Expert Grading Research Articles

Validation and Clinical Applicability of Whole-Volume Automated Segmentation of Optical Coherence Tomography in Retinal Disease Using Deep Learning

Quantitative volumetric measures of retinal disease in optical coherence tomography (OCT) scans are infeasible to perform owing to the time required for manual grading. Expert-level deep learning systems for automatic OCT segmentation have recently been developed. However, the potential clinical applicability of these systems is largely unknown. To evaluate a deep learning model for whole-volume segmentation of 4 clinically important pathological features and assess clinical applicability. This diagnostic study used OCT data from 173 patients with a total of 15 558 B-scans, treated at Moorfields Eye Hospital. The data set included 2 common OCT devices and 2 macular conditions: wet age-related macular degeneration (107 scans) and diabetic macular edema (66 scans), covering the full range of severity, and from 3 points during treatment. Two expert graders performed pixel-level segmentations of intraretinal fluid, subretinal fluid, subretinal hyperreflective material, and pigment epithelial detachment, including all B-scans in each OCT volume, taking as long as 50 hours per scan. Quantitative evaluation of whole-volume model segmentations was performed. Qualitative evaluation of clinical applicability by 3 retinal experts was also conducted. Data were collected from June 1, 2012, to January 31, 2017, for set 1 and from January 1 to December 31, 2017, for set 2; graded between November 2018 and January 2020; and analyzed from February 2020 to November 2020. Rating and stack ranking for clinical applicability by retinal specialists, model-grader agreement for voxelwise segmentations, and total volume evaluated using Dice similarity coefficients, Bland-Altman plots, and intraclass correlation coefficients. Among the 173 patients included in the analysis (92 [53%] women), qualitative assessment found that automated whole-volume segmentation ranked better than or comparable to at least 1 expert grader in 127 scans (73%; 95% CI, 66%-79%). A neutral or positive rating was given to 135 model segmentations (78%; 95% CI, 71%-84%) and 309 expert gradings (2 per scan) (89%; 95% CI, 86%-92%). The model was rated neutrally or positively in 86% to 92% of diabetic macular edema scans and 53% to 87% of age-related macular degeneration scans. Intraclass correlations ranged from 0.33 (95% CI, 0.08-0.96) to 0.96 (95% CI, 0.90-0.99). Dice similarity coefficients ranged from 0.43 (95% CI, 0.29-0.66) to 0.78 (95% CI, 0.57-0.85). This deep learning-based segmentation tool provided clinically useful measures of retinal disease that would otherwise be infeasible to obtain. Qualitative evaluation was additionally important to reveal clinical applicability for both care management and research.

Clinical Evidence of Cell-Targeted Topical Therapy for Treating Skin Dyspigmentation.

New development of cell-targeted therapies to enable site-specific skin tissue drug delivery may reduce off-target effects, decrease unwanted toxicities, and enhance drug efficacy. These efforts have led to several targeting strategies that modulate active product delivery to include small molecule-, nucleic acid-, peptide-, antibody-, and cell-based strategies. Tissue specific cell-targeting strategies such as these may be useful in cosmetic dermatologic applications. The aim of this 16-week clinical trial of a skin brightening composition containing melanocyte cell-targeted biodelivery was to assess its effectiveness in restoring the skin complexion evenness by modulating melanocyte activity in a cohort of 50 Fitzpatrick type I–VI subjects with moderate to severe dyspigmentation. Data from expert grading, skin surface colorimetry, and subject self-assessments reflected significant improvement in facial skin tone as early as 2 weeks after treatment initiation, with continual improvement through week 16. The most dramatic pigmentation improvement, based on investigator assessments, was a statistically significant improvement in skin brightness at week 2 that progressed to week 8 with significant improvement in skin evenness and brightness. By weeks 12 and 16, progressive levels of significant improvement in skin evenness and brightening became apparent. Colorimetry demonstrated progressive improvement in skin dyspigmentation starting at 2 weeks and continuing to week 16. Subject self-assessment data supported similar improvements in skin dyspigmentation. These results demonstrate the ability of a cell-targeted topical therapy to achieve improvements in skin pigmentation through site-specific suppression of melanocyte activity. J Drugs Dermatol. 2021;20(8):865-867. oi:10.36849/JDD.6037 THIS ARTICLE HAD BEEN MADE AVAILABLE FREE OF CHARGE. PLEASE SCROLL DOWN TO ACCESS THE FULL fTEXT OF THIS ARTICLE WITHOUT LOGGING IN. NO PURCHASE NECESSARY. PLEASE CONTACT THE PUBLISHER WITH ANY QUESTIONS.

