AI Supported Lactation Care

Abstract

Breastfeeding is incredibly important for the health and well-being of both infants and mothers. However, many new parents face difficulties in accessing proper support from lactation consultants. Unfortunately, there is often a shortage of consultants, particularly in lower-income countries or for parents living far from urban centers. This lack of accessible support can be challenging and discouraging for families wanting to breastfeed. With the help of AI, The UC San Diego DigiHealth Lab and UCSD Lactation Program has developed a tele-lactation care system, aiming to provide just-in-time automated screening using a smartphone.

Why?

Pain experienced during breastfeeding can often be a cause of concern. Although some discomfort is natural, physical damage as a result of improper latching or clogged pores can lead to issues needing of elevated attention. By using AI to help identify potential breast conditions as a result of breastfeeding means that parents can receive immediate guidance and assistance, regardless of their geographical location or economic status. By leveraging AI, breastfeeding support can become more widely available and empower families. Such a technology can be used to help support lactation consultants to help more families or act as a tool to help connect parents to support as they need it.

How it works

With a regular smartphone, the breastfeeding parent is guided to capture an image of their breast to assess its condition. A computer vision deep learning algorithm leveraging computation of the smartphone or on the cloud analyzes the image. Trained on a large database of breast condition issues like mastitis and nipple damage, the algorithm is capable of identifying potential issues. Having been trained on a large variety of skintones, body types, and lighting conditions, the algorithm is robust to changing naturalistic use.