Random Fourier Features-Based Deep Learning Improvement with Class Activation Interpretability for Nerve Structure Segmentation

• Published in: Sensors (Basel, Switzerland) Switzerland), 2021-11, Vol.21 (22), p.7741
• Main Authors: Jimenez-Castaño, Cristian Alfonso, Álvarez-Meza, Andrés Marino, Aguirre-Ospina, Oscar David, Cárdenas-Peña, David Augusto, Orozco-Gutiérrez, Álvaro Angel
• Format: Article
• Language: English
• Subjects: Accuracy; Algorithms; Anesthesia; Approximation; Attenuation; class activation mapping; Deep Learning; Image Processing, Computer-Assisted; Machine learning; nerve structure segmentation; Neural networks; Neural Networks, Computer; random Fourier features; Semantic segmentation; Semantics; Ultrasonic imaging; Ultrasonography; ultrasound images; Wavelet transforms
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s21227741

Peripheral nerve blocking (PNB) is a standard procedure to support regional anesthesia. Still, correct localization of the nerve's structure is needed to avoid adverse effects; thereby, ultrasound images are used as an aid approach. In addition, image-based automatic nerve segmentation from deep learning methods has been proposed to mitigate attenuation and speckle noise ultrasonography issues. Notwithstanding, complex architectures highlight the region of interest lacking suitable data interpretability concerning the learned features from raw instances. Here, a kernel-based deep learning enhancement is introduced for nerve structure segmentation. In a nutshell, a random Fourier features-based approach was utilized to complement three well-known semantic segmentation architectures, e.g., fully convolutional network, U-net, and ResUnet. Moreover, two ultrasound image datasets for PNB were tested. Obtained results show that our kernel-based approach provides a better generalization capability from image segmentation-based assessments on different nerve structures. Further, for data interpretability, a semantic segmentation extension of the GradCam++ for class-activation mapping was used to reveal relevant learned features separating between nerve and background. Thus, our proposal favors both straightforward (shallow) and complex architectures (deeper neural networks).