Search

This paper introduces, and reports on the performance of, a novel combination of algorithms for automated microaneurysm (MA) detection in retinal images. The presence of MAs in retinal images is a pathognomonic sign of Diabetic Retinopathy (DR) which is one of the leading causes of blindness amongst the working age population. An extensive survey of the literature is presented and current techniques in the field are summarised. The proposed technique first detects an initial set of candidates using a Gaussian Matched Filter and then classifies this set to reduce the number of false positives. A Tree Ensemble classifier is used with a set of 70 features (the most commons features in the literature). A new set of 32 MA groundtruth images (with a total of 256 labelled MAs) based on images from the MESSIDOR dataset is introduced as a public dataset for benchmarking MA detection algorithms. We evaluate our algorithm on this dataset as well as another public dataset (DIARETDB1 v2.1) and compare it against the best available alternative. Results show that the proposed classifier is superior in terms of eliminating false positive MA detection from the initial set of candidates. The proposed method achieves an ROC score of 0.415 compared to 0.2636 achieved by the best available technique. Furthermore, results show that the classifier model maintains consistent performance across datasets, illustrating the generalisability of the classifier and that overfitting does not occur.

Summary Background The built environment can affect health behaviours, but longitudinal evidence is limited. We aimed to examine the effect of moving into East Village, the former London 2012 Olympic and Paralympic Games Athletes' Village that was repurposed on active design principles, on adult physical activity and adiposity. Methods In this cohort study, we recruited adults seeking new accommodation in East Village and compared physical activity and built environment measures with these data in control participants who had not moved to East Village. At baseline and after 2 years, we objectively measured physical activity with accelerometry and adiposity with body-mass index and bioimpedance, and we assessed objective measures of and participants' perceptions of change in their built environment. We examined the change in physical activity and adiposity between the East Village and control groups, after adjusting for sex, age group, ethnicity, housing tenure, and household (as a random effect). Findings We recruited participants for baseline assessment between Jan 24, 2013, and Jan 7, 2016, and we followed up the cohort after 2 years, between Feb 24, 2015, and Oct 24, 2017. At baseline, 1819 households (one adult per household) consented to initial contact by the study team. 1278 adults (16 years and older) from 1006 (55%) households participated at baseline; of these participants, 877 (69%) adults from 710 (71%) households were assessed after 2 years, of whom 441 (50%) participants from 343 (48%) households had moved to East Village. We found no effect associated with moving to East Village on daily steps, the time spent doing moderate-to-vigorous physical activity (either in total or in 10-min bouts or more), daily sedentary time, body-mass index, or fat mass percentage between participants who had moved to East Village and those in the control group, despite sizeable improvements in walkability and neighbourhood perceptions of crime and quality among the East Village group relative to their original neighbourhood at baseline. Interpretation Despite large improvements in neighbourhood perceptions and walkability, we found no clear evidence that moving to East Village was associated with increased physical activity. Improving the built environment on its own might be insufficient to increase physical activity. Funding National Institute for Health Research and National Prevention Research Initiative.

Causes and effects of differences in image quality with respect to retinal fundus camera images are described. Examples of applications including diabetic retinopathy screening and epidemiology studies are given that demonstrate how assessment of image quality is linked closely to the clinical application in which the images are used. An overview of work in the field is given which covers image quality assessment algorithms, datasets and evaluation metrics. Finally, a selection of image quality assessment algorithms are summarized to illustrate different approaches such as image structure clustering and deep learning.

Background The National Child Measurement Programme (NCMP) assesses patterns of overweight–obesity in English children. It uses body mass index (BMI), which overestimates body fatness in South Asian children and underestimates body fatness in black children of presumed African ethnicity. Using previously derived BMI adjustments, ensuring that adjusted BMI was similarly related to body fatness in South Asian, black, and white children, we reassessed population body fatness, overweight, and obesity patterns in these ethnic groups in the NCMP. Methods Analyses were based on 2012–13 NCMP data in 582 899 children aged 4–5 years and 485 362 children aged 10–11 years. Standard centile-based approaches defined weight status in each age-group before and after applying BMI adjustments, derived previously with the deuterium dilution method, for South Asian and black children. Findings Among white children, overweight–obesity prevalences were 23% (45 508/197 691) for boys and 21% (39 411/188 663) for girls in 4–5 year olds and 33% (52 635/160 278) and 30% (45 978/151 146), respectively, in 10–11 year olds. Before adjustment, South Asian children had lower overweight–obesity prevalences at 4–5 years (19% of boys [4485/23 191], 19% of girls [4176/22 109]; both p Interpretation Adjusted BMI data reveal a substantial excess of overweight–obesity among English South Asian children (at 4–5 years and especially at 10–11 years) and among black girls (aged 10–11 years), with important implications for prevention. These patterns are not apparent with the use of unadjusted BMI data, which systematically underestimated overweight–obesity prevalences in South Asian children and overestimated them in black children. Funding British Heart Foundation, National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care (South London & North Thames).