AI-Based Oxytocin-Stimulated Smart Lactation Support Vest

This randomized controlled trial is the second phase of a three-stage project aimed at developing and clinically evaluating an artificial intelligence (AI)-based oxytocin-stimulated smart lactation support vest with thermo-massage and vibration technologies. The overall project includes prototype development, randomized controlled clinical evaluation, and a qualitative assessment of mothers' experiences with the device.

The randomized controlled trial will evaluate the effectiveness of the AI-based oxytocin-stimulated smart lactation support vest in mothers who have delivered by cesarean section at 32-37 weeks of gestation and whose newborns are hospitalized in the Neonatal Intensive Care Unit (NICU). Because separation of the mother and newborn and the inability to directly breastfeed may interfere with the initiation and maintenance of lactation in this population, the intervention is designed to support milk production and milk flow during the early postpartum period.

The smart lactation support vest is a wearable system incorporating controlled thermo-massage and vibration stimulation. The vest includes heating elements, vibration units, massage and heating components, temperature and bioelectrical impedance sensors, an embedded electronic control system, a rechargeable battery, a control panel, and a mobile application. The system is designed to provide synchronized stimulation to the breast and upper back/scapular region. The breast-area components provide controlled thermal and vibration stimulation, while the back-area massage and heating components are intended to promote relaxation and support the physiological milk ejection reflex. The intervention is designed to support endogenous oxytocin release and facilitate milk flow without administration of exogenous oxytocin.

The AI-based decision-support module integrated into the mobile application will evaluate device-use parameters and user-related information and will provide individualized recommendations regarding the duration and intensity of thermo-massage and vibration stimulation. The system will use information obtained from temperature and bioelectrical impedance sensors together with milk volume, breast fullness, stimulation parameters, duration of use, session frequency, and user feedback. The AI module is intended to function as a supportive decision-support system and will not provide medical diagnosis or treatment decisions. If predefined warning indicators suggesting increased breast temperature and fullness are detected, the system will provide an alert advising the mother to seek evaluation by a healthcare professional.

Eligible participants will be randomly allocated in a 1:1 ratio to either the smart lactation support vest group or the control group. Stratified randomization according to gestational age (32-34 weeks and 35-37 weeks) will be used to promote balance between groups with respect to gestational age.

Participants assigned to the intervention group will use the smart lactation support vest for 10 days, three times daily. Each session will consist of a standardized sequence of 5 minutes of heating, 15 minutes of heating combined with massage, and 10 minutes of heating combined with massage and vibration. During the initial hospitalization period, participants will receive standardized training on how to wear and operate the vest. Following discharge, adherence to the intervention will be monitored through scheduled follow-up, reminder messages, telephone contact, and the mobile application. The first daily session will be supervised by the researcher during the mothers' scheduled visits to the NICU.

Participants in the control group will receive standard lactation care and counseling according to the hospital protocol. Standard care will include routine lactation counseling, guidance regarding breastfeeding and milk expression, breast care, storage of expressed breast milk, daily reminders to express milk, and lactation support when clinically indicated. Both groups will receive the same breast milk expression equipment and breast milk storage materials. The primary difference between the groups will be the use of the smart lactation support vest by participants assigned to the intervention group.

Maternal and newborn outcomes will be monitored for 10 days. The primary outcome is the amount of expressed breast milk, measured in cubic centimeters (cc) per day. Secondary maternal outcomes include oxytocin, prolactin, and cortisol levels; breast fullness; breast temperature; breast discomfort or stinging; lactation-related symptoms; anxiety; and psychological well-being. Hormonal measurements will be obtained at baseline, day 5, and day 10. Breast fullness, breast temperature, and breast discomfort/lactation-related symptoms will be assessed during the 10-day follow-up period using standardized assessment methods.

Newborn outcomes will include daily intake of expressed breast milk, daily weight change, blood glucose levels, and bilirubin levels, which will be monitored and recorded throughout the 10-day follow-up period.

The study will also assess the safety, usability, and acceptability of the wearable system during clinical use. Device-related observations, user feedback, and adherence to the intervention will be recorded. Mothers assigned to the intervention group will complete a user experience and satisfaction assessment at the end of the 10-day intervention period.

The study is designed to determine whether use of the AI-based oxytocin-stimulated smart lactation support vest, in addition to standard lactation care, improves expressed breast milk production and selected maternal and newborn outcomes compared with standard lactation care alone.