Exploring the potential of high-resolution LC-MS in combination with ion mobility separation and surrogate minimal depth for enhanced almond origin authentication

Almonds (Prunus dulcisMill.) are consumed worldwide and their geographical origin plays a crucial role in determining their market value. In the present study, a total of 250 almond reference samples from six countries (Australia, Spain, Iran, Italy, Morocco, and the USA) were non-polar extracted and analyzed by UPLC-ESI-IM-qToF-MS. Four harvest periods, more than 30 different varieties, including both sweet and bitter almonds, were considered in the method development. Principal component analysis showed that there are three groups of samples with similarities: Australia/USA, Spain/Italy and Iran/Morocco. For origin determination, a random forest achieved an accuracy of 88.8 %. Misclassifications occurred mainly between almonds from the USA and Australia, due to similar varieties and similar external influences such as climate conditions. Metabolites relevant for classification were selected using Surrogate Minimal Depth, with triacylglycerides containing oxidized, odd chained or short chained fatty acids and some phospholipids proven to be the most suitable marker substances. Our results show that focusing on the identified lipids (e. g., using a QqQ-MS instrument) is a promising approach to transfer the origin determination of almonds to routine analysis.

Development and validation of a machine learning model for prediction of type 2 diabetes in patients with mental illness

BACKGROUND

Type 2 diabetes (T2D) is approximately twice as common among individuals with mental illness compared with the background population, but may be prevented by early intervention on lifestyle, diet, or pharmacologically. Such prevention relies on identification of those at elevated risk (prediction). The aim of this study was to develop and validate a machine learning model for prediction of T2D among patients with mental illness.

METHODS

The study was based on routine clinical data from electronic health records from the psychiatric services of the Central Denmark Region. A total of 74,880 patients with 1.59 million psychiatric service contacts were included in the analyses. We created 1343 potential predictors from 51 source variables, covering patient-level information on demographics, diagnoses, pharmacological treatment, and laboratory results. T2D was operationalised as HbA1c ≥48 mmol/mol, fasting plasma glucose ≥7.0 mmol/mol, oral glucose tolerance test ≥11.1 mmol/mol or random plasma glucose ≥11.1 mmol/mol. Two machine learning models (XGBoost and regularised logistic regression) were trained to predict T2D based on 85% of the included contacts. The predictive performance of the best performing model was tested on the remaining 15% of the contacts.

RESULTS

The XGBoost model detected patients at high risk 2.7 years before T2D, achieving an area under the receiver operating characteristic curve of 0.84. Of the 996 patients developing T2D in the test set, the model issued at least one positive prediction for 305 (31%).

CONCLUSION

A machine learning model can accurately predict development of T2D among patients with mental illness based on routine clinical data from electronic health records. A decision support system based on such a model may inform measures to prevent development of T2D in this high-risk population.

AcquisitionFocus: Joint Optimization of Acquisition Orientation and Cardiac Volume Reconstruction Using Deep Learning

In cardiac cine imaging, acquiring high-quality data is challenging and time-consuming due to the artifacts generated by the heart's continuous movement. Volumetric, fully isotropic data acquisition with high temporal resolution is, to date, intractable due to MR physics constraints. To assess whole-heart movement under minimal acquisition time, we propose a deep learning model that reconstructs the volumetric shape of multiple cardiac chambers from a limited number of input slices while simultaneously optimizing the slice acquisition orientation for this task. We mimic the current clinical protocols for cardiac imaging and compare the shape reconstruction quality of standard clinical views and optimized views. In our experiments, we show that the jointly trained model achieves accurate high-resolution multi-chamber shape reconstruction with errors of <13 mm HD95 and Dice scores of >80%, indicating its effectiveness in both simulated cardiac cine MRI and clinical cardiac MRI with a wide range of pathological shape variations.

Anatomy-guided domain adaptation for 3D in-bed human pose estimation

3D human pose estimation is a key component of clinical monitoring systems. The clinical applicability of deep pose estimation models, however, is limited by their poor generalization under domain shifts along with their need for sufficient labeled training data. As a remedy, we present a novel domain adaptation method, adapting a model from a labeled source to a shifted unlabeled target domain. Our method comprises two complementary adaptation strategies based on prior knowledge about human anatomy. First, we guide the learning process in the target domain by constraining predictions to the space of anatomically plausible poses. To this end, we embed the prior knowledge into an anatomical loss function that penalizes asymmetric limb lengths, implausible bone lengths, and implausible joint angles. Second, we propose to filter pseudo labels for self-training according to their anatomical plausibility and incorporate the concept into the Mean Teacher paradigm. We unify both strategies in a point cloud-based framework applicable to unsupervised and source-free domain adaptation. Evaluation is performed for in-bed pose estimation under two adaptation scenarios, using the public SLP dataset and a newly created dataset. Our method consistently outperforms various state-of-the-art domain adaptation methods, surpasses the baseline model by 31%/66%, and reduces the domain gap by 65%/82%. Source code is available at https://github.com/multimodallearning/da-3dhpe-anatomy.

Lexical stability of psychiatric clinical notes from electronic health records over a decade

OBJECTIVE

Natural language processing (NLP) methods hold promise for improving clinical prediction by utilising information otherwise hidden in the clinical notes of electronic health records. However, clinical practice - as well as the systems and databases in which clinical notes are recorded and stored - change over time. As a consequence, the content of clinical notes may also change over time, which could degrade the performance of prediction models. Despite its importance, the stability of clinical notes over time has rarely been tested.

METHODS

The lexical stability of clinical notes from the Psychiatric Services of the Central Denmark Region in the period from January 1, 2011, to November 22, 2021 (a total of 14,811,551 clinical notes describing 129,570 patients) was assessed by quantifying sentence length, readability, syntactic complexity and clinical content. Changepoint detection models were used to estimate potential changes in these metrics.

RESULTS

We find lexical stability of the clinical notes over time, with minor deviations during the COVID-19 pandemic. Out of 2988 data points, 17 possible changepoints (corresponding to 0.6%) were detected. The majority of these were related to the discontinuation of a specific note type.

CONCLUSION

We find lexical and syntactic stability of clinical notes from psychiatric services over time, which bodes well for the use of NLP for predictive modelling in clinical psychiatry.

