Grant: Operating Grant: Data Analysis Using Existing CDN Databases and Cohorts

Pre-eclampsia (PE) is a progressive, multi-system disorder of pregnancy which causes high blood pressure, protein in the urine and potentially, injury to multiple organs. PE can result in fetal growth restriction and stillbirth. With no cure for PE available, prevention, early diagnosis, and close monitoring by a special are the best available options for patients. Research into ways of identifying women at high risk of PE early enough in pregnancy to intervene with aspirin or other preventive treatments is an urgent priority. Artificial intelligence (AI) tools have been shown to be capable of identifying individuals at high risk of a number of medical conditions using large clinical and health databases, and even using medical images from ultrasounds. In this proposed study, our goal is to use AI tools and methods to identify pregnancies at high risk of developing PE from screening ultrasound images taken during pregnancy. In order to accomplish this, we will need to link information from various existing databases from Ottawa together. Specialized AI tools, called neural networks, will compare ultrasound images from healthy pregnancies and pregnancies affected by PE, and identify differences in these images. The differences can then be used to predict which pregnancies will develop PE in the future. If our AI models are successful in finding differences in ultrasound images between PE and healthy pregnancies, we could use this information to identify high-risk pregnancies early. Early detection of women at increased risk would allow health care providers to ensure the pregnancy is monitored by a specialist. The next step in this research program would be to expand our work to additional databases, beyond Ottawa, and possibly internationally.