Breast calcification detection based on multichannel radiofrequency signals via a unified deep learning framework

作者：

Highlights：

• Multichannel ultrasound radiofrequency signals fusion and spectrogram conversion.

• A unified deep learning framework to detect breast calcifications.

• A calcification tracking mechanism to further improve the detection accuracy.

• Superior performance compared to state-of-the-art works.

摘要

•Multichannel ultrasound radiofrequency signals fusion and spectrogram conversion.•A unified deep learning framework to detect breast calcifications.•A calcification tracking mechanism to further improve the detection accuracy.•Superior performance compared to state-of-the-art works.

论文关键词：Breast tumor,Calcification detection,Convolutional neural network,Spectrogram,Long short-term memory,Kalman filter

论文评审过程：Received 3 June 2019, Revised 25 November 2019, Accepted 1 November 2020, Available online 21 November 2020, Version of Record 24 January 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114218