Robust and Realtime Large Deformation Ultrasound Registration Using End-to-End Differentiable Displacement Optimisation

Image registration for temporal ultrasound sequences can be very beneficial for image-guided diagnostics and interventions. Cooperative human-machine systems that enable seamless assistance for both inexperienced and expert users during ultrasound examinations rely on robust, realtime motion estimation. Yet rapid and irregular motion patterns, varying image contrast and domain shifts in imaging devices pose a severe challenge to conventional realtime registration approaches. While learning-based registration networks have the promise of abstracting relevant features and delivering very fast inference times, they come at the potential risk of limited generalisation and robustness for unseen data; in particular, when trained with limited supervision. In this work, we demonstrate that these issues can be overcome by using end-to-end differentiable displacement optimisation. Our method involves a trainable feature backbone, a correlation layer that evaluates a large range of displacement options simultaneously and a differentiable regularisation module that ensures smooth and plausible deformation. In extensive experiments on public and private ultrasound datasets with very sparse ground truth annotation the method showed better generalisation abilities and overall accuracy than a VoxelMorph network with the same feature backbone, while being two times faster at inference.

Learn2Reg: Comprehensive Multi-Task Medical Image Registration Challenge, Dataset and Evaluation in the Era of Deep Learning

Image registration is a fundamental medical image analysis task, and a wide variety of approaches have been proposed. However, only a few studies have comprehensively compared medical image registration approaches on a wide range of clinically relevant tasks. This limits the development of registration methods, the adoption of research advances into practice, and a fair benchmark across competing approaches. The Learn2Reg challenge addresses these limitations by providing a multi-task medical image registration data set for comprehensive characterisation of deformable registration algorithms. A continuous evaluation will be possible at https://learn2reg.grand-challenge.org. Learn2Reg covers a wide range of anatomies (brain, abdomen, and thorax), modalities (ultrasound, CT, MR), availability of annotations, as well as intra- and inter-patient registration evaluation. We established an easily accessible framework for training and validation of 3D registration methods, which enabled the compilation of results of over 65 individual method submissions from more than 20 unique teams. We used a complementary set of metrics, including robustness, accuracy, plausibility, and runtime, enabling unique insight into the current state-of-the-art of medical image registration. This paper describes datasets, tasks, evaluation methods and results of the challenge, as well as results of further analysis of transferability to new datasets, the importance of label supervision, and resulting bias. While no single approach worked best across all tasks, many methodological aspects could be identified that push the performance of medical image registration to new state-of-the-art performance. Furthermore, we demystified the common belief that conventional registration methods have to be much slower than deep-learning-based methods.

CrossMoDA 2021 challenge: Benchmark of cross-modality domain adaptation techniques for vestibular schwannoma and cochlea segmentation

Domain Adaptation (DA) has recently been of strong interest in the medical imaging community. While a large variety of DA techniques have been proposed for image segmentation, most of these techniques have been validated either on private datasets or on small publicly available datasets. Moreover, these datasets mostly addressed single-class problems. To tackle these limitations, the Cross-Modality Domain Adaptation (crossMoDA) challenge was organised in conjunction with the 24th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2021). CrossMoDA is the first large and multi-class benchmark for unsupervised cross-modality Domain Adaptation. The goal of the challenge is to segment two key brain structures involved in the follow-up and treatment planning of vestibular schwannoma (VS): the VS and the cochleas. Currently, the diagnosis and surveillance in patients with VS are commonly performed using contrast-enhanced T1 (ceT₁) MR imaging. However, there is growing interest in using non-contrast imaging sequences such as high-resolution T2 (hrT₂) imaging. For this reason, we established an unsupervised cross-modality segmentation benchmark. The training dataset provides annotated ceT₁ scans (N=105) and unpaired non-annotated hrT₂ scans (N=105). The aim was to automatically perform unilateral VS and bilateral cochlea segmentation on hrT₂ scans as provided in the testing set (N=137). This problem is particularly challenging given the large intensity distribution gap across the modalities and the small volume of the structures. A total of 55 teams from 16 countries submitted predictions to the validation leaderboard. Among them, 16 teams from 9 different countries submitted their algorithm for the evaluation phase. The level of performance reached by the top-performing teams is strikingly high (best median Dice score - VS: 88.4%; Cochleas: 85.7%) and close to full supervision (median Dice score - VS: 92.5%; Cochleas: 87.7%). All top-performing methods made use of an image-to-image translation approach to transform the source-domain images into pseudo-target-domain images. A segmentation network was then trained using these generated images and the manual annotations provided for the source image.

Stability of diagnostic coding of psychiatric outpatient visits across the transition from the second to the third version of the Danish National Patient Registry

OBJECTIVE

In Denmark, data on hospital contacts are reported to the Danish National Patient Registry (DNPR). The ICD-10 main diagnoses from the DNPR are often used as proxies for mental disorders in psychiatric research. With the transition from the second version of the DNPR (DNPR2) to the third (DNPR3) in February-March 2019, the way main diagnoses are coded in relation to outpatient treatment changed substantially. Specifically, in the DNPR2, each outpatient treatment course was labelled with only one main diagnosis. In the DNPR3, however, each visit during an outpatient treatment course is labelled with a main diagnosis. We assessed whether this change led to a break in the diagnostic time-series represented by the DNPR, which would pose a threat to the research relying on this source.

METHODS

All main diagnoses from outpatients attending the Psychiatric Services of the Central Denmark Region from 2013 to 2021 (n = 100,501 unique patients) were included in the analyses. The stability of the DNPR diagnostic time-series at the ICD-10 subchapter level was examined by comparing means across the transition from the DNPR2 to the DNPR3.

RESULTS

While the proportion of psychiatric outpatients with diagnoses from some ICD-10 subchapters changed statistically significantly from the DNPR2 to the DNPR3, the changes were small in absolute terms (e.g., +0.6% for F2-psychotic disorders and +0.6% for F3-mood disorders).

CONCLUSION

The change from the DNPR2 to the DNPR3 is unlikely to pose a substantial threat to the validity of most psychiatric research at the diagnostic subchapter level.

Predicting involuntary admission among patients with psychotic disorder

Introduction

Involuntary admissions are increasing in numbers across Europe.¹ They can be traumatic for the patients² and are associated with large societal costs.³ Individuals with psychotic disorder are at particularly elevated risk of involuntary admission.

Objectives

This study aims to investigate whether machine learning methods including natural language processing can predict involuntary admission among patients with psychotic disorder.

