Automated segmentation of abdominal subcutaneous adipose tissue and visceral adipose tissue in obese adolescent in MRI

Gamma-glutamyl transferase: A potential biomarker for pancreas steatosis in patients with concurrent obesity, insulin resistance and metabolic dysfunction-associated steatotic liver disease

To evaluate the relationship between serum gamma-glutamyl transferase (GGT) levels and fatty pancreas in subjects with concurrent obesity, insulin resistance and metabolic dysfunction-associated steatotic liver disease (MASLD) without a history of pancreatitis. From March 2019 to September 2021, 31 adult subjects with concurrent obesity and MASLD were recruited as part of the study investigating the biological impact of bariatric surgery and lifestyle modification on obesity. Chemical shift encoded MRI of the abdomen, LiverMultiScan, anthropometric, clinical and blood biochemistry analyses were performed prior to any intervention at baseline. GGT (p <.001) was significantly different between those 'with fatty pancreas' and 'without fatty pancreas' groups. GGT (p <.001) was significantly different between those 'with both metabolic syndrome and fatty pancreas' and those 'with metabolic syndrome but without fatty pancreas.' GGT (p <.001) was also significantly different between those 'with both diabetes and fatty pancreas' and those 'with diabetes but without fatty pancreas'. Logistic regression analysis showed that abnormal GGT levels (p = .010) and Hypertension (p = .045) were significant independent predictors of fatty pancreas. GGT was associated with fatty pancreas by an odds ratio 7.333 (95% [CI]: 1.467-36.664), while the AUROC of GGT in determining fatty pancreas was 0.849. Elevation in serum GGT might be a potential marker to identify fatty pancreas.

Automated abdominal adipose tissue segmentation and volume quantification on longitudinal MRI using 3D convolutional neural networks with multi-contrast inputs

OBJECTIVE

Increased subcutaneous and visceral adipose tissue (SAT/VAT) volume is associated with risk for cardiometabolic diseases. This work aimed to develop and evaluate automated abdominal SAT/VAT segmentation on longitudinal MRI in adults with overweight/obesity using attention-based competitive dense (ACD) 3D U-Net and 3D nnU-Net with full field-of-view volumetric multi-contrast inputs.

MATERIALS AND METHODS

920 adults with overweight/obesity were scanned twice at multiple 3 T MRI scanners and institutions. The first scan was divided into training/validation/testing sets (n = 646/92/182). The second scan from the subjects in the testing set was used to evaluate the generalizability for longitudinal analysis. Segmentation performance was assessed by measuring Dice scores (DICE-SAT, DICE-VAT), false negatives (FN), and false positives (FP). Volume agreement was assessed using the intraclass correlation coefficient (ICC).

RESULTS

ACD 3D U-Net achieved rapid (< 4.8 s/subject) segmentation with high DICE-SAT (median ≥ 0.994) and DICE-VAT (median ≥ 0.976), small FN (median ≤ 0.7%), and FP (median ≤ 1.1%). 3D nnU-Net yielded rapid (< 2.5 s/subject) segmentation with similar DICE-SAT (median ≥ 0.992), DICE-VAT (median ≥ 0.979), FN (median ≤ 1.1%) and FP (median ≤ 1.2%). Both models yielded excellent agreement in SAT/VAT volume versus reference measurements (ICC > 0.997) in longitudinal analysis.

