Quantification of muscle, bones, and fat on single slice thigh CT.

Muscle, bone, and fat segmentation of CT thigh slice is essential for body composition research. Voxel-wise image segmentation enables quantification of tissue properties including area, intensity and texture. Deep learning approaches have had substantial success in medical image segmentation, but they typically require substantial data. Due to high cost of manual annotation, training deep learning models with limited human labelled data is desirable but also a challenging problem. Inspired by transfer learning, we proposed a two-stage deep learning pipeline to address this issue in thigh segmentation. We study 2836 slices from Baltimore Longitudinal Study of Aging (BLSA) and 121 slices from Genetic and Epigenetic Signatures of Translational Aging Laboratory Testing (GESTALT). First, we generated pseudo-labels based on approximate hand-crafted approaches using CT intensity and anatomical morphology. Then, those pseudo labels are fed into deep neural networks to train models from scratch. Finally, the first stage model is loaded as initialization and fine-tuned with a more limited set of expert human labels. We evaluate the performance of this framework on 56 thigh CT scans and obtained average Dice of 0.979,0.969,0.953,0.980 and 0.800 for five tissues: muscle, cortical bone, internal bone, subcutaneous fat and intermuscular fat respectively. We evaluated generalizability by manually reviewing external 3504 BLSA single thighs from 1752 thigh slices. The result is consistent and passed human review with 150 failed thigh images, which demonstrates that the proposed method has strong generalizability.

Diffuse Alveolar Hemorrhage Secondary to Human Metapneumovirus Pulmonary Infection.

A 67-year-old man presented with a week of flu-like symptoms, hypoxia, and fever. Respiratory viral panel was positive for human metapneumovirus. Initial chest imaging showed left lower lobe opacification, suggesting a bacterial superimposed on viral pneumonia. Despite antibiotics, the patient became tachycardic and increasingly hypoxic, requiring 40 L high-flow nasal cannula. Repeat imaging demonstrated worsening of a left lower lobe process. Elective bronchoscopy with bronchoalveolar lavage revealed hemorrhage. Subsequent autoimmune, bacterial, and fungal workup was negative. The patient was diagnosed with diffuse alveolar hemorrhage (DAH) secondary to human metapneumovirus pneumonia. DAH is defined as bleeding into the alveolar spaces of the lungs, a process which carries high rates of morbidity and mortality.1 While dramatic in name and often associated with hemoptysis, DAH may only present with clinically subtle and nonspecific features with a variety of alternative etiologies to consider. We present this case of DAH secondary to human metapneumovirus (hMPV) to promote discussion of etiologies of DAH aside from systemic vasculitis.

Label efficient segmentation of single slice thigh CT with two-stage pseudo labels.

Purpose: Muscle, bone, and fat segmentation from thigh images is essential for quantifying body composition. Voxelwise image segmentation enables quantification of tissue properties including area, intensity, and texture. Deep learning approaches have had substantial success in medical image segmentation, but they typically require a significant amount of data. Due to the high cost of manual annotation, training deep learning models with limited human label data is desirable, but it is a challenging problem. Approach: Inspired by transfer learning, we proposed a two-stage deep learning pipeline to address the thigh and lower leg segmentation issue. We studied three datasets, 3022 thigh slices and 8939 lower leg slices from the BLSA dataset and 121 thigh slices from the GESTALT study. First, we generated pseudo labels for thigh based on approximate handcrafted approaches using CT intensity and anatomical morphology. Then, those pseudo labels were fed into deep neural networks to train models from scratch. Finally, the first stage model was loaded as the initialization and fine-tuned with a more limited set of expert human labels of the thigh. Results: We evaluated the performance of this framework on 73 thigh CT images and obtained an average Dice similarity coefficient (DSC) of 0.927 across muscle, internal bone, cortical bone, subcutaneous fat, and intermuscular fat. To test the generalizability of the proposed framework, we applied the model on lower leg images and obtained an average DSC of 0.823. Conclusions: Approximated handcrafted pseudo labels can build a good initialization for deep neural networks, which can help to reduce the need for, and make full use of, human expert labeled data.

Regular low-calorie sweetener consumption is associated with increased secretion of glucose-dependent insulinotropic polypeptide.

