Year: 2024
Type: conference-abstract
Abstract: Creation of a data set for the development of an AI service that improves the quality of diagnosis of fetal heart disease. Formation of an algorithm for examining the fetal heart using AI, the result of which should be one of the conclusions: “norm” - the correct structure of the heart / abnormal heart (CHD?). The study per patient should contain video files of 5 of fetal heart standard views. AI conclusion: normal structure/abnormal heart structure. Verification of the conclusion is confirmation of the diagnosis after birth. The total number of studies - 2016, the number of personnel - 659324. The number of frames from the CHD - 27886. The results showed that the AI algorithm can improve the accuracy of ultrasound imaging in the diagnosis of fetal heart pathology and has good application value. The use of AI will improve the quality of ultrasound examinations of the fetal heart due to its standardization and will increase the accuracy of pathology detection in general obstetric practice or in less experienced hands.
DOI: 10.1002/uog.28445
Language: English
Subfield: Health Informatics
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4402358298