DISCUSSION

ACD 3D U-Net and 3D nnU-Net can be automated tools to quantify abdominal SAT/VAT volume rapidly, accurately, and longitudinally in adults with overweight/obesity.

analysis combined with sagittal abdominal diameter for abdominal subcutaneous fat measurement

Abdominal subcutaneous fat tissue (ASFT) is an independent predictor of mortality. This prospective observational study aimed to establish a rapid, safe, and convenient estimation equation for abdominal subcutaneous fat area (SFA) using bioimpedance analysis (BIA) combined with sagittal abdominal diameter (SAD). A total of 520 adult subjects were recruited and were randomly divided into 2/3 (n = 346) and 1/3 (n = 174) to form a modeling group (MG) and a validation group (VG), respectively. Each subject's abdomen was scanned using computed tomography to obtain target variables (SFA_CT). Predictor variables for all subjects included bioimpedance index (h²/Z), anthropometric parameters height (h), weight (W), waist circumference (WC), hip circumference (HC), and SAD, along with age and sex (male =1, female = 0). SFA estimation equation SFA_BIA+SAD was established for the MG using stepwise multiple regression analysis. Cross-validation was performed using VG to evaluate the performance of the SFA_BIA+SAD estimation equation. Stepwise multiple regression analysis was applied from the MG, including SFA_BIA+SAD = 49.89 + 1.09 SAD-29.90 Sex + 4.71 W-3.63 h²/Z-1.50 h (r = 0.92, SEE = 28.10 cm², n = 346, p < 0.001). Mean differences in SFA_BIA+SAD relative to SFA_CT were -1.21 ± 21.53, 2.85 ± 27.16, and -0.98 ± 36.6 cm² at different levels of obesity (eutrophic, overweight, obese), respectively. This study did not have a large number of samples in different fields, so it did not have completely external validity. Application of BIA combined with SAD in anthropometric parameters achieves fast, accurate and convenient SAF measurement. Results of this study provide a simple, reliable, and practical measurement that can be widely used in epidemiological studies and in measuring individual SFA.

Generation of quantification maps and weighted images from synthetic magnetic resonance imaging using deep learning network

OBJECTIVE

The generation of quantification maps and weighted images in synthetic MRI techniques is based on complex fitting equations. This process requires longer image generation times. The objective of this study is to evaluate the feasibility of deep learning method for fast reconstruction of synthetic MRI.

APPROACH

A total of 44 healthy subjects were recruited and random divided into a training set (30 subjects) and a testing set (14 subjects). A multiple-dynamic, multiple-echo (MDME) sequence was used to acquire synthetic MRI images. Quantification maps (T1, T2, and proton density (PD) maps) and weighted (T1W, T2W, and T2W FLAIR) images were created with MAGiC software and then used as the ground truth images in the deep learning (DL) model. An improved multichannel U-Net structure network was trained to generate quantification maps and weighted images from raw synthetic MRI imaging data (8 module images). Quantitative evaluation was performed on quantification maps. Quantitative evaluation metrics, as well as qualitative evaluation were used in weighted image evaluation. Nonparametric Wilcoxon signed-rank tests were performed in this study.

MAIN RESULTS

The results of quantitative evaluation show that the error between the generated quantification images and the reference images is small. For weighted images, no significant difference in overall image quality or SNR was identified between DL images and synthetic images. Notably, the DL images achieved improved image contrast with T2W images, and fewer artifacts were present on DL images than synthetic images acquired by T2W FLAIR.

SIGNIFICANCE

The DL algorithm provides a promising method for image generation in synthetic MRI techniques, in which every step of the calculation can be optimized and faster, thereby simplifying the workflow of synthetic MRI techniques.

3D Neural Networks for Visceral and Subcutaneous Adipose Tissue Segmentation using Volumetric Multi-Contrast MRI

