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

• Published in: Healthcare technology letters Vol. 6; no. 6; pp. 204 - 209
• Main Authors: Groves, Leah A, VanBerlo, Blake, Peters, Terry M, Chen, Elvis C.S
• Format: Journal Article
• Language: English
• Published: England The Institution of Engineering and Technology 01.12.2019; John Wiley & Sons, Inc; Wiley
• Subjects: Accuracy; Algorithms; automatic centroid localisation algorithm; automatic needle centroid localisation; biomedical ultrasonics; Calibration; calibration method; convolutional neural nets; convolutional neural network; Data collection; Deep learning; deep learning algorithm; image registration; learning (artificial intelligence); Localization; mean square error methods; medical image processing; Methods; needle reflection; needles; Neural networks; out-of-plane needle localisation; pixel localisation; probe calibration algorithm; probe calibrations; Registration; semiautomatic implementation; semiautomatic ultrasound probe calibration; Sensors; size 4.0 cm to 8.0 cm; size 6.0 cm; Special Issue: Papers from the 13th Workshop on Augmented Environments for Computer Assisted Interventions; target registration errors; Ultrasonic imaging; Usability; deep learning algorithm; needles; out-of-plane needle localisation; image registration; semiautomatic implementation; convolutional neural network; biomedical ultrasonics; semiautomatic ultrasound probe calibration; probe calibrations; size 6.0 cm; calibration method; probe calibration algorithm; pixel localisation; target registration errors; automatic centroid localisation algorithm; convolutional neural nets; size 4.0 cm to 8.0 cm; learning (artificial intelligence); needle reflection; calibration; mean square error methods; automatic needle centroid localisation; medical image processing
• DOI: 10.1049/htl.2019.0075
• ISSN: 2053-3713

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.