Methods

We have obtained a dataset based on electronic health records for all patients having had at least one contact with the psychiatric services in the Central Denmark Region from 2011 to 2021. This dataset covers more than 120,000 patients, of which approximately 10,000 have been diagnosed with a psychotic disorder. The dataset contains both structured data, such as diagnoses, blood tests etc., as well as unstructured data (text). We will train machine learning models, basic logistic regression-models as well as state-of-the-art neural networks, to predict involuntary admission after contacts to the psychiatric services.

Results

As the machine learning models are under development, no results are available at this time. Preliminary results are expected in spring 2022.

Conclusions

If involuntary admission can be predicted among patients with psychotic disorder based on data from electronic health records, it will pave the way for potentially preventive interventions. References: 1. Sheridans-Rains, L et al., 2019 2. Frueh, B.C et al., 2005 3. Smith,S., 2020

Disclosure

No significant relationships.

Learning a Metric for Multimodal Medical Image Registration without Supervision Based on Cycle Constraints

Deep learning based medical image registration remains very difficult and often fails to improve over its classical counterparts where comprehensive supervision is not available, in particular for large transformations-including rigid alignment. The use of unsupervised, metric-based registration networks has become popular, but so far no universally applicable similarity metric is available for multimodal medical registration, requiring a trade-off between local contrast-invariant edge features or more global statistical metrics. In this work, we aim to improve over the use of handcrafted metric-based losses. We propose to use synthetic three-way (triangular) cycles that for each pair of images comprise two multimodal transformations to be estimated and one known synthetic monomodal transform. Additionally, we present a robust method for estimating large rigid transformations that is differentiable in end-to-end learning. By minimising the cycle discrepancy and adapting the synthetic transformation to be close to the real geometric difference of the image pairs during training, we successfully tackle intra-patient abdominal CT-MRI registration and reach performance on par with state-of-the-art metric-supervision and classic methods. Cyclic constraints enable the learning of cross-modality features that excel at accurate anatomical alignment of abdominal CT and MRI scans.

A generalizable speech emotion recognition model reveals depression and remission

OBJECTIVE

Affective disorders are associated with atypical voice patterns; however, automated voice analyses suffer from small sample sizes and untested generalizability on external data. We investigated a generalizable approach to aid clinical evaluation of depression and remission from voice using transfer learning: We train machine learning models on easily accessible non-clinical datasets and test them on novel clinical data in a different language.

METHODS

A Mixture of Experts machine learning model was trained to infer happy/sad emotional state using three publicly available emotional speech corpora in German and US English. We examined the model's predictive ability to classify the presence of depression on Danish speaking healthy controls (N = 42), patients with first-episode major depressive disorder (MDD) (N = 40), and the subset of the same patients who entered remission (N = 25) based on recorded clinical interviews. The model was evaluated on raw, de-noised, and speaker-diarized data.

RESULTS

The model showed separation between healthy controls and depressed patients at the first visit, obtaining an AUC of 0.71. Further, speech from patients in remission was indistinguishable from that of the control group. Model predictions were stable throughout the interview, suggesting that 20-30 s of speech might be enough to accurately screen a patient. Background noise (but not speaker diarization) heavily impacted predictions.

CONCLUSION

A generalizable speech emotion recognition model can effectively reveal changes in speaker depressive states before and after remission in patients with MDD. Data collection settings and data cleaning are crucial when considering automated voice analysis for clinical purposes.

Seeing under the cover with a 3D U-Net: point cloud-based weight estimation of covered patients

PURPOSE

Body weight is a crucial parameter for patient-specific treatments, particularly in the context of proper drug dosage. Contactless weight estimation from visual sensor data constitutes a promising approach to overcome challenges arising in emergency situations. Machine learning-based methods have recently been shown to perform accurate weight estimation from point cloud data. The proposed methods, however, are designed for controlled conditions in terms of visibility and position of the patient, which limits their practical applicability. In this work, we aim to decouple accurate weight estimation from such specific conditions by predicting the weight of covered patients from voxelized point cloud data.

METHODS

We propose a novel deep learning framework, which comprises two 3D CNN modules solving the given task in two separate steps. First, we train a 3D U-Net to virtually uncover the patient, i.e. to predict the patient's volumetric surface without a cover. Second, the patient's weight is predicted from this 3D volume by means of a 3D CNN architecture, which we optimized for weight regression.

RESULTS

We evaluate our approach on a lying pose dataset (SLP) under two different cover conditions. The proposed framework considerably improves on the baseline model by up to [Formula: see text] and reduces the gap between the accuracy of weight estimates for covered and uncovered patients by up to [Formula: see text].

CONCLUSION

We present a novel pipeline to estimate the weight of patients, which are covered by a blanket. Our approach relaxes the specific conditions that were required for accurate weight estimates by previous contactless methods and thus constitutes an important step towards fully automatic weight estimation in clinical practice.

GraphRegNet: Deep Graph Regularisation Networks on Sparse Keypoints for Dense Registration of 3D Lung CTs

In the last two years learning-based methods have started to show encouraging results in different supervised and unsupervised medical image registration tasks. Deep neural networks enable (near) real time applications through fast inference times and have tremendous potential for increased registration accuracies by task-specific learning. However, estimation of large 3D deformations, for example present in inhale to exhale lung CT or interpatient abdominal MRI registration, is still a major challenge for the widely adopted U-Net-like network architectures. Even when using multi-level strategies, current state-of-the-art DL registration results do not yet reach the high accuracy of conventional frameworks. To overcome the problem of large deformations for deep learning approaches, in this work, we present GraphRegNet, a sparse keypoint-based geometric network for dense deformable medical image registration. Similar to the successful 2D optical flow estimation of FlowNet or PWC-Net we leverage discrete dense displacement maps to facilitate the registration process. In order to cope with enormously increasing memory requirements when working with displacement maps in 3D medical volumes and to obtain a well-regularised and accurate deformation field we 1) formulate the registration task as the prediction of displacement vectors on a sparse irregular grid of distinctive keypoints and 2) introduce our efficient GraphRegNet for displacement regularisation, a combination of convolutional and graph neural network layers in a unified architecture. In our experiments on exhale to inhale lung CT registration we demonstrate substantial improvements (TRE below 1.4 mm) over other deep learning methods. Our code is publicly available at https://github.com/multimodallearning/graphregnet.

