was funded by a Swedish Research Council ”bidrag till framstående yngre forskare”.

Prescriptive Prescriptions. Pharmaceuticals and 'Healthy' Subjectivities

This research explored how ‘healthy’ subjects are prescribed - defined and framed - by medical pharmaceuticals. It examined the cultural meanings and expectations attached to four prescription drugs, and compared the policies and practices around their use in two European countries, Sweden and UK. The drugs studied were the HPV vaccine (to prevent cervical cancer), hormone therapies to delay puberty, treatments for benign prostate hyperplasia, and drugs against Alzheimer’s disease. Methodologically, the work employed qualitative methods: textual analysis of marketing material, policy documents, and patient advocacy group material; observational fieldwork; and interviews with policy makers, medical practitioners and patient advocates.

The project had seven participants and was spread across two universities: Linköping University (Sweden) and Lancaster University (UK). Working on this team were Prof. Cecilia Åsberg, Prof. Celia Roberts, Dr. Oscar Maldonado Castaneda, Ali Hanbury, Lisa Lindén, and Tara Mehrabi.

The research was funded by an ERC Starting Grant. Completed 2017.