Individuals with obesity have larger amounts of visceral (VAT) and subcutaneous adipose tissue (SAT) in their body, increasing the risk for cardiometabolic diseases. The reference standard to quantify SAT and VAT uses manual annotations of magnetic resonance images (MRI), which requires expert knowledge and is time-consuming. Although there have been studies investigating deep learning-based methods for automated SAT and VAT segmentation, the performance for VAT remains suboptimal (Dice scores of 0.43 to 0.89). Previous work had key limitations of not fully considering the multi-contrast information from MRI and the 3D anatomical context, which are critical for addressing the complex spatially varying structure of VAT. An additional challenge is the imbalance between the number and distribution of pixels representing SAT/VAT. This work proposes a network based on 3D U-Net that utilizes the full field-of-view volumetric T₁-weighted, water, and fat images from dual-echo Dixon MRI as the multi-channel input to automatically segment SAT and VAT in adults with overweight/obesity. In addition, this work extends the 3D U-Net to a new Attention-based Competitive Dense 3D U-Net (ACD 3D U-Net) trained with a class frequency-balancing Dice loss (FBDL). In an initial testing dataset, the proposed 3D U-Net and ACD 3D U-Net with FBDL achieved 3D Dice scores of (mean ± standard deviation) 0.99 ±0.01 and 0.99±0.01 for SAT, and 0.95±0.04 and 0.96 ±0.04 for VAT, respectively, compared to manual annotations. The proposed 3D networks had rapid inference time (<60 ms/slice) and can enable automated segmentation of SAT and VAT.Clinical relevance- This work developed 3D neural networks to automatically, accurately, and rapidly segment visceral and subcutaneous adipose tissue on MRI, which can help to characterize the risk for cardiometabolic diseases such as diabetes, elevated glucose levels, and hypertension.

Assessing visceral and subcutaneous adiposity using segmented T2-MRI and multi-frequency segmental bioelectrical impedance: A sex-based comparative study

BACKGROUND AND AIM

This study aims to quantify abdominal visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) using T2-weighted magnetic resonance imaging (MRI), and assess the extent of its concordance with VAT surface-area measured by a state-of-the-art segmental multi-frequency bioelectrical impedance analysis (BIA) device. A comparison between manual and semi-automated segmentation was conducted. Further, abdominal VAT and SAT sex-based comparison in healthy Arab adults was piloted.

METHODS

A cross-sectional design was followed to recruit subjects. Abdominal VAT and SAT were determined on T2-weighted MRI manually and semi-automatically. Body composition was assessed using a BIA machine. Statistical differences between the abdominal VAT areas defined by BIA, manual, and semi-automated MRI were compared. Correlation between all methods was assessed, and statistical differences between sex abdominal VAT/SAT defined areas were compared.

RESULTS

A total of 165 abdominal T2-weighted MR images taken for 55 overweight/obese adult subjects were analyzed Differences between manual and semi-automated MRI-obtained abdominal VAT and SAT were found statistically significant (P<0.001) for all subjects. Mean abdominal VAT using the BIA technique was found to correlate significantly with manually and semi-automated T2-weighted MRI defined VAT (r=0.7436; P<0.001 and r=0.8275; P<0.001, respectively). Abdominal VAT was significantly (P<0.001) different between male and female subjects accumulating at different abdominal levels.

CONCLUSION

Semi-automatic segmentation showed a stronger significant correlation with BIA compared to manual segmentation, implying a more reliable quantification of abdominal VAT/SAT. Segmental BIA technique may serve as a feasible and convenient assessment tool for the visceral adiposity in obese subjects.

Abdominal adipose tissue components quantification in MRI as a relevant biomarker of metabolic profile

INTRODUCTION

Abnormal accumulation of adipose tissue (AT) alters the metabolic profile and underlies cardiovascular complications. Conventional measures provide global measurements for the entire body. The purpose of this study was to propose a new approach to quantify the amount and type of truncal AT automatically from MRI in metabolic patients and controls.

MATERIALS AND METHODS

DIXON acquisitions were performed at 1.5 T in 30 metabolic syndrome (MS) (59 ± 6 years), 12 obese (50 ± 11 years), 35 type 2 diabetes (T2DM) patients (56 ± 11 years) and 19 controls (52 ± 11 years). AT was segmented into: subcutaneous AT "SAT", visceral AT "VAT", deep VAT "dVAT", peri-organ VAT "pVAT" using active contours and k-means clustering algorithms. Subsequently, organ AT infiltration index "oVAT" was calculated as the normalized fat signal magnitude in organs.