Prevalence, geographic distribution, and impact on lifespan of a dilated cardiomyopathy-associated RNA-binding motif protein 20 variant in genotyped dogs

INTRODUCTION

The study objectives were to determine the prevalence and geographic distribution of a dilated cardiomyopathy (DCM)-associated RNA-binding motif protein 20 (RBM20) variant in canine DNA samples submitted for testing and to evaluate the influence of the genotype on cardiac phenotype and lifespan.

ANIMALS

Samples from 2136 dogs including 1834 Standard Schnauzers (SSNZ), 266 Giant Schnauzers (GSNZ), and 36 dogs of other breeds.

METHODS

The University of Missouri Canine Genetics Laboratory's sample-accession spreadsheet and Orthopedic Foundation for Animals' database were retrospectively reviewed for samples submitted for RBM20 genotyping from November, 2013, through May, 2018. Data analyzed included breed, date of birth, RBM20 genotype (homozygous wild-type, heterozygous variant [HET], or homozygous variant [HOM]), geographic origin of submission, pedigree, cardiac phenotype, and date of death or current age if alive.

RESULTS AND DISCUSSION

The RBM20 variant was only detected in SSNZ and GSNZ. A total of 389 SSNZ were variant-positive (prevalence = 21.2%), with 361 HET (19.7%) and 28 HOM (1.5%). Of the HOM SSNZ, DCM was confirmed in 26 of 28 (92.9%), with the remainder lost to follow-up. The median lifespan of HOM SSNZ (3.06 years) was significantly shorter than that for HET (15.11 years) and wild-type (15.18 years) SSNZ. Twenty-six GSNZ were variant-positive (prevalence = 9.8%), with 23 HET (8.6%) and three HOM (1.1%). Nine GSNZ belonged to one family, including the three HOM GSNZ that all had DCM.

CONCLUSIONS

The HOM genotype is associated with DCM and premature death in SSNZ and GSNZ.

Epidemiology and transmission characteristics of early COVID-19 cases, 20 January-19 March 2020, in Bavaria, Germany

Severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) led to a significant disease burden and disruptions in health systems. We describe the epidemiology and transmission characteristics of early coronavirus disease 2019 (COVID-19) cases in Bavaria, Germany. Cases were reverse transcription polymerase chain reaction (RT-PCR)-confirmed SARS-CoV-2 infections, reported from 20 January-19 March 2020. The incubation period was estimated using travel history and date of symptom onset. To estimate the serial interval, we identified pairs of index and secondary cases. By 19 March, 3546 cases were reported. A large proportion was exposed abroad (38%), causing further local transmission. Median incubation period of 256 cases with exposure abroad was 3.8 days (95%CI: 3.5-4.2). For 95% of infected individuals, symptom onset occurred within 10.3 days (95%CI: 9.1-11.8) after exposure. The median serial interval, using 53 pairs, was 3.5 days (95%CI: 3.0-4.2; mean: 3.9, s.d.: 2.2). Travellers returning to Germany had an important influence on the spread of SARS-CoV-2 infections in Bavaria in early 2020. Especially in times of low incidence, public health agencies should identify holiday destinations, and areas with ongoing local transmission, to monitor potential importation of SARS-CoV-2 infections. Travellers returning from areas with ongoing community transmission should be advised to quarantine to prevent re-introductions of COVID-19.

Weakly-supervised learning of multi-modal features for regularised iterative descent in 3D image registration

Methods for deep learning based medical image registration have only recently approached the quality of classical model-based image alignment. The dual challenge of both a very large trainable parameter space and often insufficient availability of expert supervised correspondence annotations has led to slower progress compared to other domains such as image segmentation. Yet, image registration could also more directly benefit from an iterative solution than segmentation. We therefore believe that significant improvements, in particular for multi-modal registration, can be achieved by disentangling appearance-based feature learning and deformation estimation. In this work, we examine an end-to-end trainable, weakly-supervised deep learning-based feature extraction approach that is able to map the complex appearance to a common space. Our results on thoracoabdominal CT and MRI image registration show that the proposed method compares favourably well to state-of-the-art hand-crafted multi-modal features, Mutual Information-based approaches and fully-integrated CNN-based methods - and handles even the limitation of small and only weakly-labeled training data sets.

A Novel Strategy for Automatic Error Classification and Error Recovery for Robotic Assembly in Flexible Production

In this article, we develop a novel strategy for automatic error classification and recovery in robotic assembly tasks. The strategy does not require error diagnosis. It allows for effective reduction of an undetermined number of error states to 4, without the need for further operator updates of error space. The strategy integrates existing methods for computer vision, active vision and active manipulation. Our solution is implemented in a generic software framework, which is independent from software and hardware for implementing error detection and allows for application in other assembly types and components. The value of our strategy was experimentally validated on a simple case, where we inserted a battery into a cell phone. The experiment was performed on 1500 assembly attempts and included 500 detected errors. The whole experiment ran for 42 hours, with no need for operator assistance or supervision. The resulting classification rate is 99.6% and the resulting recovery rate is 98.8%. The 6 unrecovered errors were successfully resolved in a successive assembly attempt.

Capture-recapture study to estimate the true incidence of tuberculosis in Portugal, 2018

Accurate estimation of the true number of cases of an infectious disease is essential to plan and efficiently allocate available resources. This study aims to improve the Portuguese surveillance system for tuberculosis (TB) by identifying gaps in TB epidemiological surveillance at the national level. We estimated annual TB incidence using a capture-recapture method to assess the sensitivity of national TB surveillance.

Using probabilistic record linkage between two data sources, the National Epidemiological Surveillance System (SINAVE) and National Tuberculosis Program Surveillance System (SVIG-TB), we extracted TB diagnosed cases data for calendar year 2018. All reported TB cases were included, classified as confirmed, probable or possible. A two-source capture-recapture analysis using a log-linear model was performed to estimate the number of unobserved TB cases in Portugal and of the proportion identified by the current TB surveillance system.

Between the two datasets, we found 896 TB cases (of a total of 2170 cases) that could not be matched (37.5% SINAVE only, 62.5% SVIG only). Based on the log-linear model, it was estimated that there were 148 unobserved TB cases (95% confidence interval 127.96 - 171.31). Therefore, the estimated true number of TB cases in 2018 is 2318, so current surveillance has a sensitivity of 93.6%. Based on these findings, the TB incidence in Portugal is estimated to be 22.55 cases per 100 000 inhabitants.