Low-calorie sweeteners (LCSs) are widely used for weight control despite limited evidence of their effectiveness and studies linking LCS consumption with incident obesity. We tested the hypothesis that regular LCS consumption is associated with higher postprandial glucose-dependent insulinotropic polypeptide (GIP) secretion, which has been linked to obesity. We used data from participants in the Baltimore Longitudinal Study of Aging who had completed a diet diary, had at least one visit during which they underwent an oral glucose tolerance test (OGTT), and had no diabetes. Of 232 participants, 166 contributed 1, 39 contributed 2, and 27 contributed 3 visits, and 96 (41%) reported using LCS. Plasma OGTT samples were analysed for glucose, insulin and GIP. Fasting glucose, insulin and GIP levels were no different between LCS users and non-users. The association of LCS use with 2-hour OGTT responses after adjustment for covariates was non-significant for glucose (P = .98) and insulin (P = .18), but significant for greater increase in GIP in LCS users (P = .037). Regular consumption of LCSs was associated with greater increases in GIP secretion after food intake, which may potentially lead to weight gain through the lipogenic properties of GIP.

Chronic Low-Calorie Sweetener Use and Risk of Abdominal Obesity among Older Adults: A Cohort Study.

INTRODUCTION: Low-calorie sweetener use for weight control has come under increasing scrutiny as obesity, especially abdominal obesity, remain entrenched despite substantial low-calorie sweetener use. We evaluated whether chronic low-calorie sweetener use is a risk factor for abdominal obesity. PARTICIPANTS AND METHODS: We used 8268 anthropometric measurements and 3096 food diary records with detailed information on low-calorie sweetener consumption in all food products, from 1454 participants (741 men, 713 women) in the Baltimore Longitudinal Study of Aging collected from 1984 to 2012 with median follow-up of 10 years (range: 0-28 years). At baseline, 785 were low-calorie sweetener non-users (51.7% men) and 669 participants were low-calorie sweetener users (50.1% men). Time-varying low-calorie sweetener use was operationalized as the proportion of visits since baseline at which low-calorie sweetener use was reported. We used marginal structural models to determine the association between baseline and time-varying low-calorie sweetener use with longitudinal outcomes-body mass index, waist circumference, obesity and abdominal obesity-with outcome status assessed at the visit following low-calorie sweetener ascertainment to minimize the potential for reverse causality. All models were adjusted for year of visit, age, sex, age by sex interaction, race, current smoking status, dietary intake (caffeine, fructose, protein, carbohydrate, and fat), physical activity, diabetes status, and Dietary Approaches to Stop Hypertension score as confounders. RESULTS: With median follow-up of 10 years, low-calorie sweetener users had 0.80 kg/m2 higher body mass index (95% confidence interval [CI], 0.17-1.44), 2.6 cm larger waist circumference (95% CI, 0.71-4.39), 36.7% higher prevalence (prevalence ratio = 1.37; 95% CI, 1.10-1.69) and 53% higher incidence (hazard ratio = 1.53; 95% CI 1.10-2.12) of abdominal obesity than low-calorie sweetener non-users. CONCLUSIONS: Low-calorie sweetener use is independently associated with heavier relative weight, a larger waist, and a higher prevalence and incidence of abdominal obesity suggesting that low-calorie sweetener use may not be an effective means of weight control.

The association between leptin and depressive symptoms is modulated by abdominal adiposity.

BACKGROUND: Evidence for a role of leptin in depression is limited and conflicting. Inconclusive findings may be explained by the complex effect of obesity on leptin signaling. In particular, both hyperleptinemia due to leptin resistance in obese persons as well as low leptin in lean persons can imply that low leptin biological signaling is associated with an increased risk of significant depressive symptoms. We tested whether the relationship between leptin and depressive symptoms is modulated by abdominal adiposity in two population-based studies. METHODS: Data were from 851 participants (65-94 years) of the InCHIANTI Study and 1064 (26-93 years) of the Baltimore Longitudinal Study of Aging (BLSA). Plasma concentrations of leptin, waist circumference and depressive symptoms via the Center for Epidemiological Studies-Depression scale (CES-D) were assessed. In longitudinal InCHIANTI analyses onset of depressed mood (CES-D≥20) was evaluated over a 9-year follow-up. RESULTS: In pooled cross-sectional analyses the interaction between leptin and waist circumference was significantly associated with CES-D scores ((log)leptin-by-waist interaction p=0.01). Also in longitudinal analyses, the (log)leptin-by-waist interaction term significantly (p=0.04) predicted depressed mood onset over time; depressed mood risk was especially increased for high levels of both leptin and waist circumference. CONCLUSIONS: The present findings suggest that low leptin signaling rather than low leptin concentration is a risk factor for depression. Future studies should develop proxy measures of leptin signaling by combining information on abdominal adiposity and leptin level to be used for clinical and research applications.