RESULTS

Excellent intra- and inter-operator reproducibility was obtained for AT segmentation. MS and obese patients had the highest amount of total AT. SAT increased in MS (1144 ± 621 g) and T2DM patients (1024 ± 634 g), and twice the level of SAT in controls (505 ± 238 g), and further increased in obese patients (1429 ± 621 g). While VAT, pVAT and dVAT increased to a similar degree in the metabolic patients compared to controls, the oVAT index was able to differentiate controls from MS and T2DM patients and to discriminate the three metabolic patient groups (p < 0.01). Local AT sub-types were not related to BMI in all groups except for SAT in controls (p = 0.03).

CONCLUSION

Reproducible truncal AT sub-types quantification using 3D MRI was able to characterize patients with metabolic diseases. It may serve in the future as a non-invasive predictor of cardiovascular complications in such patients.

Toward a Routine Assessment of Visceral Adipose Tissue Volume from Computed Tomographic Data

OBJECTIVE

The study's aim was to determine to what extent total visceral adipose tissue (VAT) volume (V_VAT-T ) measured from segmented VAT areas (A_VAT ) on all axial computed tomography (CT) sections (thickness of 5 mm) between the diaphragm and pelvic floor can be predicted by the A_VAT of commonly selected landmark sections in patients with overweight or obesity.

METHODS

A total of 113 patients (31 females, 82 males) with images of full abdominopelvic coverage and proper image quality were included (BMI = 25.0-64.1 kg/m² , 29.5 ± 4.9 kg/m² ). Linear regression between A_VAT and V_VAT-T (reference) was used to determine approximate equations for VAT volume for all parameters (single sex, different anatomical landmarks or lumbar intervertebral disc spaces, one or five axial sections). Agreement was evaluated by the multivariate coefficient of determination and by the SD of the percentage difference (s_d% ) between the estimated VAT volume on one or five sections and V_VAT-T .

RESULTS

The V_VAT-T was 0.9 to 8.4 (3.8 ± 2.2) L for females and 2.7 to 11.7 (5.6 ± 2.1) L for males. Best agreement was found at L2-3 (s_d%  = 14.3%-15.5%) for females and at L1-2 or L2-3 (11.7%-12.4%) for males. Agreement at the umbilicus or the femoral heads was poor (20.2%-57.9%). Segmentation of one or five sections was substantially faster (11/70 seconds) than whole-abdomen processing (15 minutes).

CONCLUSIONS

V_VAT-T can be rapidly estimated by VAT segmentation of axial CT sections at sex-specific lumbar intervertebral disc spaces.

The relationship between pancreas steatosis and the risk of metabolic syndrome and insulin resistance in Chinese adolescents with concurrent obesity and non-alcoholic fatty liver disease

BACKGROUND

The incidence of childhood obesity and associated comorbidities are on an increasing trend worldwide. More than 340 million children and adolescents aged between 5 and 19 years old were overweight or had obesity in 2016, from which over 124 million children and adolescents (6% of girls and 8% of boys) had obesity.

OBJECTIVE

To describe the relationship between pancreas steatosis, body fat and the risk of metabolic syndrome, insulin resistance in Hong Kong Chinese adolescents with both obesity and non-alcoholic fatty liver disease (NAFLD).

METHODS

Fifty two adolescents with obesity and NAFLD were analysed (14-18 years), stratified into fatty and non-fatty pancreas groups using chemical shift encoded MRI-pancreas proton density fat fraction ≥5%. Pancreatic, abdominal subcutaneous adipose tissue (SAT)/visceral adipose tissue (VAT) volumes, biochemical and anthropometric parameters were measured. Mann-Whitney U test, multiple linear/binary logistic regression analyses and odds ratios were used.