Capture-recapture methods are useful in estimating annual TB incidence in high-resource settings. Although the two TB surveillance systems capture the majority of TB cases in Portugal, we might still be underestimating the true number of TB cases. Because TB is a high impact infectious disease, precise incidence estimates are crucial to allocate treatment and prevention resources and guide health policies.

Key messages

CRC method showed that Portugal is a TB low incidence country. Epidemiological surveillance systems should have a high sensitivity in order to allocate efficiently resources available.

Deep Learning for Prediction of Diaphragm Activity from the Surface Electromyogram

The electrical activity of the diaphragm (EAdi) is a novel monitoring parameter for patients under assisted ventilation and is used for assessing the patient’s neural respiratory drive. It is recorded by an array of electrodes placed inside the esophagus at the level of the diaphragm. A noninvasive alternative is the measurement of the electromyogram by means of skin surface electrodes (sEMG). The respiratory sEMG signal, however, is subject to electrocardiographic interference and crosstalk from other muscles and may also pick up a different part of the muscular activity. In this work, we propose to use a deep neural network to predict the electrical activity of the diaphragm as measured by a nasogastric catheter from sEMG measurements. We use a ResNet based architecture and train the network to directly regress the EAdi as a supervised learning task - we further investigate a heatmap based regression approach. The proposed methods are evaluated on a clinical dataset consisting of 77 recordings from mechanically ventilated patients. For the direct regression task, the network’s predictions reach a Pearson correlation coefficient (PCC) of 0.818 with EAdi on the hold-out set. The heatmap regression increases the PCC to 0.830 while at the same time achieving a lower mean absolute error, indicating a superior performance. From our results we conclude that sEMG measurements may be used to predict the internal activity of the diaphragm as measured invasively using a nasogastric catheter.

[Content of a weekly morbidity and mortality conference in visceral surgery]

BACKGROUND

Morbidity and mortality (M&M) conferences are essential instruments for quality improvement in surgical departments; however, publications concerning the detailed contents of M&M conferences are rare and have not been published in the German language.

OBJECTIVE

Detailed analysis of the content of a weekly M&M conference in a department of general and visceral surgery.

MATERIAL AND METHODS

Data from a weekly M&M conference were prospectively collected. Epidemiological data, diagnosis, type of surgery, morbidity, postoperative course and mortality were documented for each patient. Type (surgical vs. medicinal) and severity (I-V according to Clavien-Dindo classification) of complications were analyzed.

RESULTS

From 1 January 2010 to 31 December 2014 a total of 761 out of 11,470 patients with a mean age of 62.2 (15.9) years were discussed in the M&M conferences. Of the M&M patients 39.4% were female, 88.9% showed surgical complications while 28.9% were diagnosed with a medical complication and 91 patients (12.0%) died. Complications were classified as Clavien-Dindo I: 27.9%, II: 10.5%, III: 37.7%, IV: 12% and V: 12.0%. Most surgical complications were classified as Clavien-Dindo I (30.0%) and Clavien-Dindo III (40.9%), medical complications were most often classified as Clavien-Dindo IV (29.6%) and V (34.6%). Wound healing impairment (41%), pulmonary complications (16.6%), anastomotic leakage (15.6%), septic (8.9%) and cardiac (8.0%) complications were discussed most often. Cardiac, pulmonary and septic complications were the main cause of morbidity in deceased patients.

CONCLUSION

The M&M conferences display a morbidity profile of each surgical department. Depending on the speciality and focus of a surgical department, the content of the M&M conferences will vary. Detailed knowledge about the content of M&M conferences enable specific measures to be taken to improve quality and patient safety.

Social inequality and fractures-secular trends in the Danish population: a case-control study

We examined links between markers of social inequality and fracture risk in the Danish population, demonstrating that high income and being married are associated with a significantly lower risk.

INTRODUCTION

We explored whether the risk of hip, humerus, and wrist fracture was associated with markers of inequality using data from Danish health registries.

METHODS

All patients 50 years or older with a primary hip (ICD10 S720, S721, S722, and S729) humerus (ICD10 S422, S423, S424, S425, S426, and S427), or wrist (ICD10: S52) fracture were identified from 1/1/1995 to 31/12/2011. Fracture patients were matched 1:1 by age, sex, and year of fracture, to a non-fracture control. Markers of inequality were as follows: income (fifths); marital status (married, divorced, widowed, or unmarried); area of residence (remote, rural, intermediate, or urban). Conditional logistic regression was used to investigate associations between these exposures, and risk of fracture, adjusting for covariates (smoking, alcohol, and Charlson co-morbidity). Interactions were fitted between exposure and covariates where appropriate.

RESULTS

A total of 189,838 fracture patients (37,500 hip, 45,602 humerus, and 106,736 wrist) and 189,838 controls were included. Mean age was 73.9 years (hip), 67.5 years (humerus), and 65.3 years (wrist). High income (5th quintile) was significantly associated with a lower odds ratio of all three fractures, compared to average income (3rd quintile). Married subjects had a significantly decreased odds ratio across all three fractures. However, no overall secular difference was observed regarding the influence of the markers of inequality.

CONCLUSION

In conclusion, we have demonstrated important, stable associations between social inequality, assessed using income, marital status, and area of residence, and fracture at the population level. These findings can inform approaches to healthcare, and suggest that much thought should be given to novel interventions aimed especially at those living alone, and ideally societal measures to reduce social inequality.

Marker free model-based radiostereometric analysis for evaluation of hip joint kinematics: A validation study

Objectives

To validate the precision of digitally reconstructed radiograph (DRR) radiostereometric analysis (RSA) and the model-based method (MBM) RSA with respect to benchmark marker-based (MM) RSA for evaluation of kinematics in the native hip joint.

Methods

Seven human cadaveric hemipelves were CT scanned and bone models were segmented. Tantalum beads were placed in the pelvis and proximal femoral bone. RSA recordings of the hips were performed during flexion, adduction and internal rotation. Stereoradiographic recordings were all analyzed with DRR, MBM and MM. Migration results for the MBM and DRR with respect to MM were compared. Precision was assessed as systematic bias (mean difference) and random variation (Pitman’s test for equal variance).

