The CALOPUS project (Computer Assisted low-cost point of care ultrasound) is a UK-India Collaborative comprising a team of Clinicians and Engineers from our department, the Institute of Biomechanical Engineering in Oxford and the Translational Health Science and Technology Institute in Delhi.
(Images: The Calopus Team, 2020).
We are developing a low-cost point of care ultrasound device that would enable minimally trained healthcare providers to obtain clinically relevant information from a simple scanning protocol. We will realise this through the integration of point-of-care ultrasound and automation technology. Our aims are to automate key factors for delivery of pregnancy care: assessing viability; fetal presentation; identify multiple pregnancies; placental location; assess the adequacy of amniotic fluid; and basic fetal measurement.
The CALOPUS protocol involves the recording videos of very simple ultrasound sweeps. We annotate the structures within the videos, and use this information to train Deep Learning models that can automatically recognise useful maternal and fetal anatomy, the placenta and amniotic fluid. Integral to this work is the development of metrics to assess quality of annotations and accelerating the labour-intensive process of human annotation with automated annotations.
A CALOPUS could an important impact in low resource settings, where many women are unable to access an ultrasound scan. For more information see the study website https://eng.ox.ac.uk/calopus/
Funding: co-funded by the Global Challenge Research Fund (GCRF) and the Engineering and Physical Sciences Research Council (EPSRC).