Correlative Assessment of Machine Learning-Based Cobb Angle Measurements and Human-Based Measurements in Adolescent Idiopathic and Congenital Scoliosis.

Purpose: Scoliosis is a complex spine deformity with direct functional and cosmetic impacts on the individual. The reference standard for assessing scoliosis severity is the Cobb angle which is measured on radiographs by human specialists, carrying interobserver variability and inaccuracy of measurements. These limitations may result in lack of timely referral for management at a time the scoliotic deformity progression can be saved from surgery. We aimed to create a machine learning (ML) model for automatic calculation of Cobb angles on 3-foot standing spine radiographs of children and adolescents with clinical suspicion of scoliosis across 2 clinical scenarios (idiopathic, group 1 and congenital scoliosis, group 2). Methods: We retrospectively measured Cobb angles of 130 patients who had a 3-foot spine radiograph for scoliosis within a 10-year period for either idiopathic or congenital anomaly scoliosis. Cobb angles were measured both manually by radiologists and by an ML pipeline (segmentation-based approach-Augmented U-Net model with non-square kernels). Results: Our Augmented U-Net architecture achieved a Symmetric Mean Absolute Percentage Error (SMAPE) of 11.82% amongst a combined idiopathic and congenital scoliosis cohort. When stratifying for idiopathic and congenital scoliosis individually a SMAPE of 13.02% and 11.90% were achieved, respectively. Conclusion: The ML model used in this study is promising at providing automated Cobb angle measurement in both idiopathic scoliosis and congenital scoliosis. Nevertheless, larger studies are needed in the future to confirm the results of this study prior to translation of this ML algorithm into clinical practice.

Machine Learning Detection and Characterization of Splenic Injuries on Abdominal Computed Tomography.

BACKGROUND: Multi-detector contrast-enhanced abdominal computed tomography (CT) allows for the accurate detection and classification of traumatic splenic injuries, leading to improved patient management. Their effective use requires rapid study interpretation, which can be a challenge on busy emergency radiology services. A machine learning system has the potential to automate the process, potentially leading to a faster clinical response. This study aimed to create such a system. METHOD: Using the American Association for the Surgery of Trauma (AAST), spleen injuries were classified into 3 classes: normal, low-grade (AAST grade I-III) injuries, and high-grade (AAST grade IV and V) injuries. Employing a 2-stage machine learning strategy, spleens were initially segmented from input CT images and subsequently underwent classification via a 3D dense convolutional neural network (DenseNet). RESULTS: This single-centre retrospective study involved trauma protocol CT scans performed between January 1, 2005, and July 31, 2021, totaling 608 scans with splenic injuries and 608 without. Five board-certified fellowship-trained abdominal radiologists utilizing the AAST injury scoring scale established ground truth labels. The model achieved AUC values of 0.84, 0.69, and 0.90 for normal, low-grade injuries, and high-grade splenic injuries, respectively. CONCLUSIONS: Our findings demonstrate the feasibility of automating spleen injury detection using our method with potential applications in improving patient care through radiologist worklist prioritization and injury stratification. Future endeavours should concentrate on further enhancing and optimizing our approach and testing its use in a real-world clinical environment.

Automated Adolescence Scoliosis Detection Using Augmented U-Net With Non-square Kernels.

Purpose: Scoliosis is a deformity of the spine, and as a measure of scoliosis severity, Cobb angle is fundamental to the diagnosis of deformities that require treatment. Conventional Cobb angle measurement and assessment is usually done manually, which is inherently time-consuming, and associated with high inter- and intra-observer variability. While there exist automatic scoliosis measurement methods, they suffer from insufficient accuracy. In this work, we propose a two-step segmentation-based deep learning architecture to automate Cobb angle measurement for scoliosis assessment using X-Ray images. Methods: The proposed architecture involves two steps. In the first step, we utilize a novel Augmented U-Net architecture to generate segmentations of vertebrae. The second step includes a non-learning-based pipeline to extract landmark coordinates from the segmented vertebrae and filter undesirable landmarks. Results: Our proposed Augmented U-Net architecture achieved a Symmetric Mean Absolute Percentage Error of 9.2%, with approximately 90% of estimations having less than 10 degrees difference compared with the AASCE-MICCAI challenge 2019 dataset ground truths. We further validated the model using an internal dataset and achieved almost the same level of performance. Conclusion: The proposed architecture is robust in providing automated spinal vertebrae segmentations and Cobb angle measurement, and is potentially generalizable to real-world clinical settings.

A Novel Ureteroscopy Training Platform That Utilizes CT Urograms to Replicate Complex Renal Collecting System Anatomies.

Introduction: Flexible ureteroscopy (fURS) is a one-person surgical technique, limiting trainees' ability to practice intraoperatively. Although well suited for simulation training, few existing fURS simulators can accurately reproduce complex renal collecting system anatomies. We developed an anatomically accurate fURS simulator using three-dimensional (3D) reconstruction of CT urograms and 3D printing technology to address this need. Materials and Methods: Patient-specific CT urograms were used to create 3D reconstruction of the renal collecting system using Slicer™. 3D models were modified using Blender™. Hollow, elastomer kidney models were created using an Objet 3D™ printer. To test and evaluate the new fURS simulator, 25 volunteers were recruited (5 novices, 13 residents, and 7 urologists). Participants were asked to explore the model with fURS and were evaluated on their ability to deduce its 3D anatomy, their ability to navigate to prespecified calices, and their time to task completion. Furthermore, participants were asked to compare the anatomical model with existing fURS benchtop models (Cook Medical™ and Limbs & Things™) on several criteria, including internal visualization, tactile feedback, and overall functional and teaching fidelity, in a survey. Results: We were able to create a fURS simulator that accurately replicates anatomically complex renal collecting systems. In exploring the model, we noted that unlike staff urologists, novices and residents often completely missed lower pole calices. A survey comparison between our simulator and comparable benchtop simulators revealed consistently better ratings of our simulator on all criteria (p < 0.05). Conclusions: We were able to create an anatomically accurate fURS simulator that provides a more realistic scoping experience. Preliminary testing revealed that trainees will benefit from this simulator, particularly with respect to learning how to navigate challenging collecting systems.

Video assessment of laparoscopic skills by novices and experts: implications for surgical education.

BACKGROUND: Previous investigators have shown that novices are able to assess surgical skills as reliably as expert surgeons. The purpose of this study was to determine how novices and experts arrive at these graded scores when assessing laparoscopic skills and the potential implications this may have for surgical education. METHODS: Four novices and four general laparoscopic surgeons evaluated 59 videos of a suturing task using a 5-point scale. Average novice and expert evaluator scores for each video and the average number of times that scores were changed were compared. Intraclass correlation coefficients were used to determine inter-rater and test-retest reliability. Evaluators were asked to define the number of videos they needed to watch before they could confidently grade and to describe how they were able to distinguish between different levels of expertise. RESULTS: There were no significant differences in mean scores assigned by the two evaluator groups. Novices changed their scores more frequently compared to experts, but this did not reach statistical significance. There was excellent inter-rater reliability between the two groups (ICC = 0.91, CI 0.85-0.95) and good test-retest reliability (ICC > 0.83). On average, novices and experts reported that they needed to watch 13.8 ± 2.4 and 8.5 ± 2.5 videos, respectively, before they could confidently grade. Both groups also identified similar qualitative indicators (e.g., instrument control). CONCLUSION: Evaluators with varying levels of expertise can reliably grade performance of an intracorporeal suturing task. While novices were less confident in their grading, both groups were able to assign comparable scores and identify similar elements of a suturing skill as being important in terms of assessment.