Results

A total of 288 dynamic RSA images were analyzed. Systematic bias for DRR and MBM with respect to MM in translations (p < 0.018 mm) and rotations (p < 0.009°) were approximately 0. Pitman’s test showed lower random variation in all degrees of freedom for DRR compared with MBM (p < 0.001).

Conclusion

Systematic error was approximately 0 for both DRR or MBM. However, precision of DRR was statistically significantly better than MBM. Since DRR does not require marker insertion it can be used for investigation of preoperative hip kinematics in comparison with the postoperative results after joint preserving hip surgery. 

Cite this article: L. Hansen, S. De Raedt, P. B. Jørgensen, B. Mygind-Klavsen, B. Kaptein, M. Stilling. Marker free model-based radiostereometric analysis for evaluation of hip joint kinematics: A validation study. Bone Joint Res 2018;7:379–387. DOI: 10.1302/2046-3758.76.BJR-2017-0268.R1.

Controlled population-based comparative study of USA and international adult [55-74] neurological deaths 1989-2014

Objectives

A population‐based controlled study to determine whether adult (55‐74 years) neurological disease deaths are continuing to rise and are there significant differences between America and the twenty developed countries 1989‐91 and 2012‐14.

Method

Total Neurological Deaths (TND) rates contrasted against control Cancer and Circulatory Disease Deaths (CDD) extrapolated from WHO data. Confidence intervals compare USA and the other countries over the period. The Over‐75's TND and population increases are examined as a context for the 55‐74 outcomes.

Results

Male neurological deaths rose >10% in eleven countries, the other countries average rose 20% the USA 43% over the period. Female neurological deaths rose >10% in ten counties, averaging 14%, the USA up 68%. USA male and female neurological deaths increased significantly more than twelve and seventeen countries, respectively. USA over‐75s population increased by 49%, other countries 56%. Other countries TND up 187% the USA rose fourfold. Male and female cancer and CDD fell in every country averaging 26% and 21%, respectively, and 64% and 67% for CDD. Male neurological rates rose significantly more than Cancer and CCD in every country; Female neurological deaths rose significantly more than cancer in 17 countries and every country for CDD. There was no significant correlation between increases in neurological deaths and decreases in control mortalities.

Conclusions

There are substantial increases in neurological deaths in most countries, significantly so in America. Rises in the 55‐74 and over‐75's rates are not primarily due to demographic changes and are a matter of concern warranting further investigation.

Immunization with HIV-1 envelope T20-encoding DNA vaccines elicits cross-clade neutralizing antibody responses

Background: Genetic immunization is expected to induce the expression of antigens in a native form. The encoded peptide epitopes are presented on endogenous MHC molecules, mimicking antigen presentation during a viral infection. We have explored the potential of enfuvirtide (T20), a short HIV peptide with antiviral properties, to enhance immune response to HIV antigens. To generate an expression vector, the T20 sequence was cloned into a conventional plasmid, the novel minicircle construct, and a replicon plasmid. In addition, 3 conventional plasmids that express the envelope of HIV-1 subtypes A, B and C and contain T20 in their gp41 sequences were also tested. Results: All combinations induced HIV-specific antibodies and cellular responses. The addition of T20 as a peptide and as an expression cassette in the 3 DNA vectors enhanced antibody responses. The highest anti-HIV-1 Env titers were obtained by the replicon T20 construct. This demonstrates that besides its known antiviral activity, T20 promotes immune responses. We also confirm that the combination of slightly divergent antigens improves immune responses. Conclusions: The antiretroviral T20 HIV-1 sequence can be used as an immunogen to elicit binding and neutralizing antibodies against HIV-1. These, or similarly modified gp41 genes/peptides, can be used as priming or boosting components for induction of broadly neutralizing anti-HIV antibodies. Future comparative studies will reveal the optimal mode of T20 administration.

Subjects with osteoporosis to remain at high risk for fracture despite benefit of prior bisphosphonate treatment-a Danish case-control study

The benefits of pharmaceutical treatment for osteoporosis may be limited for a number of patients, as they continue to experience fractures. Alternative treatments may be considered for subjects whom remain at high risk for future fractures.

INTRODUCTION

Previous studies have investigated the effects of good adherence to anti-osteoporosis medication. However, very few studies have described why some patients experience fractures and loss of BMD despite adherence to treatment. The aim of this study was to estimate the proportion of patients at high risk for fracture despite being compliant to bisphosphonate treatment and examine which factors influence why some osteoporotic patients remain at a high risk for fracture despite being compliant to bisphosphonate treatment.

METHODS

This case-control study is based on Danish national health registry data. The subjects had to have either one BMD test or a fracture prior to inclusion. "High-risk" subjects (cases) were defined as BMD t-score < =-2.5 SD, any drop in BMD from baseline or a fracture 24-36 months following inclusion.

RESULTS

A total of 2406 subjects (66.3 % women; 33.7 % men) fulfilled the inclusion criteria, and of these, 352 (14.6 %) were identified as high risk subjects. A multiple logistical regression analysis showed that high risk subjects were more likely to have lower plasma calcium and/or vitamin D levels (OR: 2.9) and were more frequently diagnosed with hyperparathyroidism (OR: 2.6).

CONCLUSION

Based on Danish national health registry data, 14.6 % patients were identified as patients remaining at high risk for fracture despite being compliant to bisphosphonate treatment. Lower plasma calcium and/or vitamin D level is the greatest predictor of high risk for fracture despite persistent bisphosphonate treatment. Secondary causes of osteoporosis should be considered and alternative treatments may be advised for subjects whom remain at high risk.

One-year efficacy and safety of saxagliptin add-on in patients receiving dapagliflozin and metformin

AIMS

Greater reductions in glycated haemoglobin (HbA1c) with saxagliptin, a dipeptidyl peptidase-4 inhibitor, versus placebo add-on in patients with type 2 diabetes who had inadequate glycaemic control with dapagliflozin 10 mg/d plus metformin were demonstrated after 24 weeks of treatment. Results over 52 weeks of treatment were assessed in this analysis.

MATERIALS AND METHODS

Patients (mean baseline HbA1c 7.9%) receiving open-label dapagliflozin 10 mg/d plus metformin were randomized to double-blind saxagliptin 5 mg/d or placebo add-on.

