FDA Clears AI for Delivery Date Estimation
A new Ultrasound AI technology for estimating delivery dates has received FDA De Novo clearance. The tool promises more precise dating than traditional methods, which could reduce unnecessary interventions for pregnancies considered post-term.
Traditional due date calculations, like Naegele's Rule, rely on the first day of the last menstrual period and assume a 28-day cycle, an approach over 200 years old. This method can be unreliable for individuals with irregular cycles, leading to potential inaccuracies in estimating the delivery date. While early ultrasound is more precise, only about 5% of babies are born on their exact due date. Inaccurate dating can lead to pregnancies being misclassified as post-term, which is after 42 weeks of gestation. This may trigger unnecessary medical interventions such as labor induction or Cesarean deliveries. Post-term pregnancies also carry genuine risks, including placental insufficiency, meconium aspiration, and an increased likelihood of C-section, making accurate dating crucial for both maternal and fetal well-being. The newly cleared "Delivery Date AI" by Ultrasound AI analyzes entire ultrasound images, looking at both fetal and maternal characteristics that correlate with delivery timing. The AI was trained on a massive dataset of over 2 million de-identified images from diverse pregnancies to generate a Predicted Delivery Date. The supporting PAIR study showed the AI's predictions improved with retraining, demonstrating its ability to learn. The technology received clearance through the FDA's De Novo pathway. This specific route is for novel, low-to-moderate-risk medical devices that have no existing equivalent on the market. Essentially, the De Novo process creates a new category of device, establishing a baseline for safety and effectiveness that future technologies can be compared against. This type of AI-driven tool enters the field amidst a significant, nationwide shortage of sonographers. From 2011 to 2021, the number of ultrasound exams grew 55.1%, while the number of practicing sonographers only increased by 43.6%. The demand for sonographers continues to outpace the number of graduates, straining healthcare resources. For midwifery, the integration of AI presents both opportunities and considerations. The Australian College of Midwives notes AI's potential to improve accuracy and efficiency but stresses that such tools should augment, not replace, the expertise of midwives. The core principle is that AI-assisted recommendations should support, rather than override, the choices of the mother and the clinical judgment of healthcare professionals.