RESULTS

Fifty percent had fatty pancreas, 38% had metabolic syndrome and 81% had insulin resistance. Liver proton density fat fraction (PDFF) and VAT were independent predictors of insulin resistance (P = .006, .016). Pancreas and liver PDFF were both independent predictors of beta cells dysfunction (P = .015, .050) and metabolic syndrome (P = .021, .041). Presence of fatty pancreas in obesity was associated with insulin resistance (OR = 1.58, 95% CI = 0.39-6.4) and metabolic syndrome (OR = 1.70, 95% CI = 0.53-5.5).

CONCLUSION

A significant causal relationship exists between fatty pancreas, fatty liver, body fat and the risk of developing metabolic syndrome and insulin resistance.

KEY POINTS

Fatty pancreas is a common finding in adolescents with obesity, with a prevalence rate of 50% in this study cohort. Liver PDFF and VAT are independent predictors of insulin resistance while pancreas PDFF and liver PDFF are independent predictors of both beta cells dysfunction and metabolic syndrome. Presence of fatty pancreas at imaging should not be considered as a benign finding but rather as an imaging biomarker of emerging pancreatic metabolic and endocrine dysfunction.

Assessment of body composition in pediatric overweight and obesity: A systematic review of the reliability and validity of common techniques

Accurate measurement of body composition is required to improve health outcomes in children and adolescents with overweight or obesity. This systematic review aimed to summarize the reliability and validity of field and laboratory body composition techniques employed in pediatric obesity studies to facilitate technique selection for research and clinical practice implementation. A systematic search in MEDLINE (via PubMed), EMBASE, CINAHL, and SPORTDiscus from inception up to December 2019 was conducted, using a combination of the following concepts: body composition, pediatric overweight/obesity, and reliability/validity. The search strategy resulted in 66 eligible articles reporting reliability (19.7%), agreement between body composition techniques cross sectionally (80.3%), and/or diagnostic test accuracy (10.6%) in children and adolescents with overweight or obesity (mean age range = 7.0-16.5 years). Skinfolds, air-displacement plethysmography (ADP), dual-energy X-ray absorptiometry (DXA), and ultrasound presented as reliable techniques. DXA, ADP, and isotope dilution showed similar and the best agreement with reference standards. Compared with these laboratory techniques, the validity of estimating body composition by anthropometric equations, skinfolds, and BIA was inferior. In conclusion, the assessment of body composition by laboratory techniques cannot be replaced by field techniques due to introduction of measurement errors, which potentially conceal actual changes in body components.

Automated and accurate quantification of subcutaneous and visceral adipose tissue from magnetic resonance imaging based on machine learning

Accurate measuring of subcutaneous adipose tissue (SAT) and visceral adipose tissue (VAT) is vital for the research of many diseases. The localization and quantification of SAT and VAT by computed tomography (CT) expose patients to harmful ionizing radiation. Magnetic resonance imaging (MRI) is a safe and painless test. The aim of this paper is to explore a practical method for the segmentation of SAT and VAT based on the iterative decomposition of water and fat with echo asymmetry and least square estimation‑iron quantification (IDEAL-IQ) technology and machine learning. The approach involves two main steps. First, a deep network is designed to segment the inner and outer boundaries of SAT in fat images and the peritoneal cavity contour in water images. Second, after mapping the peritoneal cavity contour onto the fat images, the assumption-free K-means++ with a Markov chain Monte Carlo (AFK-MC²) clustering method is used to obtain the VAT content. An MRI data set from 75 subjects is utilized to construct and evaluate the new strategy. The Dice coefficients for the SAT and VAT content obtained from the proposed method and the manual measurements performed by experts are 0.96 and 0.97, respectively. The experimental results indicate that the proposed method and the manual measurements exhibit high reliability.

