Ankita Reddy

From smartphone ultrasound images to skin photos, digital images of the human body are increasingly ubiquitous in U.S. clinical care. To help clinicians and patients process the proliferation of these images, computer scientists and technology companies are rapidly embedding artificial intelligence (AI) computer vision algorithms within these imaging technologies. While promises of faster and more accurate diagnoses propel the use of AI computer vision in clinical care, phenotyping pathologies and “markers” of patient characteristics such as gender and race become reformulated in the design and use of image-based algorithms. How are computer vision technologies reshaping how people come to see and act on social and biological difference through images? I will conduct ethnographic work investigating a computer vision research lab, a commercial AI imaging company, and a hospital using imaging in everyday care. In these sites, I examine how computer scientists, technology developers, clinicians, and patients reformulate assemblages of images and algorithms to sense, represent, and attend to bodies. In treating algorithmic development, commercial organization, and medical practices as sites of inquiry, I interrogate how digital and computational abstractions of humans interact with the affective and political dimensions of reading bodies as images to render difference legible in the clinic.