Determinants of grader agreement: an analysis of multiple short answer corpora

The ’short answer’ question format is a widely used tool in educational assessment, in which students write one to three sentences in response to an open question. The answers are subsequently rated by expert graders. The agreement between these graders is crucial for reliable analysis, both in terms of educational strategies and in terms of developing automatic models for short answer grading (SAG), an active research topic in NLP. This makes it important to understand the properties that influence grader agreement (such as question difficulty, answer length, and answer correctness). However, the twin challenges towards such an understanding are the wide range of SAG corpora in use (which differ along a number of dimensions) and the hierarchical structure of potentially relevant properties (which can be located at the corpus, answer, or question levels). This article uses generalized mixed effects models to analyze the effect of various such properties on grader agreement in six major SAG corpora for two main assessment tasks (language and content assessment). Overall, we find broad agreement among corpora, with a number of properties behaving similarly across corpora (e.g., shorter answers and correct answers are easier to grade). Some properties show more corpus-specific behavior (e.g., the question difficulty level), and some corpora are more in line with general tendencies than others. In sum, we obtain a nuanced picture of how the major short answer grading corpora are similar and dissimilar from which we derive suggestions for corpus development and analysis.

Sacubitril-valsartan initiation in chronic heart failure patients impacts sleep apnea: the ENTRESTO-SAS study.

AimsOptimizing medical cardiac treatment for sleep apnoea (SA) in patients with chronic heart failure and reduced ejection fraction (HFrEF) is an expert Grade C recommendation based on six studies encompassing a total of 67 patients only. Whether sacubitril–valsartan (SV), a cornerstone of HFrEF medical treatment, impacts SA is unknown and requires evaluation.Methods and resultsThe ENTRESTO‐SAS trial is a six‐centre, prospective, open‐label real‐life cohort study (NCT02916160). Ambulatory patients eligible for SV (i.e. HFrEF adults who remain symptomatic despite optimal treatment) were evaluated before and after 3 months of SV (including nocturnal ventilatory polygraphy); 118 patients were final analysed [median age was 66 (IQ25–75: 56–73) years, 81.4% male, 36.5% New York Heart Association III–IV, N‐terminal pro‐B‐type natriuretic peptide level of 1564 (701–3376) ng/L, left ventricular ejection fraction of 30 (25–34)%, 60.7% ischaemic HFrEF, 97.5% initially treated with angiotensin‐converting enzyme inhibitors or angiotensin II receptor blockers, 83.9% with beta‐blockers, 64.4% with mineralocorticoid receptor antagonists, and 74.6% with diuretics]. Three groups were defined according to initial central/obstructive apnoea–hypopnoea indices (AHIs): G1 (n = 49, AHIcentral ≥ 5/h and AHIobstructive < 15/h); G2 (n = 27, AHIobstructive ≥ 15/h); and G3 (n = 42, AHIcentral < 5/h and AHIobstructive < 15/h). At 3 months, the AHI (main predefined outcome) decreased significantly by −7.10/h (IQ25–75: −16.10 to 0.40; P < 0.001) in G1 + G2 without positive airway pressure treatment (45 patients, median initial AHI of 24.20 (IQ25–75: 16.40–43.50)/h). Of these, 24.4% presented an AHI decrease ≥50% and 37.78% had a final AHI < 15/h (tendency for improvement from an initial value of 20%: P = 0.0574). For G1 patients (n = 37), AHI significantly decreased from a median of 22.90 (16.00–43.50)/h to 19.20 (12.70–31.10)/h (P = 0.002). For G2 patients (n = 8), AHI decreased from a median of 30.10 (26.40–47.60)/h to 22.75 (14.60–36.90)/h (statistically non‐significant, P = 0.059).ConclusionsIn this real‐life population, SV treatment for 3 months in SA patients is associated with a significant decrease in AHI. These results support the current guidelines that recommend first an optimization of the HFrEF treatment in patients with HFrEF and central SA. A potential positive airway pressure sparing effect merits further investigation.

