The successful management of type-1 diabetes (T1D) and insulin-dependent type-2 diabetes depends on the ability to accurately calibrate bolus and basal insulin doses, and minimize the time spent with high postprandial blood glucose (BG) levels while avoiding dangerously low hypoglycemic excursions. Precise calibration of the insulin pump helps with successful disease management; however, people with T1D may still experience prolonged, potentially damaging BG levels due to postprandial highs. Insulin pumps offer large, currently underexploited degrees of freedom in insulin delivery, which can have a dramatic impact on average postprandial BG levels. Building on existing models of the glucose-insulin system, we first propose a simple, automated, and individualized insulin-pump calibration system based on a series of measures taken before and a few hours after carbohydrate ingestion. We then modulate the shape of pre-meal insulin dosing to explicitly reduce postprandial BG levels while minimizing the likelihood of dangerously low BG. Such an optimal insulin delivery time course can potentially improve postprandial BG levels and rapidly bring BG to the target level. We evaluated our methods in 20 patients over 4 days, with BG levels sampled frequently.
I. Aganj, N. Bryant, Leah (Morgan) Panaro et al.· medRxiv· 0 citations
Advances in image registration and machine learning have recently enabled volumetric analysis of postmortem brain tissue from conventional photographs of coronal slabs, which are routinely collected in brain banks and neuropathology laboratories around the world. One caveat of this methodology is the requirement of segmentation of the tissue from the background and out-of-slice tissue in photographs, which currently requires laborious manual intervention. Manual delineation is a bottleneck in this process and poses challenges in scalability, and resources, restricting adoption of these methods. In this article, we present a deep learning model to automate this process. The automatic segmentation tool relies on a U-Net architecture that was trained with a combination of 1,414 manually segmented images of both fixed and fresh tissue, from specimens with varying diagnoses, photographed at two different sites. Automated model predictions on a subset of photographs not seen in training were analyzed to estimate performance compared to manual labels, including both inter- and intra-rater variability. Our model achieved a median Dice score over 0.98, mean surface distance under 0.4 mm, and 95% Hausdorff distance under 1.60 mm, which approaches inter-/intra-rater levels. Our tool is publicly available at surfer.nmr.mgh.harvard.edu/fswiki/PhotoTools and training data is available at https://zenodo.org/records/20647553.
Jonathan Williams Ramirez, Dina Zemlyanker, Lucas J. Deden-Binder et al.· PLoS ONE· 0 citations
