Ultrasound images acquired during axillary nerve block procedures can be difficult to interpret. Highlighting the important structures, such as nerves and blood vessels, may be useful for the training of inexperienced users. A deep convolutional neural network is used to identify the musculocutaneous, median, ulnar, and radial nerves, as well as the blood vessels in ultrasound images. A dataset of 49 subjects is collected and used for training and evaluation of the neural network. Several image augmentations, such as rotation, elastic deformation, shadows, and horizontal flipping, are tested. The neural network is evaluated using cross validation. The results showed that the blood vessels were the easiest to detect with a precision and recall above 0.8. Among the nerves, the median and ulnar nerves were the easiest to detect with an F-score of 0.73 and 0.62, respectively. The radial nerve was the hardest to detect with an F-score of 0.39. Image augmentations proved effective, increasing F-score by as much as 0.13. A Wilcoxon signed-rank test showed that the improvement from rotation, shadow, and elastic deformation augmentations were significant and the combination of all augmentations gave the best result. The results are promising; however, there is more work to be done, as the precision and recall are still too low. A larger dataset is most likely needed to improve accuracy, in combination with anatomical and temporal models.
Most brain-shift compensation methods address the problem of updating preoperative images to reflect brain deformations following the craniotomy and dura opening. However, fewer enable to take into account the resection-induced deformations occuring all along the tumor removal procedure. This paper evaluates the use of two existing methods to tackle that problem. Both techniques rely on blood vessels segmented then skeletonized from preoperative MR Angiography and navigated Doppler Ultrasound images acquired during resection. While the first one proposes to register the vascular trees using a rigid modified ICP algorithm, the second method relies on a non-rigid constrained-based biomechanical approach. Quantitative results are provided, based on distances between paired landmarks set on blood vessels then anatomical structures delineated on medical images. A qualitative evaluation of the compensation is also presented using initial and updated images. An analysis on three cases of surface tumor shows both methods, especially the biomechanical one, can compensate up to 63% of the brain-shift, with an error in the range of 2 mm. However, these results are not reproduced on a more complex case of deep tumor. While more patients must be included, these preliminary results show that vesselbased methods are well suited to compensate for resection-induced brain-shift, but better outcomes in complex cases still require to improve the methods to take the resection into account.