Artificial Intelligence Algorithms for Benign vs. Malignant Dermoscopic Skin Lesion Image Classification.

In recent decades, the incidence of melanoma has grown rapidly. Hence, early diagnosis is crucial to improving clinical outcomes. Here, we propose and compare a classical image analysis-based machine learning method with a deep learning one to automatically classify benign vs. malignant dermoscopic skin lesion images. The same dataset of 25,122 publicly available dermoscopic images was used to train both models, while a disjointed test set of 200 images was used for the evaluation phase. The training dataset was randomly divided into 10 datasets of 19,932 images to obtain an equal distribution between the two classes. By testing both models on the disjoint set, the deep learning-based method returned accuracy of 85.4 ± 3.2% and specificity of 75.5 ± 7.6%, while the machine learning one showed accuracy and specificity of 73.8 ± 1.1% and 44.5 ± 4.7%, respectively. Although both approaches performed well in the validation phase, the convolutional neural network outperformed the ensemble boosted tree classifier on the disjoint test set, showing better generalization ability. The integration of new melanoma detection algorithms with digital dermoscopic devices could enable a faster screening of the population, improve patient management, and achieve better survival rates.

Leptin resistance before and after obesity: evidence that tissue glucose uptake underlies adipocyte enlargement and liver steatosis/steatohepatitis in Zucker rats from early-life stages.

BACKGROUND: Leptin resistance occurs in obese patients, but its independent contribution to adiposity and the accompanying metabolic diseases, i.e., diabetes, liver steatosis, and steatohepatitis, remains to be established. This study was conducted in an extreme model of leptin resistance to investigate mechanisms initiating diabetes, fat expansion, liver steatosis, and inflammatory disease, focusing on the involvement of glucose intolerance and organ-specific glucose uptake in brown and subcutaneous adipose tissues (BAT, SAT) and in the liver. METHODS: We studied preobese and adult Zucker rats (fa/fa, fa/+ ) during fasting or glucose loading to assess glucose tolerance. Relevant pancreatic and intestinal hormonal levels were measured by Milliplex. Imaging of 18F-fluorodeoxyglucose by positron emission tomography was used to quantify glucose uptake in SAT, BAT, and liver, and evaluate its relationship with adipocyte size and biopsy-proven nonalcoholic fatty liver disease (NAFLD) and steatohepatitis (NASH). RESULTS: Preobese fa/fa pups showed impaired glucose tolerance, adipocyte enlargement, hepatic microsteatosis, and lobular inflammation, with elevated hepatic post-glucose load glucose uptake and production. Adult fa/fa rats had more severe glucose intolerance, fasting hyperglycemia, hormonal abnormalities, elevated glucose uptake in SAT and BAT, and more markedly in the liver, together with macrosteatosis, and highly prevalent hepatic inflammation. Organ glucose uptake was proportional to the degree of fat accumulation and tissue inflammation and was able to dissect healthy from NAFLD and NAFLD/NASH livers. Most severe NASH livers showed a decline in glucose uptake and liver enzymes. CONCLUSIONS: In fa/fa Zucker rats, leptin resistance leads to glucose intolerance, mainly due to hepatic glucose overproduction, preceding obesity, and explaining pancreatic and intestinal hormonal changes and fat accumulation in adipocytes and hepatocytes. Our data support the involvement of liver glucose uptake in the pathogenesis of liver inflammatory disease. Its potential as more generalized biomarker or diagnostic approach remains to be established outside of our leptin-receptor-deficient rat model.

Maternal High-Fat Feeding Affects the Liver and Thymus Metabolic Axis in the Offspring and Some Effects Are Attenuated by Maternal Diet Normalization in a Minipig Model.

Maternal high-fat diet (HFD) affects metabolic and immune development. We aimed to characterize the effects of maternal HFD, and the subsequent diet-normalization of the mothers during a second pregnancy, on the liver and thymus metabolism in their offspring, in minipigs. Offspring born to high-fat (HFD) and normal diet (ND) fed mothers were studied at week 1 and months 1, 6, 12 of life. Liver and thymus glucose uptake (GU) was measured with positron emission tomography during hyperinsulinemic-isoglycemia. Histological analyses were performed to quantify liver steatosis, inflammation, and hepatic hematopoietic niches (HHN), and thymocyte size and density in a subset. The protocol was repeated after maternal-diet-normalization in the HFD group. At one week, HFDoff were characterized by hyperglycemia, hyperinsulinemia, severe insulin resistance (IR), and high liver and thymus GU, associating with thymocyte size and density, with elevated weight-gain, liver IR, and steatosis in the first 6 months of life. Maternal diet normalization reversed thymus and liver hypermetabolism, and increased HHN at one week. It also normalized systemic insulin-sensitivity and liver fat content at all ages. Instead, weight-gain excess, hyperglycemia, and hepatic IR were still observed at 1 month, i.e., end-lactation. We conclude that intra-uterine HFD exposure leads to time-changing metabolic and immune-correlated abnormalities. Maternal diet-normalization reversed most of the effects in the offspring.

The Role of Prep1 in the Regulation of Mesenchymal Stromal Cells.

Molecular mechanisms governing cell fate decision events in bone marrow mesenchymal stromal cells (MSC) are still poorly understood. Herein, we investigated the homeobox gene Prep1 as a candidate regulatory molecule, by adopting Prep1 hypomorphic mice as a model to investigate the effects of Prep1 downregulation, using in vitro and in vivo assays, including the innovative single cell RNA sequencing technology. Taken together, our findings indicate that low levels of Prep1 are associated to enhanced adipogenesis and a concomitant reduced osteogenesis in the bone marrow, suggesting Prep1 as a potential regulator of the adipo-osteogenic differentiation of mesenchymal stromal cells. Furthermore, our data suggest that in vivo decreased Prep1 gene dosage favors a pro-adipogenic phenotype and induces a "browning" effect in all fat tissues.