RESULTS

The adjusted mean change from baseline to week 52 in HbA1c was greater with saxagliptin than with placebo add-on -0.38% vs 0.05%; difference -0.42% (95% confidence interval -0.64, -0.20)]. More patients achieved the HbA1c target of <7% with saxagliptin than with placebo add-on (29% vs 13%), and fewer patients were rescued or discontinued the study for lack of glycaemic control with saxagliptin than with placebo add-on (19% vs 28%). Reductions from baseline in body weight (≤1.5 kg) occurred in both groups. Similar proportions of patients reported ≥1 adverse event with saxagliptin (58.2%) and placebo add-on (58.0%); no new safety signals were detected. Hypoglycaemia was infrequent in both treatment groups (≤2.5%), with no major episodes. The rate of urinary tract infections was similar in the saxagliptin and placebo add-on groups (7.8% vs 7.4%). The incidence of genital infections was 3.3% with saxagliptin versus 6.2% with placebo add-on.

CONCLUSIONS

Triple therapy with saxagliptin add-on to dapagliflozin plus metformin for 52 weeks resulted in sustained improvements in glycaemic control without an increase in body weight or increased risk of hypoglycaemia.

Efficacy and safety of triple therapy with dapagliflozin add-on to saxagliptin plus metformin over 52 weeks in patients with type 2 diabetes

We previously reported that dapagliflozin versus placebo as add-on to saxagliptin plus metformin resulted in greater reductions in glycated haemoglobin (A1C), fasting plasma glucose (FPG) and body weight (BW) after 24 weeks of treatment in patients with type 2 diabetes (T2D). Here we report results after 52 weeks of treatment. Patients stabilized on open-label metformin and saxagliptin 5 mg/day for 8-16 weeks were randomized to placebo or dapagliflozin 10 mg/day plus open-label saxagliptin plus metformin for 52 weeks. Changes from baseline to week 52 were greater with dapagliflozin versus placebo in A1C (-0.74% vs. 0.07%), FPG (-27 vs. 10 mg/dL) and BW (-2.1 vs. -0.4 kg). More patients achieved A1C <7% with dapagliflozin (29.4%) versus placebo (12.6%). Adverse events were similar with dapagliflozin (66%) and placebo (71%), and hypoglycaemia was rare (≤2%). Genital infections occurred more often with dapagliflozin (6%) than with placebo (1%); frequency of urinary tract infections was similar between the two groups (9% vs. 10%). Triple therapy with dapagliflozin add-on to saxagliptin plus metformin is a durable, effective and well-tolerated intervention for the treatment of T2D.

The epidemiology of fractures in Denmark in 2011

In the present study, we used national health care databases to estimate fracture incidence rates (IRs) and compared these IRs based on imputed data. We showed that imputation could lead to both over- and underestimation of IRs, and future research should therefore focus on how to improve those imputations.

INTRODUCTION

Osteoporosis is a major public health burden through associated (osteoporotic) fractures. In Denmark, the incidence rates (IRs) of hip fracture are widely available. However, there is limited data about other fracture sites. A recent report could only provide imputed IRs, although nationwide data is readily available in electronic healthcare databases. Therefore, our aim was to estimate fracture site-specific IRs for Denmark in 2011 and to compare those to the previously reported imputed data.

METHODS

Data from the Danish National Hospital Discharge Register was used to estimate age- and gender-specific IRs for any fracture as well as for different fracture sites in the Danish population aged 20 years and older in 2011. Hip fracture IRs were stratified to sub-sites, and IRs were determined for all hip fractures which were confirmed by surgery.

RESULTS

The total number of incident fractures in 2011 was 80,760 (IR 191, 95 % confidence interval (CI) 190-192 (per 10,000 person-years)), of which 35,398 (43.8 %, IR 171, 95 % CI 169-173) occurred in men and 45,362 (56.2 %, IR 211, 95 % CI 209-213) in women. The majority of the fractures occurred in the population aged 50 years and older (n = 50,470, IR 249, 95 % CI 247-251). The numbers of any hip fracture were lower than the previously imputed estimates, whereas the number of forearm fractures was higher.

CONCLUSION

We showed age- and gender-specific fracture rates for any fracture as well as for different fracture sites. The IRs of most fracture sites increased with age. Estimating the number of fractures for Denmark based on imputation of data from other countries led to both over- and underestimation. Future research should therefore focus on how to improve those imputations as not all countries have nationwide registry data.

Early observations on the health of Syrian refugees in Canada

Between November 4, 2015 and February 29, 2016, a total of 26,166 refugees came to Canada. Of those, only two (0.01%) were found to have signs of a notifiable disease in the Immigration Medical Examination and these individuals were referred to public health for follow-up. Most refugees - 24,640 (94.17%) - arrived by government-coordinated charter flights and underwent enhanced screening. Upon arrival in Canada, 274 refugees (1.11%) were assessed by Quarantine Officers for signs of a potential communicable disease (such as fever) and 10 (0.04%) were referred to hospital. Paramedics onsite at the airport assessed 1,212 refugees (4.92%). Fifty-four (0.22%) were transferred to hospital and many of these were known to require urgent medical care. Provincial and local public health authorities and community networks have been instrumental in providing immediate and longer-term health care to arriving refugees. The two most immediate care needs were catch-up immunizations and dental care. Arriving in Canada at the height of the influenza season, a number of refugees experienced time-limited upper respiratory infections. When referring refugees to Canadian authorities, the United Nations High Commissioner for Refugees (UNHCR) advised that the Syrian refugee population may be expected to have high medical needs. These were not necessarily identified beforehand and may include diabetes, developmental disabilities, conflict-related injuries or mental health issues. These health care needs of Syrians will be identified and addressed as they integrate into the local health care systems. The arrival of Syrian refugees in Canada has not resulted in any urgent public health concerns or need for public health intervention. Canada's experience to date indicates that the arrival of Syrian refugees in this country can be managed in a way that will integrate them into the health care system without increased risk to public health.

An epidemiologic risk prediction model for ovarian cancer in Europe: the EPIC study

BACKGROUND

Ovarian cancer has a high case-fatality ratio, largely due to late diagnosis. Epidemiologic risk prediction models could help identify women at increased risk who may benefit from targeted prevention measures, such as screening or chemopreventive agents.

METHODS

We built an ovarian cancer risk prediction model with epidemiologic risk factors from 202,206 women in the European Prospective Investigation into Cancer and Nutrition study.

RESULTS

Older age at menopause, longer duration of hormone replacement therapy, and higher body mass index were included as increasing ovarian cancer risk, whereas unilateral ovariectomy, longer duration of oral contraceptive use, and higher number of full-term pregnancies were decreasing risk. The discriminatory power (overall concordance index) of this model, as examined with five-fold cross-validation, was 0.64 (95% confidence interval (CI): 0.57, 0.70). The ratio of the expected to observed number of ovarian cancer cases occurring in the first 5 years of follow-up was 0.90 (293 out of 324, 95% CI: 0.81-1.01), in general there was no evidence for miscalibration.

CONCLUSION

Our ovarian cancer risk model containing only epidemiological data showed modest discriminatory power for a Western European population. Future studies should consider adding informative biomarkers to possibly improve the predictive ability of the model.

Subsequent fracture rates in a nationwide population-based cohort study with a 10-year perspective

SUMMARY

Fractures after the age of 50 are frequently observed in Denmark, and many of these may be osteoporotic. This study examined the incidence of all and subsequent fractures in a 10-year period from 2001 to 2011. The incidence of subsequent fractures was high, especially following hip fracture.

INTRODUCTION

The purpose of this study is to examine patterns of subsequent fractures and mortality rates over a 10-year period in patients already suffering from fracture.

METHODS

The study was designed as a nationwide, register-based follow-up study. Patients were included if diagnosed with an index fracture (ICD-10 codes: S22.x, S32.x, S42.x, S52.x, S62.x, S72.x, S82.x, S92.x, T02.x, T08.x, T10.x and T12.x) between January 1st, 2001 and December 31st, 2001 and if older than 50 years at time of fracture. The patients were investigated for future subsequent fractures from January 1st, 2002 to December 31st, 2011.

RESULTS

In this study, we demonstrated that patients with fractures (especially hip fractures) have a high risk of subsequent fractures, especially hip fracture. Other fractures, which are not commonly considered as osteoporotic fractures, such as lower leg, were frequently observed in the 10 years following index fracture. The cumulative incidence proportion (CIP) of subsequent fractures during the 10-year follow-up period was high for all recurrent fractures (9-46 %). Subsequent hip fracture, regardless of index fracture, had the highest CIP across the study period, ranging from 9 to 40 %. Appendicular fractures were often followed by a recurrent fracture, or subsequent fractures at a more proximal location in the same limb, i.e. forearm fractures were followed by humerus fractures. These results have not been previously demonstrated to this extent, and according to our knowledge, no previous studies have estimated cumulative 10-year subsequent fracture incidences for any non-hip fractures.

CONCLUSION

Patients suffering a fracture (and especially a hip fracture) have a high incidence of subsequent fracture. Fractures after the age of 50 may be considered an early warning of increased risk for future fractures in many patients.

Plasma methionine, choline, betaine, and dimethylglycine in relation to colorectal cancer risk in the European Prospective Investigation into Cancer and Nutrition (EPIC)

BACKGROUND

Disturbances in one carbon metabolism may contribute to carcinogenesis by affecting methylation and synthesis of DNA. Choline and its oxidation product betaine are involved in this metabolism and can serve as alternative methyl group donors when folate status is low.

PATIENTS AND METHODS

We conducted a case-control study nested within the European Prospective Investigation into Cancer and Nutrition (EPIC), to investigate plasma concentrations of the methyl donors methionine, choline, betaine (trimethylglycine), and dimethylglycine (DMG) in relation to colorectal cancer (CRC) risk. Our study included 1367 incident CRC cases (965 colon and 402 rectum) and 2323 controls matched by gender, age group, and study center. Multivariate-adjusted odds ratios (ORs) and 95% confidence intervals (95% CIs) for CRC risk were estimated by conditional logistic regression, comparing the fifth to the first quintile of plasma concentrations.

RESULTS

Overall, methionine (OR: 0.79, 95% CI: 0.63-0.99, P-trend = 0.05), choline (OR: 0.77, 95% CI: 0.60-0.99, P-trend = 0.07), and betaine (OR: 0.85, 95% CI: 0.66-1.09, P-trend = 0.06) concentrations were inversely associated with CRC risk of borderline significance. In participants with folate concentration below the median of 11.3 nmol/l, high betaine concentration was associated with reduced CRC risk (OR: 0.71, 95% CI: 0.50-1.00, P-trend = 0.02), which was not observed for those having a higher folate status. Among women, but not men, high choline concentration was associated with decreased CRC risk (OR: 0.62, 95% CI: 0.43-0.88, P-trend = 0.01). Plasma DMG was not associated with CRC risk.

CONCLUSIONS

Individuals with high plasma concentrations of methionine, choline, and betaine may be at reduced risk of CRC.

A homozygous KCNJ10 mutation in Jack Russell Terriers and related breeds with spinocerebellar ataxia with myokymia, seizures, or both

Background

Juvenile‐onset spinocerebellar ataxia has been recognized in Jack Russell Terriers and related Russell group terriers (RGTs) for over 40 years. Ataxia occurs with varying combinations of myokymia, seizures, and other signs of neurologic disease. More than 1 form of the disease has been suspected.

Hypothesis/Objectives

The objective was to identify the mutation causing the spinocerebellar ataxia associated with myokymia, seizures, or both and distinguish the phenotype from other ataxias in the RGTs.

Animals

DNA samples from 16 RGTs with spinocerebellar ataxia beginning from 2 to 12 months of age, 640 control RGTs, and 383 dogs from 144 other breeds along with the medical records of affected dogs were studied.

Methods

This case‐control study compared the frequencies of a KCNJ10 allele in RGTs with spinocerebellar ataxia versus control RGTs. This allele was identified in a whole‐genome sequence of a single RGT with spinocerebellar ataxia and myokymia by comparison to whole‐genome sequences from 81 other canids that were normal or had other diseases.

Results

A missense mutation in the gene coding for the inwardly rectifying potassium channel Kir4.1 (KCNJ10:c.627C>G) was significantly (P < .001) associated with the disease. Dogs homozygous for the mutant allele all had spinocerebellar ataxia with varying combinations of myokymia and seizures.

Conclusions and Clinical Importance

Identification of the KCNJ10 mutation in dogs with spinocerebellar ataxia with myokymia, seizures, or both clarifies the multiple forms of ataxia seen in these breeds and provides a DNA test to identify carriers.