Kasai Portoenterostomy Delays: Is a HIDA Scan Worth the Wait?

BACKGROUND: Biliary Atresia (BA), an obstructive cholangiopathy, is the most common cause of end-stage liver disease and liver transplantation in children. Timely differentiation of BA from other causes of neonatal jaundice remains a challenge, yet is critical to improving outcomes. METHODS: Clinical characteristics including demographics, age at jaundice presentation, age at hepatobiliary scintigraphy, age at surgery, severity of liver fibrosis, and native-liver survival were reviewed in infants with hyperbilirubinemia and suspected BA for this single center retrospective cohort study. We investigated the accuracy of hepatobiliary scintigraphy as well as elapsed time from jaundice presentation to diagnostic intervention. RESULTS: BA was suspected in 234 infants. BA was identified in 17% of infants with hepatobiliary scintigraphy and 72% of infants who underwent operative exploration without hepatobiliary scintigraphy. Elapsed time from jaundice presentation to Kasai Portoenterostomy (KPE) for BA patients was 2.1x longer if hepatobiliary scintigraphy was obtained (p = 0.084). The mean age at KPE for this cohort was 66.8 days (n = 54), with a significantly higher mean age at KPE (75.2 days) for infants who were later listed or underwent liver transplantation (p = 0.038). Histologically, the lowest liver fibrosis scores were seen in infants undergoing KPE <30 days old and worsened significantly with increased age (p < 0.001). CONCLUSION: Hepatobiliary scintigraphy compared to operative exploration for the diagnostic evaluation of infants with suspected BA introduces significant time delays to KPE but enables avoidance of surgery in some infants. The temporal pattern of worsening cholestatic liver injury from BA with each day of increased age highlights the importance of intervening as early as possible for the best prognosis. TYPE OF STUDY: Retrospective study, Level of evidence: III.

Deep Learning-Based Abdominal Muscle Segmentation on CT Images of Surgical Patient Populations.

Computed tomography (CT) is commonly used for the characterization and tracking of abdominal muscle mass in surgical patients for both pre-surgical outcome predictions and post-surgical monitoring of response to therapy. In order to accurately track changes of abdominal muscle mass, radiologists must manually segment CT slices of patients, a time-consuming task with potential for variability. In this work, we combined a fully convolutional neural network (CNN) with high levels of preprocessing to improve segmentation quality. We utilized a CNN based approach to remove patients' arms and fat from each slice and then applied a series of registrations with a diverse set of abdominal muscle segmentations to identify a best fit mask. Using this best fit mask, we were able to remove many parts of the abdominal cavity, such as the liver, kidneys, and intestines. This preprocessing was able to achieve a mean Dice similarity coefficient (DSC) of 0.53 on our validation set and 0.50 on our test set by only using traditional computer vision techniques and no artificial intelligence. The preprocessed images were then fed into a similar CNN previously presented in a hybrid computer vision-artificial intelligence approach and was able to achieve a mean DSC of 0.94 on testing data. The preprocessing and deep learning-based method is able to accurately segment and quantify abdominal muscle mass on CT images.

Computational and image processing methods for analysis and automation of anatomical alignment and joint spacing in reconstructive surgery.

PURPOSE: Reconstructive surgeries to treat a number of musculoskeletal conditions, from arthritis to severe trauma, involve implant placement and reconstructive planning components. Anatomically matched 3D-printed implants are becoming increasingly patient-specific; however, the preoperative planning and design process requires several hours of manual effort from highly trained engineers and clinicians. Our work mitigates this problem by proposing algorithms for the automatic re-alignment of unhealthy anatomies, leading to more efficient, affordable, and scalable treatment solutions. METHODS: Our solution combines global alignment techniques such as iterative closest points with novel joint space refinement algorithms. The latter is achieved by a low-dimensional characterization of the joint space, computed from the distribution of the distance between adjacent points in a joint. RESULTS: Experimental validation is presented on real clinical data from human subjects. Compared with ground truth healthy anatomies, our algorithms can reduce misalignment errors by 22% in translation and 19% in rotation for the full foot-and-ankle and 37% in translation and 39% in rotation for the hindfoot only, achieving a performance comparable to expert technicians. CONCLUSION: Our methods and histogram-based metric allow for automatic and unsupervised alignment of anatomies along with techniques for global alignment of complex arrangements such as the foot-and-ankle system, a major step toward a fully automated and data-driven re-positioning, designing, and diagnosing tool.

Development of Algorithms for Automated Detection of Cervical Pre-Cancers With a Low-Cost, Point-of-Care, Pocket Colposcope.

GOAL: In this paper, we propose methods for (1) automatic feature extraction and classification for acetic acid and Lugol's iodine cervigrams and (2) methods for combining features/diagnosis of different contrasts in cervigrams for improved performance. METHODS: We developed algorithms to pre-process pathology-labeled cervigrams and extract simple but powerful color and textural-based features. The features were used to train a support vector machine model to classify cervigrams based on corresponding pathology for visual inspection with acetic acid, visual inspection with Lugol's iodine, and a combination of the two contrasts. RESULTS: The proposed framework achieved a sensitivity, specificity, and accuracy of 81.3%, 78.6%, and 80.0%, respectively, when used to distinguish cervical intraepithelial neoplasia (CIN+) relative to normal and benign tissues. This is superior to the average values achieved by three expert physicians on the same data set for discriminating normal/benign cases from CIN+ (77% sensitivity, 51% specificity, and 63% accuracy). CONCLUSION: The results suggest that utilizing simple color- and textural-based features from visual inspection with acetic acid and visual inspection with Lugol's iodine images may provide unbiased automation of cervigrams. SIGNIFICANCE: This would enable automated, expert-level diagnosis of cervical pre-cancer at the point of care.