Describing Black-Box Medicine

Abstract

Black-box medicine is a developing form of personalized medicine that uses opaque computer models based on big data to make predictions and recommendations about health decisions. Black-box medicine has significant promise to advance medicine, but also demands substantial investments. This raises the question of intellectual property incentives for development, especially patents. For patenting black-box medical algorithms, Section 101 subject-matter eligibility is an ongoing question under evolving Supreme Court precedent, but this paper questions how black-box algorithms fare when confronted with a more prosaic hurdle: Section 112. For something which is either too complex to understand or inherently fully opaque, how can a patent applicant adequately describe and enable the claimed technology? This paper concludes that Section 112 presents major -- but not insurmountable -- hurdles for patenting black-box medical algorithms.

Note: This piece was written in connection with the Boston University Workshop on Personalized Medicine and Intellectual Property.

Keywords: personalized medicine, black box, patentability, enablement, 112