Selective contribution of regional adiposity, skeletal muscle, and adipokines to glucose disposal in older adults.

OBJECTIVES: To study the relationships between muscle mass, regional adiposity, and adipokines and glucose disposal in an older population. DESIGN: Cross-sectional analysis. SETTING: Community-dwelling volunteers from the Baltimore Longitudinal Study of Aging. PARTICIPANTS: Two hundred eighty men and 259 women with a mean age of 71.1 ± 0.4 (range 55-96) and complete data on fasting plasma adiponectin and leptin, oral glucose tolerance test (OGTT) (plasma glucose available at 0, 20, 40, 60, 80, 100, and 120 minutes), thigh computed tomography (CT), physical activity levels, and anthropometric measures. MEASUREMENTS: Participants were classified into eight groups according to the presence of global adiposity (body mass index > 27 kg/m(2)), central adiposity (waist circumference > 88 cm for women and > 102 cm for men), and low muscle mass (CT thigh, lowest sex-specific tertile (93.8 cm(2) in women and 110.7 cm(2) in men) of adjusted thigh muscle area). Linear regression models were used to estimate the contribution of these eight groups to early glucose area under the curve (AUC) (t = 0-40 minutes), late glucose AUC (t = 60-120 minutes), and total glucose AUC (t = 0-120 minutes) from the OGTT. RESULTS: Regardless of muscle mass, individuals with a combination of central and global adiposity were more likely to have delayed glucose disposal rates (P < .05). A strong negative association was also found between circulating adiponectin levels and glucose disposal rates (early AUC, ß = -0.14; late AUC, ß = -0.20; and total AUC, ß = -0.20; P < .05 for all three AUCs) after adjusting for regional adiposity, muscle mass, circulating leptin levels, physical activity, age, and sex. CONCLUSION: Older individuals with global and central adiposity may be at risk of glucose intolerance unrelated to low muscle mass.

Cross-sectional evidence of a signaling pathway from bone homeostasis to glucose metabolism.

CONTEXT: Preclinical studies suggested the existence of a signaling pathway connecting bone and glucose metabolisms. Supposedly leptin modulates osteocalcin bioactivity, which in turn stimulates insulin and adiponectin secretion, and ß-cell proliferation. OBJECTIVE: The objective of the investigation was to study the reciprocal relationships of adiponectin, leptin, osteocalcin, insulin resistance, and insulin secretion to verify whether such relationships are consistent with a signaling pathway connecting bone homeostasis and glucose metabolism. DESIGN: This was a cross-sectional analysis. SETTING: The study was conducted with community-dwelling volunteers participating in the Baltimore Longitudinal Study of Aging. PARTICIPANTS: Two hundred eighty women and 300 men with complete data on fasting plasma adiponectin, leptin, and osteocalcin, oral glucose tolerance test (plasma glucose and insulin values available at t = 0, 20, and 120 min), and anthropometric measures participated in the study. MAIN OUTCOME MEASURES: Linear regression models were used to test independent associations of adiponectin, osteocalcin, and leptin with the indices of insulin resistance and secretion. The expected reciprocal relationship between different biomarkers was verified by structural equation modeling. RESULTS: In linear regression models, leptin was strongly associated with indices of both insulin resistance and secretion. Both adiponectin and osteocalcin were negatively associated with insulin resistance. Structural equation modeling revealed a direct inverse association of leptin with osteocalcin; a direct positive association of osteocalcin with adiponectin; and an inverse relationship of osteocalcin with insulin resistance and adiponectin with insulin resistance and secretion, which is cumulatively consistent with the hypothesized model. CONCLUSIONS: Bone and glucose metabolisms are probably connected through a complex pathway that involves leptin, osteocalcin, and adiponectin. The clinical relevance of such a pathway for bone pathology in diabetes should be further investigated.