RESUMO
Data collection, curation, and cleaning constitute a crucial phase in Machine Learning (ML) projects. In biomedical ML, it is often desirable to leverage multiple datasets to increase sample size and diversity, but this poses unique challenges, which arise from heterogeneity in study design, data descriptors, file system organization, and metadata. In this study, we present an approach to the integration of multiple brain MRI datasets with a focus on homogenization of their organization and preprocessing for ML. We use our own fusion example (approximately 84,000 images from 54,000 subjects, 12 studies, and 88 individual scanners) to illustrate and discuss the issues faced by study fusion efforts, and we examine key decisions necessary during dataset homogenization, presenting in detail a database structure flexible enough to accommodate multiple observational MRI datasets. We believe our approach can provide a basis for future similarly-minded biomedical ML projects.
RESUMO
Introduction: Hyperinsulinemia in the absence of impaired glucose tolerance and normal HbA1c is considered indicative of pre-diabetes. Very few Indian studies have focused on hyperinsulinemia particularly in young adults. The present study aimed to determine whether hyperinsulinemia may be present despite HbA1c being normal. Methods: This was a cross-sectional study conducted on adolescents and young adults aged 16-25 years living in Mumbai, India. The participants attended various academic institutions and were those who underwent screening as the first step of a clinical trial for studying the efficacy of almond intake in prediabetes. Results: Among this young population (n=1313), 4.2% (n=55) of the participants were found to be prediabetic (ADA criteria) and 19.7% of them had HbA1c levels between 5.7%-6.4%. However, almost, 30.5% had hyperinsulinemia inspite of normal blood glucose levels and normal HbA1c. Among those with HbA1c<5.7 (n=533), 10.5% (n=56) participants had fasting insulin>15 mIU/L and a higher percentage (39.4%, n=260) had stimulated insulin above 80 mIU/L. These participants had higher mean anthropometric markers than those with normal fasting and/or stimulated insulin. Conclusion: Hyperinsulinaemia in the absence of impaired glucose tolerance and normal HbA1c may provide a much earlier indicator of detection for risk of metabolic disease and progression to metabolic syndrome and diabetes mellitus.
RESUMO
A large percentage of the Indian population has diabetes or is at risk of pre-diabetes. Almond consumption has shown benefits on cardiometabolic risk factors in adults. This study explored the effect of almond consumption on determinants of metabolic dysfunction-blood glucose, lipids, insulin and selected inflammatory markers in adolescents and young adults aged 16-25 years from Mumbai city. This randomized controlled trial was conducted for a period of 90 days on individuals with impaired levels of fasting glucose levels between 100-125 mg/dL (5.6-6.9 mmol/L) and 2-h post-glucose value 140-199 mg/dL (7.8-11.0 mmol/L) and/or fasting insulin (≥15 mIU/ml)/stimulated insulin (≥80 mIU/ml). Of 1,313 individuals screened, 421 met the inclusion criteria, of which 275 consented to participate and 219 completed the trial. The trial was registered with Clinical Trials Registry India (CTRI) CTRI/2018/02/011927. The almonds group (n = 107) consumed 56 g almonds daily, the control group (n = 112) was provided an iso-caloric cereal-pulse based snack. At baseline and endline, blood glucose, insulin, HbA1c, LDL-c, HDL-c, total and ox-cholesterol, triglycerides, hs-CRP, IL-6, TNF-α, adiponectin, leptin were measured and HOMA-IR and FG:FI ratios were calculated. Dietary intakes were assessed. The anthropometric measurements, biochemical markers as well as macronutrient intakes did not differ significantly between the two groups at baseline. Almond consumption significantly decreased HbA1c, total cholesterol and LDL-c. Stimulated insulin decreased post-intervention in both groups, but the decrease was greater in the almonds group. Fasting glucose was reduced post intervention in the controls with no change in the almonds group. FG:FI ratio decreased in the almonds group. TNF-α and IL-6 decreased in the almonds group, while it increased in the control group. Our results showed that almonds reduced HbA1c, LDL-c and total cholesterol levels in just 12 weeks of consumption in these adolescents and young adults who were at risk for developing diabetes. Almonds can be considered as part of food-based strategies for preventing pre-diabetes. Clinical Trial Registration: ClinicalTrials.gov, identifier: CTRI/2018/02/011927.