Prospective, Longitudinal Pilot Study: Daily Self-Imaging with Patient-Operated Home OCT in Neovascular Age-Related Macular Degeneration.

PurposeTo evaluate longitudinally the performance of the Notal Vision Home OCT (NVHO), comprising a spectral-domain OCT device for patient self-imaging at home, telemedicine infrastructure for automated data upload, and deep learning algorithm for automated OCT evaluation. The aims were to study the system’s performance in daily image acquisition and automated analysis and to characterize the dynamics of retinal fluid exudation in neovascular age-related macular degeneration (nAMD).DesignPilot prospective, observational longitudinal study.ParticipantsFour individuals (mean age, 73.8 years) with nAMD (one or both eyes) undergoing anti–vascular endothelial growth factor therapy in routine clinical practice.MethodsThe participants performed daily self-imaging at home with the NVHO for 1 month. The macular cube scans were uploaded automatically to the Notal Health Cloud. They underwent evaluation separately by the Notal OCT Analyzer (NOA) and human expert graders for fluid presence, segmentation, and volume.Main Outcome MeasuresDaily self-imaging completion, image quality, acquisition time, agreement between automated and human grading of retinal fluid, and temporal dynamics of fluid volume.ResultsOf 240 self-imaging attempts initiated, the number successfully completed was 211 (87.9%). Of these, 97.6% had satisfactory quality. For fluid presence, the NOA agreed with human grading in 94.7% of cases. From a subset of 24 scans with fluid, for agreement between NOA and human fluid volume measurements, the correlation coefficient was 0.996 and mean absolute difference was 1.5 nl (vs. 0.995 and 1.2 nl, respectively, for interhuman agreement). Graphic plots of fluid volume revealed wide variation in the dynamics of fluid exudation and treatment response.ConclusionsThe participants could perform daily self-imaging at home and generate macular cube scans of satisfactory quality. Automated quantitative OCT analysis achieved high agreement with human grading. Daily self-imaging with automated OCT analysis permitted detailed characterization of the dynamics of fluid exudation and revealed wide variation between eyes. Metrics describing these dynamics may become important disease biomarkers. Home OCT telemedicine systems represent an alternative paradigm of disease monitoring; they may allow highly personalized retreatment decisions, with fewer unnecessary injections and clinic visits.

A system for measuring the satisfaction of railway heritage tourism: the case of Yunnan–Vietnam Railway

Abstract As a representative historical railway in Southeast Asia, Yunnan–Vietnam Railway (YVR) is of historical, technological, social-cultural, landscape, architectural, and heritage values. However, some of its remains have been closed and destroyed, waiting for further protection and developing solutions. YVR is the research object in this paper for the process of measuring tourists’ satisfaction. This paper highlights the development of an indicator system for measuring the satisfaction degree of the YVR heritage tourism, with the help of the analytical hierarchy process and the expert grading method. The basic demographic information of tourists collected from the questionnaires is systemized, and the satisfaction degree is evaluated from five indicators based on previous research: natural landscape, cultural character, transportation and facilities, safety and services, as well as the sustainable development of the region. The analysis results gained by SPSS show a high satisfaction value on the safety and service and regional sustainable development. Namely, the tourists give higher ratings on the regional cooperation, personal safety, service staff, hospitality of locals, and the technology of railway, while the aspects of environmental quality, enjoyment of exploring, accessibility by public transports, the comfort of train experience, and related facilities (accommodation/restaurants/shopping) received lower ratings, which indicates the critical points for the future tourism and urban planning related to this historical railway.

An Artificial-Intelligence- and Telemedicine-Based Screening Tool to Identify Glaucoma Suspects from Color Fundus Imaging.

