®®®® SIIA Público

Título del libro: Proceedings Of The 21st International Symposium On Medical Information Processing And Analysis, Sipaim 2025
Título del capítulo: Deep Learning-based Echocardiography Enhancement: Reducing Speckle Noise and Improving Temporal Resolution

Autores UNAM:
JORGE LUIS PEREZ GONZALEZ; NIDIYARE HEVIA MONTIEL;
Autores externos:

Idioma:

Año de publicación:
2025
Palabras clave:

De-noising; Echocardiographic imaging; Intermediate flow; Learning-based approach; Net model; Real- time; Resolution improvement; Speckle noise; Speckle noise reduction; Temporal resolution


Resumen:

© 2025 IEEE.Ultrasound is widely used in medical imaging due to its accessibility and cost-effectiveness. This paper presents a Deep Learning-based approach for enhancing echocardiographic imaging by addressing two challenges: speckle noise reduction and temporal resolution improvement. For denoising task, we propose the SE-U-Net model, while for temporal enhancement we employ Real-Time Intermediate Flow Estimation for frame interpolation. Multiple datasets were used for training, validation, and testing of both models. The proposed pipeline was applied in high heart rate murine models to enhance relevant information. In final evaluation, the denoising model achieved SSIM=0.99 and PSNR=38.27dB, indicating effective contrast enhancement, preservation of homogeneous regions, and reduction of small artifacts. The interpolation model, on the other hand, achieved SSIM=0.87 and PSNR=33.23dB, enabling the increase of ultrasound temporal resolution from 37 to 296 frames per second without losing image quality.


Entidades citadas de la UNAM: