Publications by authors named "Pierre Chatelain"

60 Publications

Knowledge representation and learning of operator clinical workflow from full-length routine fetal ultrasound scan videos.

Med Image Anal 2021 Apr 23;69:101973. Epub 2021 Jan 23.

Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, Oxford, United Kingdom.

Ultrasound is a widely used imaging modality, yet it is well-known that scanning can be highly operator-dependent and difficult to perform, which limits its wider use in clinical practice. The literature on understanding what makes clinical sonography hard to learn and how sonography varies in the field is sparse, restricted to small-scale studies on the effectiveness of ultrasound training schemes, the role of ultrasound simulation in training, and the effect of introducing scanning guidelines and standards on diagnostic image quality. The Big Data era, and the recent and rapid emergence of machine learning as a more mainstream large-scale data analysis technique, presents a fresh opportunity to study sonography in the field at scale for the first time. Large-scale analysis of video recordings of full-length routine fetal ultrasound scans offers the potential to characterise differences between the scanning proficiency of experts and trainees that would be tedious and time-consuming to do manually due to the vast amounts of data. Such research would be informative to better understand operator clinical workflow when conducting ultrasound scans to support skills training, optimise scan times, and inform building better user-machine interfaces. This paper is to our knowledge the first to address sonography data science, which we consider in the context of second-trimester fetal sonography screening. Specifically, we present a fully-automatic framework to analyse operator clinical workflow solely from full-length routine second-trimester fetal ultrasound scan videos. An ultrasound video dataset containing more than 200 hours of scan recordings was generated for this study. We developed an original deep learning method to temporally segment the ultrasound video into semantically meaningful segments (the video description). The resulting semantic annotation was then used to depict operator clinical workflow (the knowledge representation). Machine learning was applied to the knowledge representation to characterise operator skills and assess operator variability. For video description, our best-performing deep spatio-temporal network shows favourable results in cross-validation (accuracy: 91.7%), statistical analysis (correlation: 0.98, p < 0.05) and retrospective manual validation (accuracy: 76.4%). For knowledge representation of operator clinical workflow, a three-level abstraction scheme consisting of a Subject-specific Timeline Model (STM), Summary of Timeline Features (STF), and an Operator Graph Model (OGM), was introduced that led to a significant decrease in dimensionality and computational complexity compared to raw video data. The workflow representations were learnt to discriminate between operator skills, where a proposed convolutional neural network-based model showed most promising performance (cross-validation accuracy: 98.5%, accuracy on unseen operators: 76.9%). These were further used to derive operator-specific scanning signatures and operator variability in terms of type, order and time distribution of constituent tasks.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.media.2021.101973DOI Listing
April 2021

Spatio-temporal visual attention modelling of standard biometry plane-finding navigation.

Med Image Anal 2020 10 20;65:101762. Epub 2020 Jun 20.

Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, Oxford, OX3 7DQ, UK.

We present a novel multi-task neural network called Temporal SonoEyeNet (TSEN) with a primary task to describe the visual navigation process of sonographers by learning to generate visual attention maps of ultrasound images around standard biometry planes of the fetal abdomen, head (trans-ventricular plane) and femur. TSEN has three components: a feature extractor, a temporal attention module (TAM), and an auxiliary video classification module (VCM). A soft dynamic time warping (sDTW) loss function is used to improve visual attention modelling. Variants of the model are trained on a dataset of 280 video clips, each containing one of the three biometry planes and lasting 3-7 seconds, with corresponding real-time recorded gaze tracking data of an experienced sonographer. We report the performances of the different variants of TSEN for visual attention prediction at standard biometry plane detection. The best model performance is achieved using bi-directional convolutional long-short term memory (biCLSTM) in both TAM and VCM, and it outperforms a previous spatial model on all static and dynamic saliency metrics. As an auxiliary task to validate the clinical relevance of the visual attention modelling, the predicted visual attention maps were used to guide standard biometry plane detection in consecutive US video frames. All spatio-temporal TSEN models achieve higher scores compared to a spatial-only baseline; the best performing TSEN model achieves F1 scores on these standard biometry planes of 83.7%, 89.9% and 81.1%, respectively.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.media.2020.101762DOI Listing
October 2020

Preoperative Topical Estrogen Treatment vs Placebo in 244 Children With Midshaft and Posterior Hypospadias.

J Clin Endocrinol Metab 2020 07;105(7)

Centre de Référence Maladies Rares Développement Génital: du Fœtus à l'Adulte, Hospices Civils de Lyon, Bron, France.

Purpose: Urethral fistula and dehiscence are common after hypospadias surgery. Preoperative androgens have been considered to reduce these complications although this consideration is not evidence-based. Dermatologists have reported the benefits of topical estrogens on skin healing. We investigated whether the preoperative use of topical promestriene could reduce healing complications in hypospadias surgery. Our primary objective was to demonstrate a reduction of healing complications with promestriene vs placebo. Impact on reoperations and other complications, clinical tolerance, bone growth, and biological systemic effects of the treatment were also considered.

Methods: We conducted a prospective, randomized, placebo-controlled, double-blind, parallel group trial between 2011 and 2015 in 4 French centers. One-stage transverse preputial island flap urethroplasty (onlay urethroplasty) was selected for severe hypospadias. Promestriene or placebo was applied on the penis for 2 months prior to surgery. The primary outcome was the presence of postoperative urethral fistula or dehiscence in the first year postsurgery. For safety reasons, hormonal and anatomical screenings were performed.

Results: Out of 241 patients who received surgery, 122 patients were randomized to receive placebo, and 119 patients received promestriene. The primary outcome was unavailable for 11 patients. Healing complications were assessed at 16.4% (19/116) in the placebo vs 14.9% (17/114) in the promestriene arm, and the odds ratio adjusted on center was 0.93 (95% confidence interval 0.45-1.94), P = 0.86.

Conclusions And Relevance: Although we observed an overall lower risk of complications compared to previous publications, postsurgery complications were not different between promestriene and placebo, because of a lack of power of the study or the inefficacy of promestriene.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1210/clinem/dgaa231DOI Listing
July 2020

Safety Indices of Ultrasound: Adherence to Recommendations and Awareness During Routine Obstetric Ultrasound Scanning.

Ultraschall Med 2020 Apr 27;41(2):138-145. Epub 2020 Feb 27.

Nuffield Department of Women's & Reproductive Health, Oxford-University, Oxford, United Kingdom of Great Britain and Northern Ireland.

Purpose:  To analyze bioeffect safety indices and assess how often operators look at these indices during routine obstetric ultrasound.

Materials And Methods:  Automated analysis of prospectively collected data including video recordings of full-length ultrasound scans coupled with operator eye tracking was performed. Using optical recognition, we extracted the Mechanical Index (MI), Thermal Index in soft tissue (TIs), and Thermal Index in bone (TIb) values and ultrasound mode. This allowed us to report the bioeffect safety indices during routine obstetric scans and assess adherence to professional organization recommendations. Eye-tracking analysis allowed us to assess how often operators look at the displayed bioeffect safety indices.

Results:  A total of 637 ultrasound scans performed by 17 operators were included, of which 178, 216, and 243 scans were first, second, and third-trimester scans, respectively. During live scanning, the mean and range were 0.14 (0.1 to 3.0) for TIb, 0.2 (0.1 to 1.2) for TIs, and 0.9 (0.1 to 1.3) for MI. The mean and standard deviation of TIb were 0.15 ± 0.03, 0.23 ± 0.09, 0.32 ± 0.24 in the first, second, and third trimester, respectively. For B-mode, the highest TIb was 0.8 in all trimesters. The highest TIb was recorded for pulsed-wave Doppler mode in all trimesters. The recommended exposure times were maintained in all scans. Analysis of eye tracking suggested that operators looked at bioeffect safety indices in only 27 (4.2 %) of the scans.

Conclusion:  In this study, recommended bioeffect indices were adhered to in all routine scans. However, eye tracking showed that operators rarely assessed safety indices during scanning.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1055/a-1074-0722DOI Listing
April 2020

Ultrasound Image Representation Learning by Modeling Sonographer Visual Attention.

Inf Process Med Imaging 2019 Jun 22;26:592-604. Epub 2019 May 22.

Department of Engineering Science, University of Oxford, UK.

Image representations are commonly learned from class labels, which are a simplistic approximation of human image understanding. In this paper we demonstrate that transferable representations of images can be learned without manual annotations by modeling human visual attention. The basis of our analyses is a unique gaze tracking dataset of sonographers performing routine clinical fetal anomaly screenings. Models of sonographer visual attention are learned by training a convolutional neural network (CNN) to predict gaze on ultrasound video frames through visual saliency prediction or gaze-point regression. We evaluate the transferability of the learned representations to the task of ultrasound standard plane detection in two contexts. Firstly, we perform transfer learning by fine-tuning the CNN with a limited number of labeled standard plane images. We find that fine-tuning the saliency predictor is superior to training from random initialization, with an average F1-score improvement of 9.6% overall and 15.3% for the cardiac planes. Secondly, we train a simple softmax regression on the feature activations of each CNN layer in order to evaluate the representations independently of transfer learning hyper-parameters. We find that the attention models derive strong representations, approaching the precision of a fully-supervised baseline model for all but the last layer.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/978-3-030-20351-1_46DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6986905PMC
June 2019

Captioning Ultrasound Images Automatically.

Med Image Comput Comput Assist Interv 2019 Oct 10;22:338-346. Epub 2019 Oct 10.

University of Oxford, Oxford, UK.

We describe an automatic natural language processing (NLP)-based image captioning method to describe fetal ultrasound video content by modelling the vocabulary commonly used by sonographers and sonologists. The generated captions are similar to the words spoken by a sonographer when describing the scan experience in terms of visual content and performed scanning actions. Using full-length second-trimester fetal ultrasound videos and text derived from accompanying expert voice-over audio recordings, we train deep learning models consisting of convolutional neural networks and recurrent neural networks in merged configurations to generate captions for ultrasound video frames. We evaluate different model architectures using established general metrics (, ) and application-specific metrics. Results show that the proposed models can learn joint representations of image and text to generate relevant and descriptive captions for anatomies, such as the spine, the abdomen, the heart, and the head, in clinical fetal ultrasound scans.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/978-3-030-32251-9_37DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6978141PMC
October 2019

Efficient Ultrasound Image Analysis Models with Sonographer Gaze Assisted Distillation.

Med Image Comput Comput Assist Interv 2019 10;22(Pt 4):394-402. Epub 2019 Oct 10.

University of Oxford, Oxford OX3 7DQ, United Kingdom.

Recent automated medical image analysis methods have attained state-of-the-art performance but have relied on memory and compute-intensive deep learning models. Reducing model size without significant loss in performance metrics is crucial for time and memory-efficient automated image-based decision-making. Traditional deep learning based image analysis only uses expert knowledge in the form of manual annotations. Recently, there has been interest in introducing other forms of expert knowledge into deep learning architecture design. This is the approach considered in the paper where we propose to combine ultrasound video with point-of-gaze tracked for expert sonographers as they scan to train memory-efficient ultrasound image analysis models. Specifically we develop teacher-student knowledge transfer models for the exemplar task of frame classification for the fetal abdomen, head, and femur. The best performing memory-efficient models attain performance within 5% of conventional models that are 1000× larger in size.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/978-3-030-32251-9_43DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6962054PMC
October 2019

Multi-task SonoEyeNet: Detection of Fetal Standardized Planes Assisted by Generated Sonographer Attention Maps.

Med Image Comput Comput Assist Interv 2018 Sep 26;11070:871-879. Epub 2018 Sep 26.

Institute of Biomedical Engineering, University of Oxford, Old Road Campus Research Building, Oxford OX3 7DQ, UK.

We present a novel multi-task convolutional neural network called Multi-task SonoEyeNet () that learns to generate clinically relevant visual attention maps using sonographer gaze tracking data on input ultrasound (US) video frames so as to assist standardized abdominal circumference (AC) plane detection. Our architecture consists of a generator and a discriminator, which are trained in an adversarial scheme. The generator learns sonographer attention on a given US video frame to predict the frame label (standardized AC plane / background). The discriminator further fine-tunes the predicted attention map by encouraging it to mimick the ground-truth sonographer attention map. The novel model expands the potential clinical usefulness of a previous model by eliminating the requirement of input gaze tracking data during inference without compromising its plane detection performance (Precision: 96.8, Recall: 96.2, F-1 score: 96.5).
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/978-3-030-00928-1_98DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6459365PMC
September 2018

Evaluation of Gaze Tracking Calibration for Longitudinal Biomedical Imaging Studies.

IEEE Trans Cybern 2020 Jan 5;50(1):153-163. Epub 2018 Sep 5.

Gaze tracking is a promising technology for studying the visual perception of clinicians during image-based medical exams. It could be used in longitudinal studies to analyze their perceptive process, explore human-machine interactions, and develop innovative computer-aided imaging systems. However, using a remote eye tracker in an unconstrained environment and over time periods of weeks requires a certain guarantee of performance to ensure that collected gaze data are fit for purpose. We report the results of evaluating eye tracking calibration for longitudinal studies. First, we tested the performance of an eye tracker on a cohort of 13 users over a period of one month. For each participant, the eye tracker was calibrated during the first session. The participants were asked to sit in front of a monitor equipped with the eye tracker, but their position was not constrained. Second, we tested the performance of the eye tracker on sonographers positioned in front of a cart-based ultrasound scanner. Experimental results show a decrease of accuracy between calibration and later testing of 0.30° and a further degradation over time at a rate of 0.13°. month. The overall median accuracy was 1.00° (50.9 pixels) and the overall median precision was 0.16° (8.3 pixels). The results from the ultrasonography setting show a decrease of accuracy of 0.16° between calibration and later testing. This slow degradation of gaze tracking accuracy could impact the data quality in long-term studies. Therefore, the results we present here can help in planning such long-term gaze tracking studies.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1109/TCYB.2018.2866274DOI Listing
January 2020

Transcriptomics and machine learning predict diagnosis and severity of growth hormone deficiency.

JCI Insight 2018 04 5;3(7). Epub 2018 Apr 5.

Division of Developmental Biology and Medicine, Faculty of Biology, Medicine and Health, University of Manchester and Manchester Academic Health Science Centre, Manchester, United Kingdom.

Background: The effect of gene expression data on diagnosis remains limited. Here, we show how diagnosis and classification of growth hormone deficiency (GHD) can be achieved from a single blood sample using a combination of transcriptomics and random forest analysis.

Methods: Prepubertal treatment-naive children with GHD (n = 98) were enrolled from the PREDICT study, and controls (n = 26) were acquired from online data sets. Whole blood gene expression was correlated with peak growth hormone (GH) using rank regression and a random forest algorithm tested for prediction of the presence of GHD and in classification of GHD as severe (peak GH <4 μg/l) and nonsevere (peak ≥4 μg/l). Performance was assessed using area under the receiver operating characteristic curve (AUC-ROC).

Results: Rank regression identified 347 probe sets in which gene expression correlated with peak GH concentrations (r = ± 0.28, P < 0.01). These 347 probe sets yielded an AUC-ROC of 0.95 for prediction of GHD status versus controls and an AUC-ROC of 0.93 for prediction of GHD severity.

Conclusion: This study demonstrates highly accurate diagnosis and disease classification for GHD using a combination of transcriptomics and random forest analysis.

Trial Registration: NCT00256126 and NCT00699855.

Funding: Merck and the National Institute for Health Research (CL-2012-06-005).
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1172/jci.insight.93247DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC5928867PMC
April 2018

Assessment of informal caregiver's needs by self-administered instruments: a literature review.

Eur J Public Health 2017 10;27(5):796-801

Fondation France Répit, Lyon, France.

Background: Clinicians, researchers and politicians are seeking to better assess caregiver's needs. Challenges exist in broadly implementing this so as to provide appropriate support. The aim of this review was to compile self-administered instruments for assessment of caregiver's needs that are deemed to be scientifically robust.

Methods: The Medline database was searched for publications reporting self-administered instruments assessing caregiver's needs with acceptable psychometric properties. These instruments were analyzed in terms of the development context, target population, concept, purpose, structure, content and psychometric properties. The dimensions of the needs were listed and categorized.

Results: A total of nine self-administered instruments were analyzed. They averaged 32 items, they were specifically developed for a targeted subpopulation of caregivers and dedicated to epidemiological research. Response devices were based on Likert scales. The main dimensions of the needs identified were 'Health and Care', 'Psychological - Emotional Support', 'Information-Knowledge', 'Social Life-Work-Finance'. None was specifically geared toward caregivers for the elderly, children or teenagers. In the absence of transcultural validation, no instrument was directly usable in Europe.

Conclusions: Assessing caregivers' needs is a key part in providing caregivers with appropriate support. The development of self-administered instruments constitutes a complex field that is still underexplored at the international level; strict specifications with psychometric validation are essential. To be efficient, the instrument should be integrated in a larger process including: upstream, recognition, identification and assessment of the overall situation of the caregiver; and downstream, guidance, establishment and follow-up of a suitable action plan.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1093/eurpub/ckx103DOI Listing
October 2017

The human bitter taste receptor T2R38 is broadly tuned for bacterial compounds.

PLoS One 2017 13;12(9):e0181302. Epub 2017 Sep 13.

ChemCom S.A., Brussels, Belgium.

T2R38 has been shown to be a specific bacterial detector implicated in innate immune defense mechanism of human upper airway. Several clinical studies have demonstrated that this receptor is associated with the development of chronic rhinosinusitis (CRS). T2R38 was previously reported to bind to homoserine lactones (HSL), quorum sensing molecules specific of Pseudomonas Aeruginosa and other gram negative species. Nevertheless, these bacteria are not the major pathogens found in CRS. Here we report on the identification of bacterial metabolites acting as new agonists of T2R38 based on a single cell calcium imaging study. Two quorum sensing molecules (Agr D1 thiolactone from Staphylococcus Aureus and CSP-1 from Streptococcus Pneumoniae) and a list of 32 bacterial metabolites from pathogens frequently implicated in CRS were tested. First, we observed that HSL failed to activate T2R38 in our experimental system, but that the dimethylsulfoxide (DMSO), used as a solvent for these lactones may, by itself, account for the agonistic effect previously described. Secondly, we showed that both Agr D1 thiolactone and CSP-1 are inactive but that at least 7 bacterial metabolites (acetone, 2-butanone, 2-pentanone, 2-methylpropanal, dimethyl disulfide, methylmercaptan, γ-butyrolactone) are able to specifically trigger this receptor. T2R38 is thus much more broadly tuned for bacterial compounds than previously thought.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0181302PLOS
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC5597121PMC
October 2017

A Randomized Phase 2 Study of Long-Acting TransCon GH vs Daily GH in Childhood GH Deficiency.

J Clin Endocrinol Metab 2017 05;102(5):1673-1682

Ascendis Pharma, 2900 Hellerup, Denmark.

Context: TransCon Growth Hormone (GH) (Ascendis Pharma) is a long-acting recombinant sustained-release human GH prodrug in development for children with GH deficiency (GHD).

Objective: To compare the pharmacokinetics, pharmacodynamics, safety, and efficacy of weekly TransCon GH to that of daily GH in prepubertal children with GHD.

Design: Randomized, open-label, active-controlled study of three doses of weekly TransCon GH versus daily Genotropin (Pfizer).

Setting: Thirty-eight centers in 14 European countries and Egypt.

Patients: Prepubertal male and female treatment-naïve children with GHD (n = 53).

Interventions: Subjects received one of three TransCon GH doses (0.14, 0.21, or 0.30 mg GH/kg/wk) or Genotropin 0.03 mg GH/kg/d for 26 weeks.

Main Outcome Measures: GH and insulinlike growth factor-1 (IGF-1) levels, growth, adverse events, and immunogenicity.

Results: Both GH maximum concentration and area under the curve were similar following TransCon GH or Genotropin administration at comparable doses. A dose response was observed, with IGF-1 standard deviation scores increasing into the normal range for all three TransCon GH doses. Annualized mean height velocity for the three TransCon GH doses ranged from 11.9 cm to 13.9 cm, which was not statistically different from 11.6 cm for Genotropin. Adverse events were mild to moderate, and most were unrelated to the study drug. Injection site tolerance was good. One TransCon GH subject developed a low-titer, nonneutralizing antibody response to GH.

Conclusions: The results suggest that long-acting TransCon GH is comparable to daily Genotropin for GH (pharmacokinetics) and IGF-1 (pharmacodynamics) levels, safety, and efficacy and support advancement into phase 3 development.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1210/jc.2016-3776DOI Listing
May 2017

Growth curves for congenital adrenal hyperplasia from a national retrospective cohort.

J Pediatr Endocrinol Metab 2016 Dec;29(12):1379-1388

Background: In congenital adrenal hyperplasia (CAH), adjusting hydrocortisone dose during childhood avoids reduced adult height. However, there are currently no CAH-specific charts to monitor growth during treatment. Our objective was to elaborate growth reference charts and bone maturation data for CAH patients.

Methods: We conducted a retrospective observational cohort study, in 34 French CAH centers. Patients were 496 children born 1970-1991 with genetically proven 21-hydroxylase deficiency. Their growth and bone maturation data were collected until age 18 together with adult height, puberty onset, parental height, and treatment. The mean (SD) heights were modeled from birth to adulthood. The median±1 SD and ±2 SDs model-generated curves were compared with the French references. A linear model for bone maturation and a logistic regression model for the probability of short adult height were built.

Results: Growth charts were built by sex for salt wasting (SW) and simple virilizing (SV) children treated before 1 year of age. In girls and boys, growth was close to that of the general French population up to puberty onset. There was almost no pubertal spurt and the mean adult height was shorter than that of the general population in girls (-1.2 SD, 156.7 cm) and boys (-1.0 SD, 168.8 cm). Advanced bone age at 8 years had a strong impact on the risk of short adult height (OR: 4.5 per year advance).

Conclusions: The 8-year bone age is a strong predictor of adult height. It will help monitoring the growth of CAH-affected children.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1515/jpem-2016-0156DOI Listing
December 2016

Assisting the examination of large histopathological slides with adaptive forests.

Med Image Anal 2017 01 5;35:655-668. Epub 2016 Oct 5.

Computer Aided Medical Procedures, Technische Universität München, Germany; Computer Aided Medical Procedures, Johns Hopkins University, USA.

The examination of biopsy samples plays a central role in the diagnosis and staging of numerous diseases, including most cancer types. However, because of the large size of the acquired images, the localization and quantification of diseased portions of a tissue is usually time-consuming, as pathologists must scroll through the whole slide to look for objects of interest which are often only scarcely distributed. In this work, we introduce an approach to facilitate the visual inspection of large digital histopathological slides. Our method builds on a random forest classifier trained to segment the structures sought by the pathologist. However, moving beyond the pixelwise segmentation task, our main contribution is an interactive exploration framework including: (i) a region scoring function which is used to rank and sequentially display regions of interest to the user, and (ii) a relevance feedback capability which leverages human annotations collected on each suggested region. Thereby, an online domain adaptation of the learned pixelwise segmentation model is performed, so that the region scores adapt on-the-fly to possible discrepancies between the original training data and the slide at hand. Three real-time update strategies are compared, including a novel approach based on online gradient descent which supports faster user interaction than an accurate delineation of objects. Our method is evaluated on the task of extramedullary hematopoiesis quantification within mouse liver slides. We assess quantitatively the retrieval abilities of our approach and the benefit of the interactive adaptation scheme. Moreover, we demonstrate the possibility of extrapolating, after a partial exploration of the slide, the surface covered by hematopoietic cells within the whole tissue.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.media.2016.09.009DOI Listing
January 2017

Validating genetic markers of response to recombinant human growth hormone in children with growth hormone deficiency and Turner syndrome: the PREDICT validation study.

Eur J Endocrinol 2016 Dec 20;175(6):633-643. Epub 2016 Sep 20.

Faculty of BiologyMedicine and Health, University of Manchester and Manchester Academic Health Science Centre, Royal Manchester Children's Hospital, Central Manchester University Hospitals NHS Foundation Trust, Manchester, UK

Objective: Single-nucleotide polymorphisms (SNPs) associated with the response to recombinant human growth hormone (r-hGH) have previously been identified in growth hormone deficiency (GHD) and Turner syndrome (TS) children in the PREDICT long-term follow-up (LTFU) study (Nbib699855). Here, we describe the PREDICT validation (VAL) study (Nbib1419249), which aimed to confirm these genetic associations.

Design And Methods: Children with GHD (n = 293) or TS (n = 132) were recruited retrospectively from 29 sites in nine countries. All children had completed 1 year of r-hGH therapy. 48 SNPs previously identified as associated with first year growth response to r-hGH were genotyped. Regression analysis was used to assess the association between genotype and growth response using clinical/auxological variables as covariates. Further analysis was undertaken using random forest classification.

Results: The children were younger, and the growth response was higher in VAL study. Direct genotype analysis did not replicate what was found in the LTFU study. However, using exploratory regression models with covariates, a consistent relationship with growth response in both VAL and LTFU was shown for four genes - SOS1 and INPPL1 in GHD and ESR1 and PTPN1 in TS. The random forest analysis demonstrated that only clinical covariates were important in the prediction of growth response in mild GHD (>4 to <10 μg/L on GH stimulation test), however, in severe GHD (≤4 μg/L) several SNPs contributed (in IGF2, GRB10, FOS, IGFBP3 and GHRHR).

Conclusions: The PREDICT validation study supports, in an independent cohort, the association of four of 48 genetic markers with growth response to r-hGH treatment in both pre-pubertal GHD and TS children after controlling for clinical/auxological covariates. However, the contribution of these SNPs in a prediction model of first-year response is not sufficient for routine clinical use.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1530/EJE-16-0357DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC5097129PMC
December 2016

Growth Hormone Research Society perspective on the development of long-acting growth hormone preparations.

Eur J Endocrinol 2016 Jun 23;174(6):C1-8. Epub 2016 Mar 23.

Swedish Neuroscience Institute Seattle, Washington, USA.

Objective: The Growth Hormone (GH) Research Society (GRS) convened a workshop to address important issues regarding trial design, efficacy, and safety of long-acting growth hormone preparations (LAGH).

Participants: A closed meeting of 55 international scientists with expertise in GH, including pediatric and adult endocrinologists, basic scientists, regulatory scientists, and participants from the pharmaceutical industry.

Evidence: Current literature was reviewed for gaps in knowledge. Expert opinion was used to suggest studies required to address potential safety and efficacy issues.

Consensus Process: Following plenary presentations summarizing the literature, breakout groups discussed questions framed by the planning committee. Attendees reconvened after each breakout session to share group reports. A writing team compiled the breakout session reports into a draft document that was discussed and revised in an open forum on the concluding day. This was edited further and then circulated to attendees from academic institutions for review after the meeting. Participants from pharmaceutical companies did not participate in the planning, writing, or in the discussions and text revision on the final day of the workshop. Scientists from industry and regulatory agencies reviewed the manuscript to identify any factual errors.

Conclusions: LAGH compounds may represent an advance over daily GH injections because of increased convenience and differing phamacodynamic properties, providing the potential for improved adherence and outcomes. Better methods to assess adherence must be developed and validated. Long-term surveillance registries that include assessment of efficacy, cost-benefit, disease burden, quality of life, and safety are essential for understanding the impact of sustained exposure to LAGH preparations.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1530/EJE-16-0111DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC5081743PMC
June 2016

GH deficiency status combined with GH receptor polymorphism affects response to GH in children.

Eur J Endocrinol 2015 Dec 4;173(6):777-89. Epub 2015 Sep 4.

Merck Serono SAGeneva, SwitzerlandDépartement de PédiatrieHôpital Mère-Enfant, Université Claude Bernard, Lyon, FranceFederal State Institution 'Endocrinology Scientific Center of Russian Medical Technology'Moscow, RussiaEndocrine ServiceHospital de Pediatría Garrahan, Ciudad Autónoma de Buenos Aires, Buenos Aires, ArgentinaIRCCS Istituto Giannina Gaslini di GenovaClinica Pediatrica, Università di Genova, Genova, ItalyPediatra d.U. Azienda Ospedaliera Universitaria IntegrataUniversità di Verona, Verona, ItalyManchester Academic Health Sciences CentreRoyal Manchester Children's Hospital, 5th Floor, Oxford Road, Manchester M13 9WL, UK

Meta-analysis has shown a modest improvement in first-year growth response to recombinant human GH (r-hGH) for carriers of the exon 3-deleted GH receptor (GHRd3) polymorphism but with significant interstudy variability. The associations between GHRd3 and growth response to r-hGH over 3 years in relation to severity of GH deficiency (GHD) were investigated in patients from 14 countries. Treatment-naïve pre-pubertal children with GHD were enrolled from the PREDICT studies (NCT00256126 and NCT00699855), categorized by peak GH level (peak GH) during provocation test: ≤4 μg/l (severe GHD; n=45) and >4 to <10 μg/l mild GHD; n=49) and genotyped for the GHRd3 polymorphism (full length (fl/fl, fl/d3, d3/d3). Gene expression (GE) profiles were characterized at baseline. Changes in growth (height (cm) and SDS) over 3 years were measured. There was a dichotomous influence of GHRd3 polymorphism on response to r-hGH, dependent on peak GH level. GH peak level (higher vs lower) and GHRd3 (fl/fl vs d3 carriers) combined status was associated with height change over 3 years (P<0.05). GHRd3 carriers with lower peak GH had lower growth than subjects with fl/fl (median difference after 3 years -3.3 cm; -0.3 SDS). Conversely, GHRd3 carriers with higher peak GH had better growth (+2.7 cm; +0.2 SDS). Similar patterns were observed for GH-dependent biomarkers. GE profiles were significantly different between the groups, indicating that the interaction between GH status and GHRd3 carriage can be identified at a transcriptomic level. This study demonstrates that responses to r-hGH depend on the interaction between GHD severity and GHRd3 carriage.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1530/EJE-15-0474DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4623334PMC
December 2015

Growth hormone treatment for childhood short stature and risk of stroke in early adulthood.

Neurology 2015 Mar;84(10):1062-3

Lyon, France.

View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1212/WNL.0000000000001385DOI Listing
March 2015

Deorphanization and characterization of human olfactory receptors in heterologous cells.

Chem Biodivers 2014 Nov;11(11):1764-81

ChemCom S.A., Route de Lennik, 802, B-1070 Brussels.

Olfaction plays an indispensable role in human and animals in self and environmental recognition, as well as intra- and interspecific communication. Following the discovery of a family of olfactory receptors (ORs) by Buck and Axel in 1991, it has been established that the sense of smell begins with the molecular recognition of a chemical odorant by one or more ORs expressed in the olfactory sensory neurons. Therefore, characterization of the molecular interactions between odorant molecules and ORs is a key step in the elucidation of the general properties of the olfactory system and in the development of applications, i.e., design of new odorants, search for blockers, etc. The process putted in place at ChemCom to improve the expression of ORs at the cytoplasmic membrane of the HEK293 cell and assays enabling large-scale deorphanization, and to characterize the interaction between chemical odorants and ORs is described. The family of human ORs includes ca. 400 putatively functional ORs which are GPCRs (G protein-coupled receptors); to date over 100 human ORs have been deorphanized.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1002/cbdv.201400083DOI Listing
November 2014

Leveraging random forests for interactive exploration of large histological images.

Med Image Comput Comput Assist Interv 2014 ;17(Pt 1):1-8

The large size of histological images combined with their very challenging appearance are two main difficulties which considerably complicate their analysis. In this paper, we introduce an interactive strategy leveraging the output of a supervised random forest classifier to guide a user through such large visual data. Starting from a forest-based pixelwise estimate, subregions of the images at hand are automatically ranked and sequentially displayed according to their expected interest. After each region suggestion, the user selects among several options a rough estimate of the true amount of foreground pixels in this region. From these one-click inputs, the region scoring function is updated in real time using an online gradient descent procedure, which corrects on-the-fly the shortcomings of the initial model and adapts future suggestions accordingly. Experimental validation is conducted for extramedullary hematopoesis localization and demonstrates the practical feasibility of the procedure as well as the benefit of the online adaptation strategy.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/978-3-319-10404-1_1DOI Listing
November 2014

Changes over time in sex assignment for disorders of sex development.

Pediatrics 2014 Sep 4;134(3):e710-5. Epub 2014 Aug 4.

Pediatric Endocrinology and Diabetes Unit, Dana Children's Hospital, Tel Aviv, Israel; Sourasky Medical Center, Tel Aviv, Israel; and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel.

Background And Objective: It is unclear whether the proportion of infants with a disorder of sex development who are raised as male or female has changed over time. The temporal trends in sex assignment of affected cases entered in the International Disorder of Sex Development (I-DSD) Registry were studied.

Methods: Cases of disorders of sex development reported as partial androgen insensitivity syndrome (PAIS; n = 118), disorder of gonadal development (DGD; n = 232), and disorder of androgen synthesis (DAS; n = 104) were divided into those who were born before 1990, 1990-1999, and after 1999. External appearance of the genitalia was described by the external masculinization score.

Results: The median (5th-95th percentile) external masculinization scores of those infants with PAIS, DGD, and DAS who were raised as boys were 6 (2-9), 6 (3-9), and 6 (1-12), respectively, and were significantly higher than in those raised as girls (2 [0-6], 2 [0-7], and 0 [0-5], respectively); this difference was maintained in the 3 temporal birth cohorts (P < .01). Of the 118 cases in the pre-1990 cohort, 41 (35%) were raised as boys; of the 148 cases in the 1990-1999 cohort, 60 (41%) were raised as boys; and of the 188 cases in the post-1999 cohort, 128 (68%) were raised as boys.

Conclusions: Although there is an association between the external appearance of the genitalia and the choice of sex assignment, there are clear temporal trends in this practice pointing toward an increased likelihood of affected infants being raised as boys. The impact of this change in practice on long-term health outcomes requires additional focus.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1542/peds.2014-1088DOI Listing
September 2014

Profiling of olfactory receptor gene expression in whole human olfactory mucosa.

PLoS One 2014 6;9(5):e96333. Epub 2014 May 6.

ChemCom S.A., Brussels, Belgium.

Olfactory perception is mediated by a large array of olfactory receptor genes. The human genome contains 851 olfactory receptor gene loci. More than 50% of the loci are annotated as nonfunctional due to frame-disrupting mutations. Furthermore haplotypic missense alleles can be nonfunctional resulting from substitution of key amino acids governing protein folding or interactions with signal transduction components. Beyond their role in odor recognition, functional olfactory receptors are also required for a proper targeting of olfactory neuron axons to their corresponding glomeruli in the olfactory bulb. Therefore, we anticipate that profiling of olfactory receptor gene expression in whole human olfactory mucosa and analysis in the human population of their expression should provide an opportunity to select the frequently expressed and potentially functional olfactory receptors in view of a systematic deorphanization. To address this issue, we designed a TaqMan Low Density Array (Applied Biosystems), containing probes for 356 predicted human olfactory receptor loci to investigate their expression in whole human olfactory mucosa tissues from 26 individuals (13 women, 13 men; aged from 39 to 81 years, with an average of 67±11 years for women and 63±12 years for men). Total RNA isolation, DNase treatment, RNA integrity evaluation and reverse transcription were performed for these 26 samples. Then 384 targeted genes (including endogenous control genes and reference genes specifically expressed in olfactory epithelium for normalization purpose) were analyzed using the same real-time reverse transcription PCR platform. On average, the expression of 273 human olfactory receptor genes was observed in the 26 selected whole human olfactory mucosa analyzed, of which 90 were expressed in all 26 individuals. Most of the olfactory receptors deorphanized to date on the basis of sensitivity to known odorant molecules, which are described in the literature, were found in the expressed olfactory receptors gene set.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0096333PLOS
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4011832PMC
January 2015

Late prenatal dexamethasone and phenotype variations in 46,XX CAH: concerns about current protocols and benefits for surgical procedures.

J Pediatr Urol 2014 Oct 15;10(5):941-7. Epub 2014 Mar 15.

Service d'Urologie Pédiatrique, Hôpital Mère-Enfant, Centre Hospitalo-Universitaire de Lyon, GHE, 59, boulevard Pinel, 69677 Bron Cedex, France. Electronic address:

Objective: To describe the action of prenatal dexamethasone (PreDex) on the anatomy of female congenital adrenal hyperplasia (CAH) genitalia when started at later stages of gestation.

Materials And Methods: Our group follows a large cohort of French CAH patients who underwent PreDex therapy, of whom 258 were recently reported. Four 46,XX patients with a delayed PreDex treatment presented with a virilized genitalia and required surgical reconstruction. This is a retrospective report on genital phenotyping at the time of surgery of these four patients who began PreDex therapy at 8, 12, 20, and 28 weeks of gestation.

Results: Although this series is limited in number, the anatomical description of the length of the genital tubercle, the height of the urethra-vaginal confluence, and the degree of fusion of the genital folds seems to be dependent upon the starting date of PreDex. Most PreDex treatments prescribed up to now have covered the full duration of gestation.

Conclusions: Our findings suggest that PreDex therapy could be limited to the period of the partitioning window. It is hoped that further prospective multicentric clinical studies will obtain ethical approval in order to elucidate the place and protocols of PreDex therapy in the management of CAH.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.jpurol.2014.02.003DOI Listing
October 2014

Learning from multiple experts with random forests: application to the segmentation of the midbrain in 3D ultrasound.

Med Image Comput Comput Assist Interv 2013 ;16(Pt 2):230-7

Computer Aided Medical Procedures, Technische Universität München, Germany.

In the field of computer aided medical image analysis, it is often difficult to obtain reliable ground truth for evaluating algorithms or supervising statistical learning procedures. In this paper we present a new method for training a classification forest from images labelled by variably performing experts, while simultaneously evaluating the performance of each expert. Our approach builds upon state-of-the-art randomized classification forest techniques for medical image segmentation and recent methods for the fusion of multiple expert decisions. By incorporating the performance evaluation within the training phase, we obtain a novel forest framework for learning from conflicting expert decisions, accounting for both inter- and intra-expert variability. We demonstrate on a synthetic example that our method allows to retrieve the correct segmentation among other incorrectly labelled images, and we present an application to the automatic segmentation of the midbrain in 3D transcranial ultrasound images.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/978-3-642-40763-5_29DOI Listing
April 2014

Pediatric perspective on pharmacogenomics.

Pharmacogenomics 2013 Nov;14(15):1889-905

Institute of Human Development, Medical & Human Sciences, University of Manchester & Royal Manchester Children's Hospital, Central Manchester University Hospitals NHS Foundation Trust, Manchester Academic Health Science Centre, 5th Floor Research, Oxford Road, Manchester, M13 9WL, UK.

The advances in high-throughput genomic technologies have improved the understanding of disease pathophysiology and have allowed a better characterization of drug response and toxicity based on individual genetic make up. Pharmacogenomics is being recognized as a valid approach used to identify patients who are more likely to respond to medication, or those in whom there is a high probability of developing severe adverse drug reactions. An increasing number of pharmacogenomic studies are being published, most include only adults. A few studies have shown the impact of pharmacogenomics in pediatrics, highlighting a key difference between children and adults, which is the contribution of developmental changes to therapeutic responses across different age groups. This review focuses on pharmacogenomic research in pediatrics, providing examples from common pediatric conditions and emphasizing their developmental context.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.2217/pgs.13.193DOI Listing
November 2013

Human growth is associated with distinct patterns of gene expression in evolutionarily conserved networks.