Results The system achieved an accuracy of 89.67% (sensitivity, 83.33%; specificity, 93.89%; and AUC, 0.93). For external validation, the Retinal Fundus Image Database for Glaucoma Analysis dataset, which has 638 gradable quality images, was used. Here, the model achieved an accuracy of 83.54% (sensitivity, 80.11%; specificity, 84.96%; and AUC, 0.85). Conclusions Having demonstrated an accurate and fully automated glaucoma-suspect screening system that can be deployed on telemedicine platforms, we plan prospective trials to determine the feasibility of the system in primary-care settings.

Capturing the clinical decision-making processes of expert and novice diabetic retinal graders using a \u2018think-aloud\u2019 approach

BackgroundDiabetic eye screening programmes have been developed worldwide based on evidence that early detection and treatment of diabetic retinopathy are crucial to preventing sight loss. However, little is known about the decision-making processes and training needs of diabetic retinal graders, particularly in low- and middle-income countries.ObjectivesTo provide data for improving evidence-based diabetic retinopathy training to help novice graders process fundus images more like experts.Subjects/methodsThis is a mixed-methods qualitative study conducted in southern Vietnam and Northern Ireland. Novice diabetic retinal graders in Vietnam (n = 18) and expert graders in Northern Ireland (n = 5) were selected through a purposive sampling technique. Data were collected from 21st February to 3rd September 2019. The interviewer used neutral prompts during think-aloud sessions to encourage participants to verbalise their thought processes while grading fundus images from anonymised patients, followed by semi-structured interviews. Thematic framework analysis was used to identify themes, supported by illustrative quotes from interviews. Mann–Whitney U tests were used to compare graders’ performance.ResultsExpert graders used a more systematic approach when grading images, considered all four images per patient and used available software tools such as red-free filters prior to making a decision on management. The most challenging features for novice graders were intra-retinal microvascular abnormalities and new vessels, which were more accurately identified by experts.ConclusionTaking more time to grade fundus images and adopting a protocol-driven “checklist” approach may help novice graders to function more like experts.

Effective reduction of post-inflammatory hyperpigmentation with the tyrosinase inhibitor isobutylamido-thiazolyl-resorcinol (Thiamidol).

ObjectivePost‐inflammatory hyperpigmentation (PIH) is a major cosmetic concern especially in individuals with darker skin complexion. Unfortunately, treatment with anti‐inflammatory ingredients alone does not prevent the development of hyperpigmented spots. Recently, isobutylamido‐thiazolyl‐resorcinol (Thiamidol) was described as a very potent inhibitor of human tyrosinase. The objective of this research was to investigate the potential of this compound to prevent PIH induced by epidermal wounding (suction blister) and related to acne.MethodsSuction blister‐induced PIH was treated with a formulation containing Thiamidol or a vehicle for 3 months, and the changes in hyperpigmentation were monitored by spectroscopic measurements. The effect of skin care formulations containing Thiamidol on acne‐related PIH was investigated in two studies, a vehicle‐controlled, double‐blinded, randomized clinical study and a clinical observational study. Both studies had a duration of 3 months and included assessments such as clinical photography, clinical grading and melanin index measurements.ResultsAlready after 2 weeks of treatment, suction blister sites treated with Thiamidol were significantly lighter than control sites and improved throughout the treatment period. Subjects´ self‐grading demonstrated that Thiamidol significantly improved the visibility of acne‐induced hyperpigmentation compared to the vehicle treatment. A skin care regimen with Thiamidol significantly improved acne‐related PIH over 12 weeks shown by Mexameter measurements, expert grading, self‐grading and clinical photography.ConclusionThiamidol represents a safe and effective ingredient for cosmetic products against post‐inflammatory hyperpigmentation.

Automatic Segmentation in Multiple OCT Layers For Stargardt Disease Characterization Via Deep Learning.

