®®®® SIIA Público

Título del libro:
Título del capítulo: Automatic Standard Plane Detection in Fetal Ultrasound Improved by Fine-Tuning

Autores UNAM:
JORGE LUIS PEREZ GONZALEZ; FERNANDO ARAMBULA COSIO; MARIO ESTANISLAO GUZMAN HUERTA; BORIS ESCALANTE RAMIREZ; JIMENA OLVERES MONTIEL;
Autores externos:

Idioma:

Año de publicación:
2025
Palabras clave:

Deep learning; Convolutional neural network; Echography; Fetal growth; Fetal ultrasound; Fine tuning; Machine-learning; Plane detection; Similarity indices; Standard plane; Ultrasound images; Convolutional neural networks


Resumen:

During evaluation of fetal growth on ultrasound images, it is a vital requirement the expert selection of standard measurement planes of the fetal brain, abdomen and femur. Manual expert plane selection is a process subject to intra- and inter-operator variability, therefore some authors have proposed deep learning-based systems for automatic detection of standard fetal planes in the second trimester of pregnancy. However, fetal growth evaluation is increasingly recommended during the third trimester of gestation. This work proposes the fine-tuning, retraining, and validation of a system based on convolutional neural networks to improve the detection of the three main fetometry planes: brain, abdomen, and femur. The system was fine-tuned by retraining using ultrasound images from the second and third trimesters. The results of the proposed system were compared with standard planes of the same patient selected by medical specialists of the National Institute of Perinatology of Mexico, obtaining average structural similarity indices of 0.79, average similarity indices based on histogram correlation of 0.99, and average Pearson correlation coefficients of 0.96. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.


Entidades citadas de la UNAM: