Deep learning approach for automatic out-of-plane needle localisation for semi-automatic ultrasound probe calibration

Healthcare Technology Letters

article

Leah A. Groves¹  Blake VanBerlo³  Terry M. Peters¹  Elvis C.S. Chen¹ 

[1] School of Biomedical Engineering, University of Western Ontario;Robarts Research Institute, University of Western Ontario;Schulich School of Medicine, University of Western Ontario

关键词: medical image processing;    calibration;    learning (artificial intelligence);    needles;    image registration;    mean square error methods;    biomedical ultrasonics;    convolutional neural nets;    out-of-plane needle localisation;    semiautomatic ultrasound probe calibration;    deep learning algorithm;    needle reflection;    probe calibration algorithm;    convolutional neural network;    automatic centroid localisation algorithm;    probe calibrations;    pixel localisation;    semiautomatic implementation;    automatic needle centroid localisation;    target registration errors;    calibration method;    size 6.0 cm;    size 4.0 cm to 8.0 cm;   

DOI  :  10.1049/htl.2019.0075

学科分类：肠胃与肝脏病学

来源: Wiley

【 摘 要 】

The authors present a deep learning algorithm for the automatic centroid localisation of out-of-plane US needle reflections to produce a semi-automatic ultrasound (US) probe calibration algorithm. A convolutional neural network was trained on a dataset of 3825 images at a 6 cm imaging depth to predict the position of the centroid of a needle reflection. Applying the automatic centroid localisation algorithm to a test set of 614 annotated images produced a root mean squared error of 0.62 and 0.74 mm (6.08 and 7.62 pixels) in the axial and lateral directions, respectively. The mean absolute errors associated with the test set were 0.50 ± 0.40 mm and 0.51 ± 0.54 mm (4.9 ± 3.96 pixels and 5.24 ± 5.52 pixels) for the axial and lateral directions, respectively. The trained model was able to produce visually validated US probe calibrations at imaging depths on the range of 4–8 cm, despite being solely trained at 6 cm. This work has automated the pixel localisation required for the guided-US calibration algorithm producing a semi-automatic implementation available open-source through 3D Slicer. The automatic needle centroid localisation improves the usability of the algorithm and has the potential to decrease the fiducial localisation and target registration errors associated with the guided-US calibration method.

【 授权许可】

CC BY|CC BY-ND|CC BY-NC|CC BY-NC-ND