PurposeThis study sought to perform automated segmentation of 11 retinal layers and Stargardt-associated features on spectral-domain optical coherence tomography (SD-OCT) images and to analyze differences between normal eyes and eyes diagnosed with Stargardt disease.MethodsAutomated segmentation was accomplished through application of the deep learning–shortest path (DL-SP) framework, a shortest path segmentation approach that is enhanced by a deep learning fully convolutional neural network. To compare normal eyes and eyes diagnosed with Stargardt disease, various retinal layer thickness and intensity feature maps associated with the outer retinal layers were generated.ResultsThe automated DL-SP approach achieved a mean difference within a subpixel accuracy range for all layers when compared to manually traced layers by expert graders. The algorithm achieved mean and absolute mean differences in border positions for Stargardt features of −0.11 ± 4.17 pixels and 1.92 ± 3.71 pixels, respectively. In several of the feature maps generated, the characteristic Stargardt features of flecks and atrophic-appearing lesions were readily visualized.ConclusionsTo the best of our knowledge, this is the first automated algorithm for 11 retinal layer segmentation on OCT in eyes with Stargardt disease, and, furthermore, the feature differences found between eyes diagnosed with Stargardt disease and normal eyes may inform new insights and the better understanding of retinal characteristic morphologic changes caused by Stargardt disease.Translational RelevanceThe automated algorithm's performance and the feature differences found using the algorithm's segmentation support the future applications of SD-OCT for the quantitative monitoring of Stargardt disease.

Retinal photograph-based deep learning algorithms for myopia and a blockchain platform to facilitate artificial intelligence medical research: a retrospective multicohort study

By 2050, almost 5 billion people globally are projected to have myopia, of whom 20% are likely to have high myopia with clinically significant risk of sight-threatening complications such as myopic macular degeneration. These are diagnoses that typically require specialist assessment or measurement with multiple unconnected pieces of equipment. Artificial intelligence (AI) approaches might be effective for risk stratification and to identify individuals at highest risk of visual loss. However, unresolved challenges for AI medical studies remain, including paucity of transparency, auditability, and traceability. In this retrospective multicohort study, we developed and tested retinal photograph-based deep learning algorithms for detection of myopic macular degeneration and high myopia, using a total of 226 686 retinal images. First we trained and internally validated the algorithms on datasets from Singapore, and then externally tested them on datasets from China, Taiwan, India, Russia, and the UK. We also compared the performance of the deep learning algorithms against six human experts in the grading of a randomly selected dataset of 400 images from the external datasets. As proof of concept, we used a blockchain-based AI platform to demonstrate the real-world application of secure data transfer, model transfer, and model testing across three sites in Singapore and China. The deep learning algorithms showed robust diagnostic performance with areas under the receiver operating characteristic curves [AUC] of 0·969 (95% CI 0·959-0·977) or higher for myopic macular degeneration and 0·913 (0·906-0·920) or higher for high myopia across the external testing datasets with available data. In the randomly selected dataset, the deep learning algorithms outperformed all six expert graders in detection of each condition (AUC of 0·978 [0·957-0·994] for myopic macular degeneration and 0·973 [0·941-0·995] for high myopia). We also successfully used blockchain technology for data transfer, model transfer, and model testing between sites and across two countries. Deep learning algorithms can be effective tools for risk stratification and screening of myopic macular degeneration and high myopia among the large global population with myopia. The blockchain platform developed here could potentially serve as a trusted platform for performance testing of future AI models in medicine. None.

ARTICLE: Efficacy of Ceramide-Containing Formulations on UV-Induced Skin Surface Barrier Alterations.

The human skin, particularly the stratum corneum, serves as a protective barrier against exogenous factors, including ultraviolet radiation (UVR) and pathogen invasions. The impact of UVR on skin cancer and photoaging has been extensively studied. However, the direct impact of UVR on skin barrier integrity under clinical settings remains poorly explored. Due to their benefits in reducing inflammation and promoting skin barrier repair, ceramide-containing formulations can provide added photoprotection benefits. In this study, the efficacy of a ceramide-containing sunscreen and moisturizer were evaluated in preventing UV-induced skin surface barrier changes. Expert grading, instrumental, and tape-stripping assessments demonstrated that UVR induced erythema and hyperpigmentation and caused changes in skin cells surface morphological organization and maturation. Treatment with a ceramide-containing sunscreen and moisturizing cream routine reduced erythema and hyperpigmentation, improved skin hydration, and maintained normal superficial skin cells morphology and turnover after UVR. Our results indicate that barrier-enforcing lipids formulations can provide additional benefits in patient&rsquo;s daily routine by strengthening the barrier and improving skin health overall against chronic sun exposure. J Drugs Dermatol. 20(4 Suppl):s29-35. doi:10.36849/JDD.S589E.

70740 Impact of fundus pigmentation on retinal layer visibility on investigational bedside optical coherence tomography in preterm infants

ABSTRACT IMPACT: This study helps translate investigational bedside optical coherence tomography into improved diagnosis and care of preterm infants at risk for retinopathy of prematurity. OBJECTIVES/GOALS: In retinopathy of prematurity screening, fundus photography may be of limited quality in eyes with dark fundus pigmentation (FP). The goal of this study was to evaluate the impact of FP on overall scan quality and retinal layer visibility on investigational bedside optical coherence tomography (OCT) in preterm infants. METHODS/STUDY POPULATION: We analyzed 846 OCT scans captured prospectively from 188 eyes of 94 preterm infants enrolled in the BabySTEPS study (NCT02887157). Trained ophthalmologists, masked to OCT findings, determined FP (blond, medium, or dark). Expert graders, masked to FP, evaluated OCT images for: 1) overall OCT quality (excellent, acceptable, poor, or unusable); and 2) all age-appropriate retinal layers visible (yes or no). To assess the association of FP with OCT quality (excellent/acceptable or poor/unusable) and retinal layer visibility, we performed multivariable logistic regression modeling, adjusting for biologic and demographic confounders and correlations from repeated OCT scans and paired eyes. RESULTS/ANTICIPATED RESULTS: Mean birthweight was 964 (SD=283) grams, mean gestational age was 27.8 (SD=2.6) weeks, 48 (51%) infants were male, and 51 (54%) were non-white. On exam, 72 (38%) eyes had blond FP, 92 (49%) had medium, and 24 (13%) had dark. OCT quality was excellent or acceptable in 725 scans (86%) and all age-appropriate retinal layers were visible in 781 scans (92%). Compared to eyes with blond FP, eyes with medium and dark FP did not have higher odds of poor/unusable OCT scan quality (adjusted OR 0.87 [95% CI 0.50-1.48] and 0.49 [95% CI 0.16-1.55], respectively) or not all age-appropriate retinal layers visible on OCT (adjusted OR 1.17 [95% CI 0.39-3.51] and 0.57 [95% CI 0.15-2.20], respectively). DISCUSSION/SIGNIFICANCE OF FINDINGS: Medium and dark FP did not impact overall scan quality or age-appropriate retinal layer visibility on investigational bedside OCT in preterm infants. This study supports the feasibility of using OCT to analyze retinal microanatomy in diverse populations of preterm infants with a range of FP.

ARTICLE: Efficacy of Ceramide-Containing Formulations on UV-Induced Skin Surface Barrier Alterations

The human skin, particularly the stratum corneum, serves as a protective barrier against exogenous factors, including ultraviolet radiation (UVR) and pathogen invasions. The impact of UVR on skin cancer and photoaging has been extensively studied. However, the direct impact of UVR on skin barrier integrity under clinical settings remains poorly explored. Due to their benefits in reducing inflammation and promoting skin barrier repair, ceramide-containing formulations can provide added photoprotection benefits. In this study, the efficacy of a ceramide-containing sunscreen and moisturizer were evaluated in preventing UV-induced skin surface barrier changes. Expert grading, instrumental, and tape-stripping assessments demonstrated that UVR induced erythema and hyperpigmentation and caused changes in skin cells surface morphological organization and maturation. Treatment with a ceramide-containing sunscreen and moisturizing cream routine reduced erythema and hyperpigmentation, improved skin hydration, and maintained normal superficial skin cells morphology and turnover after UVR. Our results indicate that barrier-enforcing lipids formulations can provide additional benefits in patient’s daily routine by strengthening the barrier and improving skin health overall against chronic sun exposure. J Drugs Dermatol. 20(4 Suppl):s29-35. doi:10.36849/JDD.S589E

Deep Learning for the Diagnosis of Stage in Retinopathy of Prematurity: Accuracy and Generalizability across Populations and Cameras

Stage is an important feature to identify in retinal images of infants at risk of retinopathy of prematurity (ROP). The purpose of this study was to implement a convolutional neural network (CNN) for binary detection of stages 1, 2, and 3 in ROP and to evaluate its generalizability across different populations and camera systems. Diagnostic validation study of CNN for stage detection. Retinal fundus images obtained from preterm infants during routine ROP screenings. Two datasets were used: 5943 fundus images obtained by RetCam camera (Natus Medical, Pleasanton, CA) from 9 North American institutions and 5049 images obtained by 3nethra camera (Forus Health Incorporated, Bengaluru, India) from 4 hospitals in Nepal. Images were labeled based on the presence of stage by 1 to 3 expert graders. Three CNN models were trained using 5-fold cross-validation on datasets from North America alone, Nepal alone, and a combined dataset and were evaluated on 2 held-out test sets consisting of 708 and 247 images from the Nepali and North American datasets, respectively. Convolutional neural network performance was evaluated using area under the receiver operating characteristic curve (AUROC), area under the precision-recall curve (AUPRC), sensitivity, and specificity. Both the North American- and Nepali-trained models demonstrated high performance on a test set from the same population: AUROC, 0.99; AUPRC, 0.98; sensitivity, 94%; and AUROC, 0.97; AUPRC, 0.91; and sensitivity, 73%; respectively. However, the performance of each model decreased to AUROC of 0.96 and AUPRC of 0.88 (sensitivity, 52%) and AUROC of 0.62 and AUPRC of 0.36 (sensitivity, 44%) when evaluated on a test set from the other population. Compared with the models trained on individual datasets, the model trained on a combined dataset achieved improved performance on each respective test set: sensitivity improved from 94% to 98% on the North American test set and from 73% to 82% on the Nepali test set. A CNN can identify accurately the presence of ROP stage in retinal images, but performance depends on the similarity between training and testing populations. We demonstrated that internal and external performance can be improved by increasing the heterogeneity of the training dataset features of the training dataset, in this case by combining images from different populations and cameras.

Printed artificial sweat as replacement for natural fingermarks: Qualitative and quantitative approach considering an amino acid reagent

The study presented in this paper aims at assessing how printed fingermarks can be used to generate realistic latent marks bearing varying quantities of materials to be detected. Considering dilution series of artificial sweat (eccrine secretion) and 1,2-indanedione/zinc as amino acid reagent, we assessed how printed marks behave in comparison to natural fingermarks provided by a set of 30 donors. The results were assessed in terms of relative intensity (contrast, luminescence) and expert grading (ridge details, overall quality). With regards to the set of 30 donors, this study brought a quantitative look to the influence of intra- and inter-variability on the relative intensity values observed when processing natural fingermarks. This provided new data to further understand the concept of “donorship”. With regards to the use of printed marks, it has been illustrated how dilution series of a concentrated solution allows covering a range of cases: unnatural marks (intensity values well above those obtained with donors), rich marks (corresponding to fingermarks left by good donors), and faint marks (associated with the kind of results observed with poor donors). Such a range of detection performance offers the possibility to generate fine-tuned detection exercises of varying difficulty levels. Printed items made of artificial sweat could hence constitute a valuable alternative to natural secretions in the context of education and proficiency testing.

