Learning local depth regression from defocus blur by soft-assignment encoding.

We present a novel, to the best of our knowledge, patch-based approach for depth regression from defocus blur. Most state-of-the-art methods for depth from defocus (DFD) use a patch classification approach among a set of potential defocus blurs related to a depth, which induces errors due to the continuous variation of the depth. Here, we propose to adapt a simple classification model using a soft-assignment encoding of the true depth into a membership probability vector during training and a regression scale to predict intermediate depth values. Our method uses no blur model or scene model; it only requires a training dataset of image patches (either raw, gray scale, or RGB) and their corresponding depth label. We show that our method outperforms both classification and direct regression on simulated images from structured or natural texture datasets, and on raw real data having optical aberrations from an active DFD experiment.

An Unsafe/Safe Typology in People with Type 2 Diabetes: Bridging Patients' Expectations, Personality Traits, Medication Adherence, and Clinical Outcomes.

Background: Support programs are provided to people with diabetes to help them manage their disease. However, adherence to and persistence in support programs are often low, making it difficult to demonstrate their effectiveness. Aim: To identify the determinants of patients' perceived interest in diabetes support programs because it may be a powerful determinant of effective participation in such programs. Patients and Methods: An online study conducted in April 2021 in metropolitan France on 600 people with diabetes recruited from a consumer panel. A 64-item psychosocial questionnaire including a question asking to evaluate the helpfulness of a support program was used. Univariate, multivariate, and multiple correspondence analyses were performed. Results: The existence of a typology, known as Unsafe/Safe, was discovered, in which patients with type 2 diabetes respond in two distinct ways. Type U (unsafe) patients, who believe that a support program would be helpful, are more likely to be nonadherent to their treatment, have high hemoglobin A1c levels, have at least one diabetic complication, lack information regarding their disease and treatment, rate the burden of their disease and impairment of their quality of life as high, worry about their future, and are pessimistic. Type S (safe) patients have the opposite characteristics. Type U patients can be dichotomized into two broad classes: one in which they lack information regarding disease and treatment and the other in which alterations in the quality of life and burden of the disease predominate. Insulin-treated patients give more importance to the lack of information, whereas noninsulin-treated patients complain primarily about the burden of the disease and impairment of quality of life. Conclusion: This study describes this new U/S typology, proposes a simple method based on a nine-item questionnaire to identify type U patients by calculating a Program Helpfulness Score described herein, and clarifies the nature of the intervention to be provided to them. This novel approach could be applied to other chronic diseases.

Factors influencing the levothyroxine dose in the hormone replacement therapy of primary hypothyroidism in adults.

Levothyroxine (LT4) is a safe, effective means of hormone replacement therapy for hypothyroidism. Here, we review the pharmaceutical, pathophysiological and behavioural factors influencing the absorption, distribution, metabolism and excretion of LT4. Any factor that alters the state of the epithelium in the stomach or small intestine will reduce and/or slow absorption of LT4; these include ulcerative colitis, coeliac disease, bariatric surgery, Helicobacter pylori infection, food intolerance, gastritis, mineral supplements, dietary fibre, resins, and various drugs. Once in the circulation, LT4 is almost fully bound to plasma proteins. Although free T4 (FT4) and liothyronine concentrations are extensively buffered, it is possible that drug- or disorder-induced changes in plasma proteins levels can modify free hormone levels. The data on the clinical significance of genetic variants in deiodinase genes are contradictory, and wide-scale genotyping of hypothyroid patients is not currently justified. We developed a decision tree for the physician faced with an abnormally high thyroid-stimulating hormone (TSH) level in a patient reporting adequate compliance with the recommended LT4 dose. The physician should review medications, the medical history and the serum FT4 level and check for acute adrenal insufficiency, heterophilic anti-TSH antibodies, antibodies against gastric and intestinal components (gastric parietal cells, endomysium, and tissue transglutaminase 2), and Helicobacter pylori infection. The next step is an LT4 pharmacodynamic absorption test; poor LT4 absorption should prompt a consultation with a gastroenterologist and (depending on the findings) an increase in the LT4 dose level. An in-depth etiological investigation can reveal visceral disorders and, especially, digestive tract disorders.