Built to Last? Reproducibility and Reusability of Deep Learning Algorithms in Computational Pathology.

Recent progress in computational pathology has been driven by deep learning. While code and data availability are essential to reproduce findings from preceding publications, ensuring a deep learning model's reusability is more challenging. For that, the codebase should be well-documented and easy to integrate into existing workflows and models should be robust toward noise and generalizable toward data from different sources. Strikingly, only a few computational pathology algorithms have been reused by other researchers so far, let alone employed in a clinical setting. To assess the current state of reproducibility and reusability of computational pathology algorithms, we evaluated peer-reviewed articles available in PubMed, published between January 2019 and March 2021, in 5 use cases: stain normalization; tissue type segmentation; evaluation of cell-level features; genetic alteration prediction; and inference of grading, staging, and prognostic information. We compiled criteria for data and code availability and statistical result analysis and assessed them in 160 publications. We found that only one-quarter (41 of 160 publications) made code publicly available. Among these 41 studies, three-quarters (30 of 41) analyzed their results statistically, half of them (20 of 41) released their trained model weights, and approximately a third (16 of 41) used an independent cohort for evaluation. Our review is intended for both pathologists interested in deep learning and researchers applying algorithms to computational pathology challenges. We provide a detailed overview of publications with published code in the field, list reusable data handling tools, and provide criteria for reproducibility and reusability.

Automated MRI Lung Segmentation and 3D Morphologic Features for Quantification of Neonatal Lung Disease.

Purpose: To analyze the performance of deep learning (DL) models for segmentation of the neonatal lung in MRI and investigate the use of automated MRI-based features for assessment of neonatal lung disease. Materials and Methods: Quiet-breathing MRI was prospectively performed in two independent cohorts of preterm infants (median gestational age, 26.57 weeks; IQR, 25.3-28.6 weeks; 55 female and 48 male infants) with (n = 86) and without (n = 21) chronic lung disease (bronchopulmonary dysplasia [BPD]). Convolutional neural networks were developed for lung segmentation, and a three-dimensional reconstruction was used to calculate MRI features for lung volume, shape, pixel intensity, and surface. These features were explored as indicators of BPD and disease-associated lung structural remodeling through correlation with lung injury scores and multinomial models for BPD severity stratification. Results: The lung segmentation model reached a volumetric Dice coefficient of 0.908 in cross-validation and 0.880 on the independent test dataset, matching expert-level performance across disease grades. MRI lung features demonstrated significant correlations with lung injury scores and added structural information for the separation of neonates with BPD (BPD vs no BPD: average area under the receiver operating characteristic curve [AUC], 0.92 ± 0.02 [SD]; no or mild BPD vs moderate or severe BPD: average AUC, 0.84 ± 0.03). Conclusion: This study demonstrated high performance of DL models for MRI neonatal lung segmentation and showed the potential of automated MRI features for diagnostic assessment of neonatal lung disease while avoiding radiation exposure.Keywords: Bronchopulmonary Dysplasia, Chronic Lung Disease, Preterm Infant, Lung Segmentation, Lung MRI, BPD Severity Assessment, Deep Learning, Lung Imaging Biomarkers, Lung Topology Supplemental material is available for this article. Published under a CC BY 4.0 license.See also the commentary by Parraga and Sharma in this issue.

Escherichia coli Bacterobilia Is Associated with Severe Postoperative Pancreatic Fistula After Pancreaticoduodenectomy.

BACKGROUND: The benefit of preoperative biliary stenting in the treatment of pancreatic ductal adenocarcinoma is controversially debated. Data from recent meta-analyses favor primary surgery for the majority of resectable pancreatic cancers. Regardless of this evidence, preoperative biliary stenting via endoscopy (EBS) is commonly performed, often before involvement of a surgeon. The goal of this study was to elucidate the association of bile duct stenting, microbiological dislocation of gut flora to the biliary compartment, and major postoperative complications. METHODS: Patient data was derived from a prospectively maintained database including all pancreatic resections between January 2006 and December 2014. Patients receiving pancreaticoduodenectomy for malignant disease in the head of the pancreas with prior EBS were included. Microbiological data were obtained through conventional culture from intraoperative bile duct swabs. RESULTS: Two hundred ninety-eight patients were enrolled in this study. Severe postoperative complications were associated with stent colonization: Postoperative pancreatic fistula type C occurred more frequently in E. coli-colonized patients (sample estimated odds ratio (OR) = 4.07), and the rate of lymphatic fistula was elevated in Enterococcus-colonized patients (OR = 3.25). Longer stenting duration (> 16 days) was associated with the prevalence of these bacteria. CONCLUSION: Major surgical complications following pancreaticoduodenectomy, including severe pancreatic fistula, are associated with bacterobilia after EBS. The indication for bile duct stenting should be evaluated in a multidisciplinary setting.

Aggregation behavior of rodlike ionogenic polymers in chloroform.

The rodlike ionogenic polymers poly(p-pyridylene-phenylene) and poly(p-pyridylene/phenylene-ethynylene) form polyelectrolytes when protonated with toluene sulfonic acid or ethane sulfonic acid in chloroform solution. This molecular modification, clearly indicated by a marked red shift of the UV absorption band, induces the formation of prolate, bundlelike aggregates, whose size and shape are obtained from their rotational dynamics as revealed by electric birefringence relaxation and their translational dynamics as measured by dynamic light scattering. The aggregates have a length of 400-600 nm and a high aspect ratio >15. In general, the polyelectrolyte molecules are arranged with their long axes parallel to the long axis of the aggregates. They probably attract each other through the electrostatic interaction with counterions. The counterions are not bound to specific sites but may be shifted under the action of an external electric field to account for the highly anisotropic electric polarizability. When inert salt or excess sulfonic acid is added, these compounds seem to accumulate within the aggregates and influence the attractive forces. This is generally leading to an elongation of the aggregates and, in the case of added salts, even to a marked reduction of birefringence.