Short-Term Variability of Proton Density Fat Fraction in Pancreas and Liver Assessed by Multiecho Chemical-Shift Encoding-Based MRI at 3 T.

BACKGROUND: Quantification of pancreatic fat (PF) and intrahepatic lipids (IHL) is of increasing interest in subjects at risk for metabolic diseases. There is limited data available on short- and medium-term variability of PF/IHL and on their dependence on nutritional status. PURPOSE: To assess short-term intraday variations of PF/IHL after a high-fat meal as well as medium-term changes after 5 days of high-caloric diet. STUDY TYPE: Prospective cohort study. SUBJECTS: A total of 12 subjects (six males) for intraday variations study, 15 male subjects for medium-term high-caloric diet study and 11 age- and body mass index (BMI)-matched controls. FIELD STRENGTH/SEQUENCE: A 3 T; chemical-shift encoded multiecho gradient echo sequence. ASSESSMENT: For the intraday study, subjects were scanned after overnight fasting and after a high fat meal on the same day. For the medium-term study, 26 subjects were scanned after overnight fasting with 15/11 rescanned after 5 days of high-calorie diet/isocaloric diet. Proton density fat fraction (PDFF) maps were generated inline on the scanner. Regions of interest were manually drawn in head, body, and tail of pancreas and in the liver by a medical physicist and a doctoral student (26/4 years of experience). PF was calculated as the average of the head, body, and tail measurements. STATISTICAL TESTS: Repeated measurements ANOVA for assessing changes in PF/IHL, linear correlation analyses for assessing relationships of PF/IHL with BMI. Significance level P < 0.05 for all. RESULTS: Nonsignificant changes in PF (2.6 ± 1.0 vs. 2.7 ± 0.9% after high-fat meal, 1.4 ± 0.8 vs. 1.5 ± 0.6% [high-caloric diet] and 1.5 ± 0.8 vs. 1.8 ± 1.0% [isocaloric control group]), nonsignificant changes in IHL after high-fat meal (2.6 ± 1.3 vs. 2.5 ± 0.9%) and in the control group (1.1 ± 0.6 vs. 1.2 ± 1.1%), significantly increased IHL after high-caloric diet (1.7 ± 2.2% vs. 2.7 ± 3.6%). Nonsignificant changes in PF (2.6 ± 1.0 vs. 2.7 ± 0.9% after high-fat meal, 1.4 ± 0.8 vs. 1.5 ± 0.6% [high-caloric diet] and 1.5 ± 0.8 vs. 1.8 ± 1.0% [isocaloric control group]), nonsignificant changes in IHL after high-fat meal (2.6 ± 1.3 vs. 2.5 ± 0.9%) and in the control group (1.1 ± 0.6 vs. 1.2 ± 1.1%), significantly increased IHL after 5-days of high-caloric diet (1.7 ± 2.2% vs. 2.7 ± 3.6%). DATA CONCLUSION: Time of day and nutritional status have no significant influence on PF/IHL and are therefore not likely to be major confounders in epidemiologic or clinical studies. EVIDENCE LEVEL: 2 TECHNICAL EFFICACY: Stage 1.

Detection of diabetes from whole-body MRI using deep learning.

Obesity is one of the main drivers of type 2 diabetes, but it is not uniformly associated with the disease. The location of fat accumulation is critical for metabolic health. Specific patterns of body fat distribution, such as visceral fat, are closely related to insulin resistance. There might be further, hitherto unknown, features of body fat distribution that could additionally contribute to the disease. We used machine learning with dense convolutional neural networks to detect diabetes-related variables from 2371 T1-weighted whole-body MRI data sets. MRI was performed in participants undergoing metabolic screening with oral glucose tolerance tests. Models were trained for sex, age, BMI, insulin sensitivity, HbA1c, and prediabetes or incident diabetes. The results were compared with those of conventional models. The area under the receiver operating characteristic curve was 87% for the type 2 diabetes discrimination and 68% for prediabetes, both superior to conventional models. Mean absolute regression errors were comparable to those of conventional models. Heatmaps showed that lower visceral abdominal regions were critical in diabetes classification. Subphenotyping revealed a group with high future diabetes and microalbuminuria risk.Our results show that diabetes is detectable from whole-body MRI without additional data. Our technique of heatmap visualization identifies plausible anatomical regions and highlights the leading role of fat accumulation in the lower abdomen in diabetes pathogenesis.

Overexpression of Id-1 is associated with poor clinical outcome in node negative breast cancer.

Id-1 is an important regulator of cellular growth and differentiation and controls malignant progression of breast cancer cells. The aim of our study was to assess the clinical impact of Id-1 expression in breast cancer, i.e., its potential impact on prognosis and prediction of treatment response. Id-1 protein expression was determined immunohistochemically in 191 patients with lymph-node negative breast cancer, and univariate and multivariate survival analysis was carried out. Fifteen (7.9%) specimens showed strong expression, 75 (39.3%) moderate, 55 (28.8%) weak expression and 46 (24.1%) cases no expression of Id-1. Patients with strong or moderate Id-1 expression had a significant shorter overall (p = 0.003, Cox regression) and disease-free survival (p = 0.01, Cox regression) compared to those with absent or low expression. Progesterone receptor density was significantly higher in breast cancers with absent/low Id-1 expression compared to those with moderate/strong expression (p < 0.001, t-test). Id-1 expression was significantly stronger in cases positive for p16(INK4a) expression compared to those negative for p16 (p = 0.049, Mann-Whitney test). The influence of Id-1 on clinical outcome seems much stronger in patients with negative estrogen receptor status compared to those with positive status, who received receptor antagonists as adjuvant therapy in most cases. Overexpression of Id-1 protein represents a strong independent prognostic marker in node negative breast cancer, and future therapies inhibiting Id-1 expression might be beneficial for these patients. Our results also suggest that due to the apparent interaction of Id-1 with the steroid-receptor system in breast cancer, hormonal therapies might influence Id-1 expression and its impact on clinical outcome.