OC08.07: Preliminary results on the assessment of the feasibility of instrumental delivery using an automatic algorithm
Year: 2024
Type: conference-abstract
Abstract: The primary outcome of the study is to preliminarily evaluate the performance of an automatic algorithm which is expected to produce a colour code (green, yellow and red) corresponding to the feasibility of instrumental delivery, starting from transperineal ultrasound (US) images acquired on the axial plane. The automatic algorithm predicts the chance of success of instrumental delivery by the assessment/measurement of the following three parameters measured from US images: the occiput position, the head-perineum distance and the midline angle. The outcome is a colour code related to the likelihood of successful vacuum extraction, from green (high chances of success) to yellow (intermediate chance of success) to red (low chance of success). An initial dataset of 26 ultrasound B-mode images was acquired in the axial plane on a cohort of women in the second stage of labour. Results were assessed in relation to the agreement between the algorithm output and the final outcome (“ground truth”). Algorithm sensitivity (Sens) and specificity (Spec) are computed for green (G) and yellow (Y) code, respectively: SensG = 0.80, SpecG = 0.82, SensY = 0.64, SpecY = 0.80. Only two cases have been classified with red colour, producing an overall accuracy of 0.73. The final algorithm will aim to support clinical decision in the management of delivery. Preliminary results are encouraging, highlighting the feasibility of the proposed approach, and lead the way for further improvements. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
DOI: 10.1002/uog.27776
Authors: A. Dall'Asta, R. Morello, C. Melito, C. Botrugno, A. Bottino, F. Conversano, P. Pisani, S. Casciaro, T. Ghi
Language: English
Subfield: Obstetrics and Gynecology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4402360087