What are the optimal targeting visualizations for performing surgical navigation of iliosacral screws? A user study.

INTRODUCTION: Complex orthopaedic procedures, such as iliosacral screw (ISS) fixations, can take advantage of surgical navigation technology to achieve accurate results. Although the impact of surgical navigation on outcomes has been studied, no studies to date have quantified how the design of the targeting display used for navigation affects ISS targeting performance. However, it is known in other contexts that how task information is displayed can have significant effects on both accuracy and time required to perform motor tasks, and that this can be different among users with different experience levels. This study aimed to investigate which visualization techniques helped experienced surgeons and inexperienced users most efficiently and accurately align a surgical tool to a target axis. METHODS: We recruited 21 participants and conducted a user study to investigate five proposed 2D visualizations (bullseye, rotated bullseye, target-fixed, tool-fixed in translation, and tool-fixed in translation and rotation) with varying representations of the ISS targets and tool, and one 3D visualization. We measured the targeting accuracy achieved by each participant, as well as the time required to perform the task using each of the visualizations. RESULTS: We found that all 2D visualizations had equivalent translational and rotational errors, with mean translational errors below 0.9 mm and rotational errors below 1.1[Formula: see text]. The 3D visualization had statistically greater mean translational and rotational errors (4.29 mm and 5.47[Formula: see text], p < 0.001) across all users. We also found that the 2D bullseye view allowed users to complete the simulated task most efficiently (mean 30.2 s; 95% CI 26.4-35.7 s), even when combined with other visualizations. CONCLUSIONS: Our results show that 2D bullseye views helped both experienced orthopaedic trauma surgeons and inexperienced users target iliosacral screws accurately and efficiently. These findings could inform the design of visualizations for use in a surgical navigation system for screw insertions for both training and surgical practice.

Ensemble of deep convolutional neural networks is more accurate and reliable than board-certified ophthalmologists at detecting multiple diseases in retinal fundus photographs.

AIMS: To develop an algorithm to classify multiple retinal pathologies accurately and reliably from fundus photographs and to validate its performance against human experts. METHODS: We trained a deep convolutional ensemble (DCE), an ensemble of five convolutional neural networks (CNNs), to classify retinal fundus photographs into diabetic retinopathy (DR), glaucoma, age-related macular degeneration (AMD) and normal eyes. The CNN architecture was based on the InceptionV3 model, and initial weights were pretrained on the ImageNet dataset. We used 43 055 fundus images from 12 public datasets. Five trained ensembles were then tested on an 'unseen' set of 100 images. Seven board-certified ophthalmologists were asked to classify these test images. RESULTS: Board-certified ophthalmologists achieved a mean accuracy of 72.7% over all classes, while the DCE achieved a mean accuracy of 79.2% (p=0.03). The DCE had a statistically significant higher mean F1-score for DR classification compared with the ophthalmologists (76.8% vs 57.5%; p=0.01) and greater but statistically non-significant mean F1-scores for glaucoma (83.9% vs 75.7%; p=0.10), AMD (85.9% vs 85.2%; p=0.69) and normal eyes (73.0% vs 70.5%; p=0.39). The DCE had a greater mean agreement between accuracy and confident of 81.6% vs 70.3% (p<0.001). DISCUSSION: We developed a deep learning model and found that it could more accurately and reliably classify four categories of fundus images compared with board-certified ophthalmologists. This work provides proof-of-principle that an algorithm is capable of accurate and reliable recognition of multiple retinal diseases using only fundus photographs.

Can uncertainty estimation predict segmentation performance in ultrasound bone imaging?

PURPOSE: Segmenting bone surfaces in ultrasound (US) is a fundamental step in US-based computer-assisted orthopaedic surgeries. Neural network-based segmentation techniques are a natural choice for this, given promising results in related tasks. However, to gain widespread use, we must be able to know how much to trust segmentation networks during clinical deployment when ground-truth data is unavailable. METHODS: We investigated alternative ways to measure the uncertainty of trained networks by implementing a baseline U-Net trained on a large dataset, together with three uncertainty estimation modifications: Monte Carlo dropout, test time augmentation, and ensemble learning. We measured the segmentation performance, calibration quality, and the ability to predict segmentation performance on test data. We further investigated the effect of data quality on these measures. RESULTS: Overall, we found that ensemble learning with binary cross-entropy (BCE) loss achieved the best segmentation performance (mean Dice: 0.75-0.78 and RMS distance: 0.62-0.86mm) and the lowest calibration errors (mean: 0.22-0.28%). In contrast to previous studies of area or volumetric segmentation, we found that the resulting uncertainty measures are not reliable proxies for surface segmentation performance. CONCLUSION: Our experiments indicate that a significant performance and confidence calibration boost can be achieved with ensemble learning and BCE loss, as tested on 13,687 US images containing various anatomies and imaging parameters. However, these techniques do not allow us to reliably predict future segmentation performance. The results of this study can be used to improve the calibration and performance of US segmentation networks.

Novel techniques for intraoperative parathyroid gland identification: a comprehensive review.

INTRODUCTION: The parathyroid glands (PGs) are critical for calcium regulation and homeostasis. The preservation of PGs during neck surgery is crucial to avoid postoperative hypoparathyroidism. There are no existing guidelines for intraoperative PG identification, and the current approach relies heavily on the experience of the operating surgeon. A technique that accurately and rapidly identifies PGs would represent a useful intraoperative adjunct. AREAS COVERED: This review aims to assess common dye and fluorescence-based PG imaging techniques and examine their utility for intraoperative PG identification. A literature search of published data on methylene blue (MB), indocyanine green (ICG) angiography, near-infrared autofluorescence (NIRAF), and the PGs between 1971 and 2020 was conducted on PubMed. EXPERT OPINION: NIRAF and near-infrared (NIR) parathyroid angiography have emerged as promising and reliable techniques for intraoperative PG identification. NIRAF may aid with real-time identification of both normal and diseased PGs and reduce the risk of postoperative complications such as hypocalcemia. Further large prospective multicenter studies should be conducted in thyroid and parathyroid surgical patient populations to confirm the clinical efficacy of these intraoperative NIR-based PG detection techniques.

Ultrasound Bone Segmentation: A Scoping Review of Techniques and Validation Practices.

Ultrasound bone segmentation is an important yet challenging task for many clinical applications. Several works have emerged attempting to improve and automate bone segmentation, which has led to a variety of computational techniques, validation practices and applied clinical scenarios. We characterize this exciting and growing body of research by reviewing published ultrasound bone segmentation techniques. We review 56 articles in detail and categorize and discuss the image analysis techniques that have been used for bone segmentation. We highlight the general trends of this field in terms of clinical motivation, image analysis techniques, ultrasound modalities and the types of validation practices used to quantify segmentation performance. Finally, we present an outlook on promising areas of research based on the unaddressed needs for solving ultrasound bone segmentation.