Nature article authors propose asking donors separately whether their cells may be used in biocomputers

Nature article authors propose asking donors separately whether their cells may be used in biocomputers

On 27 July, neuroscientists and bioethicists published proposed rules in Nature for laboratories that connect brain organoids to computing systems. They recommend obtaining separate consent from each donor for this use. If the donor cannot be contacted, they recommend referring the project to an independent committee.

A brain organoid is a small, three-dimensional cluster of neurons grown from stem cells. In biocomputing, researchers deliver an electrical signal to the organoid through microelectrodes and record the neurons' response. A computer treats this response as the result of a computation. Such systems have already performed simple pattern-recognition and prediction tasks.

Organoids are often made from cells donated by people participating in medical research. A cell line can be maintained for years and used in new projects. The article's authors argue that consent for biomedical research does not constitute consent for computational or commercial applications, such as voice-recognition or facial-recognition systems.

The authors recommend that laboratories establish new cell lines with explicit consent for computational use. For existing cell lines, researchers should contact the donor again. If this is impossible, an independent committee that includes neuroscientists and engineers should evaluate the project. Conventional stem-cell committees review biomedical research, but they usually do not consider projects conducted by computing and engineering departments.

In interviews with 67 donors and parents of child donors, two thirds of the participants wanted to be informed about uses of organoids beyond the original study, and more than half wanted to grant permission for further use. Some parents asked for the child to make that decision after reaching adulthood. These preferences support the authors' position that the original consent should not be treated as universal permission for a new engineering application.

In the authors' model, consent becomes part of the process by which a laboratory selects cells for a biocomputer. It determines which cell lines may be used for computational tasks and who decides whether they may be used when a new engineering application was not covered by the original agreement.