Facial Expression Wrinkles and Their Relaxation by a Synthetic Peptide

Expression wrinkles form over time due to repeated facial movements such as smiling and frowning. They have an imprint on facial skin in areas such as the corner of the eyes, where they take the form of crow’s feet, the forehead and the glabella, where they appear as frown lines, and around the mouth, as marionette lines. In the study presented here, we recruited two sets of volunteers. An older group of 57 volunteers aged 50 to 65 years, and a group of eight younger volunteers aged 21–35 who were the biological daughters of eight of the older volunteers. Using VISIA CR, we took images of the volunteers in relaxed, angry and smiling mode to assess similarities in expression wrinkle patterns. In addition, the older volunteers were split into a placebo group and an active group who applied a formulation of 4% of a cosmetic product containing the peptide diaminobutyroyl benzylamide diacetate (DABBA) for four weeks. Wrinkles were assessed by image analysis, expert grading and Primoslite measurements. Our study found striking similarities in the facial wrinkle patterns of mothers with relaxed faces and daughters with angry or smiling faces. We found a decrease in visible wrinkles in the group of older volunteers applying DABBA. We created a facial map for graded wrinkles showing these changes. Volunteers using the active formulation showed significantly less wrinkle area and length on the forehead when frowning compared to the placebo group (p < 0.05).

A deep-learning system for the assessment of cardiovascular disease risk via the measurement of retinal-vessel calibre.

Retinal blood vessels provide information on the risk of cardiovascular disease (CVD). Here, we report the development and validation of deep-learning models for the automated measurement of retinal-vessel calibre in retinal photographs, using diverse multiethnic multicountry datasets that comprise more than 70,000 images. Retinal-vessel calibre measured by the models and by expert human graders showed high agreement, with overall intraclass correlation coefficients of between 0.82 and 0.95. The models performed comparably to or better than expert graders in associations between measurements of retinal-vessel calibre and CVD risk factors, including blood pressure, body-mass index, total cholesterol and glycated-haemoglobin levels. In retrospectively measured prospective datasets from a population-based study, baseline measurements performed by the deep-learning system were associated with incident CVD. Our findings motivate the development of clinically applicable explainable end-to-end deep-learning systems for the prediction of CVD on the basis of the features of retinal vessels in retinal photographs.

Serial theft case linkage based on a two‐step cumulative probability model

AbstractAt present, serial theft case linkage remains at the stage of empiricism. In order to overcome this subjective arbitrariness, this study proposes using a ‘two‐step cumulative probability model’ for investigation. In the first step, based on expert grading method, the opinions of 99 policemen were combined to quantify the serial theft case characteristics. In the second step, when a new case occurred, the characteristics of it were compared with the characteristics of each serial theft case, and the corresponding probabilities were added according to the calculations of the second step; when the accumulated points exceeded the threshold, we considered concatenating the new case with the corresponding serial cases. The results demonstrated that the average accuracy of the two‐step cumulative probability model was 87.5% and that the average response rate of the irrelevant case (control group) was 12.3%. We concluded that the two‐step cumulative probability model could assist in criminal investigations.

Standardized disaster and climate resilience grading: A global scale empirical analysis of community flood resilience

Suitable and standardized indicators to track progress in disaster and climate resilience are increasingly considered a key requirement for successfully informing efforts towards effective disaster risk reduction and climate adaptation. Standardized measures of resilience which can be used across different geographical and socioeconomic contexts are however sparse. We present and analyze a standardized community resilience measurement framework for flooding. The corresponding measurement tool is modelled based on and adapted from a so-called ‘technical risk grading’ approach as used in the insurance sector. The grading approach of indicators is based on a two-step process: (i) raw data is collected, and (ii) experts grade the indicators, called sources of resilience, based on this data. We test this approach using approximately 1.25 million datapoints collected across more than 118 communities in nine countries. The quantitative analysis is complemented by content analysis to validate the results from a qualitative perspective. We find that some indicators can more easily be graded by looking at raw data alone, while others require a stronger application of expert judgement. We summarize the reasons for this through six key messages. One major finding is that resilience grades related to subjective characteristics such as ability, feel, and trust are far more dependent on expert judgment than on the actual raw data collected. Additionally, the need for expert judgement further increases if graders must extrapolate the whole community picture from limited raw data. Our findings regarding the role of data and grade specifications can inform ways forward for better, more efficient and increasingly robust standardized assessment of resilience. This should help to build global standardized and comparable, yet locally contextualized, baseline estimates of the many facets of resilience in order to track progress over time on disaster and climate resilience and inform the implementation of the Paris Agreement, Sendai Framework, and the Sustainable Development Goals.