MRI adipose tissue and muscle composition analysis-a review of automation techniques

MRI is becoming more frequently used in studies involving measurements of adipose tissue and volume and composition of skeletal muscles. The large amount of data generated by MRI calls for automated analysis methods. This review article presents a summary of automated and semi-automated techniques published between 2013 and 2017. Technical aspects and clinical applications for MRI-based adipose tissue and muscle composition analysis are discussed based on recently published studies. The conclusion is that very few clinical studies have used highly automated analysis methods, despite the rapidly increasing use of MRI for body composition analysis. Possible reasons for this are that the availability of highly automated methods has been limited for non-imaging experts, and also that there is a limited number of studies investigating the reproducibility of automated methods for MRI-based body composition analysis.

How to best assess abdominal obesity

PURPOSE OF REVIEW

Abdominal obesity, especially the increase of visceral adipose tissue (VAT), is closely associated with increased mortality related to cardiovascular disease, diabetes, and fatty liver disease. This review provides an overview of the recent advances for abdominal obesity measurement.

RECENT FINDINGS

Compared to simple waist circumference, emerging three-dimensional (3D) body-scanning techniques also measure abdominal volume and shape. Abdominal dimension measures have been implemented in bioelectrical impedance analysis to improve accuracy when estimating VAT. Geometrical models have been applied in ultrasound to convert depth measurement into VAT area. Only computed tomography (CT) and MRI can provide direct measures of VAT. Recent advances in imaging allow for evaluating functional aspects of abdominal fat such as brown adipose tissue and fatty acid composition.

SUMMARY

Waist circumference is a simple, inexpensive method to measure abdominal obesity. CT and MRI are reference methods for measuring VAT. Further studies are needed to establish the accuracy for dual-energy X-ray absorptiometry in estimating longitudinal changes of VAT. Further studies are needed to establish whether bioelectrical impedance analysis, ultrasound, or 3D body scanning is consistently superior to waist circumference in estimating VAT in different populations.

Observed changes in brown, white, hepatic and pancreatic fat after bariatric surgery: Evaluation with MRI

OBJECTIVES

To study the change in brown and white adipose tissue (BAT and WAT), as well as fat content in the liver and pancreas, in patients with morbid obesity before and after bariatric surgery.

METHODS

Twelve patients with morbid obesity (F=8, M=4, age: 45.4 years (38.4-51.2), BMI: 35.2 kg/m² (32.5-38.6)) underwent pre-op MRI at baseline and two post-op scans at 6-month and 12-month intervals after bariatric surgery. Co-registered water, fat, fat-fraction and T2* image series were acquired. Supraclavicular BAT and abdominal WAT were measured using in-house algorithms. Intrahepatic triglyceride (IHTG) was measured using MR spectroscopy and pancreatic fat was measured using a region-of-interest approach. Fat contents were compared between baseline and the first and second 6-month intervals using non-parametric analysis of Friedman's test and Wilcoxon's signed-rank test. Level of significance was selected at p=0.017 (0.05/3). Threshold of non-alcoholic fatty liver disease was set at 5.56%.

RESULTS

Results indicated that BMI (p=0.005), IHTG (p=0.005), and subcutaneous (p=0.005) and visceral adipose tissues (p=0.005) were significantly reduced 6 months after surgery. Pancreatic fat (p=0.009) was significantly reduced at 12 months. Most reduction became stable between the 6-month and 12-month interval. No significant difference was observed in BAT volume, fat-fraction and T2* values.

CONCLUSION

The results of this study suggest that bariatric surgery effectively reduced weight, mainly as a result of the reduction of abdominal WAT. Liver and pancreatic fat were deceased below the threshold possibly due to the reduction of free fatty acid. BAT volume, fat-fraction and T2* showed no significant changes, probably because surgery itself might not have altered the metabolic profile of the patients.

KEY POINTS

• No significant changes were observed in fat-fraction, T2* and volume of brown adipose tissue after bariatric surgery. • Non-alcoholic fatty liver disease was resolved after surgery. • Abdominal white fat and liver fat were significantly reduced 6 months after surgery and become stable between 6 and 12 months while pancreatic fat was significantly reduced between 0 and 12 months.