BMC Genomics 2013 Aug 13;14:547. Epub 2013 Aug 13.

Manchester Academic Health Sciences Centre, Faculty of Medical and Human Sciences, Royal Manchester Children's Hospital and the Institute of Human Development, University of Manchester, Manchester, United Kingdom.

Background: A co-ordinated tissue-independent gene expression profile associated with growth is present in rodent models and this is hypothesised to extend to all mammals. Growth in humans has similarities to other mammals but the return to active long bone growth in the pubertal growth spurt is a distinctly human growth event. The aim of this study was to describe gene expression and biological pathways associated with stages of growth in children and to assess tissue-independent expression patterns in relation to human growth.

Results: We conducted gene expression analysis on a library of datasets from normal children with age annotation, collated from the NCBI Gene Expression Omnibus (GEO) and EBI Arrayexpress databases. A primary data set was generated using cells of lymphoid origin from normal children; the expression of 688 genes (ANOVA false discovery rate modified p-value, q < 0.1) was associated with age, and subsets of these genes formed clusters that correlated with the phases of growth--infancy, childhood, puberty and final height. Network analysis on these clusters identified evolutionarily conserved growth pathways (NOTCH, VEGF, TGFB, WNT and glucocorticoid receptor--Hyper-geometric test, q < 0.05). The greatest degree of network 'connectivity' and hence functional significance was present in infancy (Wilcoxon test, p < 0.05), which then decreased through to adulthood. These observations were confirmed in a separate validation data set from lymphoid tissue. Similar biological pathways were observed to be associated with development-related gene expression in other tissues (conjunctival epithelia, temporal lobe brain tissue and bone marrow) suggesting the existence of a tissue-independent genetic program for human growth and maturation.

Conclusions: Similar evolutionarily conserved pathways have been associated with gene expression and child growth in multiple tissues. These expression profiles associate with the developmental phases of growth including the return to active long bone growth in puberty, a distinctly human event. These observations also have direct medical relevance to pathological changes that induce disease in children. Taking into account development-dependent gene expression profiles for normal children will be key to the appropriate selection of genes and pathways as potential biomarkers of disease or as drug targets.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1186/1471-2164-14-547DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3765282PMC
August 2013

Prenatal management of disorders of sex development.

J Pediatr Urol 2012 Dec 4;8(6):576-84. Epub 2012 Nov 4.

Clinical and Molecular Genetics Unit, UCL Institute of Child Health, London, UK.

Disorders of sex development (DSD) rarely present prenatally but, as they are very complex conditions, management should be directed by highly specialised medical teams to allow consideration of all aspects of diagnosis, treatment and ethical issues. In this brief review, we present an overview of the prenatal presentation and management of DSD, including the sonographic appearance of normal genitalia and methods of determining genetic sex, the prenatal management of pregnancies with the unexpected finding of genital ambiguity on prenatal ultrasound and a review of the prenatal management of pregnancies at high risk of DSD. As this is a rapidly developing field, management options will change over time, making the involvement of clinical geneticists, paediatric endocrinologists and urologists, as well as fetal medicine specialists, essential in the care of these complex pregnancies. The reader should also bear in mind that local social, ethical and legal aspects may also influence management.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.jpurol.2012.10.012DOI Listing
December 2012

Growth hormone to improve short bowel syndrome intestinal autonomy: a pediatric randomized open-label clinical trial.

JPEN J Parenter Enteral Nutr 2011 Nov 5;35(6):723-31. Epub 2011 Oct 5.

Université Lyon 1, Faculté de Médecine Lyon Est, Lyon, France.

Background: The ability of growth hormone (GH) to promote the weaning-off of parenteral nutrition (PN) in short bowel syndrome (SBS) is unclear. No randomized controlled study is available in children. This study was undertaken to determine if GH could enhance the weaning off of PN in PN-dependent children with SBS.

Methods: A prospective randomized open-label multicenter study was performed in 14 patients (mean age, 9 ± 1.4 years) with SBS (average small bowel length, 33 cm) and long-term PN dependency (8 years) on an unrestricted diet. A standardized PN decrease with and without GH (0.14 mg/kg/d) was conducted. The patients were randomized to either a GH group (4 months of GH) or a control (CTR) group (4 months without GH, followed by 4 months with GH). Blood tests and a nutrition assessment of enteral and parenteral intakes were performed. Groups were compared with the Wilcoxon test.

Results: Treatment with GH did not improve the weaning off of PN (decrease in PN caloric intake of 32.5% ± 9.6% in the GH group vs 35.2% ± 8.7% in the CTR group, nonsignificant). In the CTR group, GH treatment induced an additional but not statistically significant decrease of 8.8% ± 12.4% in daily calories. Parenteral needs returned to near basal rates 6 months after GH discontinuation (GH: 77.6% ± 10.6% vs CTR: 73.2% ± 7.4%). Weight decreased slightly in both groups. No biological parameters varied significantly.

Conclusions: GH did not improve the weaning off of PN in PN-dependent children with SBS.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1177/0148607111415531DOI Listing
November 2011

Studies of a cohort of 46,XY with DSD including steroid biosynthesis deficiencies.

Adv Exp Med Biol 2011 ;707:15-7

Endocrinologie Moléculaire et Maladies Rares, CBPE, Lyon-Bron, France.

View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/978-1-4419-8002-1_4DOI Listing
September 2011