Oleksandr Shlakhter

Placental Morphologic Features and Chorionic Surface Vasculature at Term Are Highly Correlated With 3-Dimensional Sonographic Measurements at 11 to 14 Weeks

The purpose of this study was to examine the potential for 3‐dimensional sonographic measurement of the early placenta in predicting ultimate placental morphologic features at delivery.

Acceleration Operators in the Value Iteration Algorithms for Markov Decision Processes

We study the general approach to accelerating the convergence of the most widely used solution method of Markov decision processes (MDPs) with the total expected discounted reward. Inspired by the monotone behavior of the contraction mappings in the feasible set of the linear programming problem equivalent to the MDP, we establish a class of operators that can be used in combination with a contraction mapping operator in the standard value iteration algorithm and its variants. We then propose two such operators, which can be easily implemented as part of the value iteration algorithm and its variants. Numerical studies show that the computational savings can be significant especially when the discount factor approaches one and the transition probability matrix becomes dense, in which the standard value iteration algorithm and its variants suffer from slow convergence.

Probability distributions of placental morphological measurements and origins of variability of placental shapes

INTRODUCTION While the mean shape of human placenta is round with centrally inserted umbilical cord, significant deviations from this ideal are fairly common, and may be clinically meaningful. Traditionally, they are explained by trophotropism. We have proposed a hypothesis explaining typical variations in placental shape by randomly determined fluctuations in the growth process of the vascular tree. It has been recently reported that umbilical cord displacement in a birth cohort has a log-normal probability distribution, which indicates that the displacement between an initial point of origin and the centroid of the mature shape is a result of accumulation of random fluctuations of the dynamic growth of the placenta. To confirm this, we investigate statistical distributions of other features of placental morphology. METHODS In a cohort of 1023 births at term digital photographs of placentas were recorded at delivery. Excluding cases with velamentous cord insertion, or missing clinical data left 1001 (97.8%) for which placental surface morphology features were measured. Best-fit statistical distributions for them were obtained using EasyFit. RESULTS AND DISCUSSION The best-fit distributions of umbilical cord displacement, placental disk diameter, area, perimeter, and maximal radius calculated from the cord insertion point are of heavy-tailed type, similar in shape to log-normal distributions. This is consistent with a stochastic origin of deviations of placental shape from normal. CONCLUSIONS Deviations of placental shape descriptors from average have heavy-tailed distributions similar in shape to log-normal. This evidence points away from trophotropism, and towards a spontaneous stochastic evolution of the variants of placental surface shape features.

Variable placental thickness affects placental functional efficiency independent of other placental shape abnormalities

Our previous work suggests that stressors that impact placental vascular growth result in a deformed chorionic surface shape, which reflects an abnormal placental three-dimensional shape. We propose to use variability of placental disk thickness as a reflector of deviations in placental vascular growth at the finer level of the fetal stems. We hypothesize that increased variability of thickness is associated with abnormal chorionic surface shape, but will be a predictor of reduced placental functional efficiency (smaller baby for a given placental weight) independent of shape. These measures may shed light on the mechanisms linking placental growth to risk of adult disease. The sample was drawn from the Pregnancy, Infection and Nutrition Study. In all, 94.6% of the cohort consented to placental examination. Of the 1023 delivered at term, those previously sectioned by the Pathology Department were excluded, leaving 587 (57%) cases with intact placentas that were sliced and photographed. The chorionic surface shape and the shape of a central randomly oriented placental slice were analyzed and measures were compared using correlation. Lower mean placental disk thickness and more variable disk thickness were each strongly and significantly correlated with deformed chorionic plate shapes. More variable disk thickness was strongly correlated with reduced placental efficiency independent of abnormal chorionic surface shape. Variability of placental disk thickness, simple to measure in a single randomly oriented central slice, may be an easily acquired measure that is an independent indicator of lowered placental efficiency, which may in turn program the infant and result in increased risk for development of adult diseases.

Accelerated modified policy iteration algorithms for Markov decision processes

We propose a new approach to accelerate the convergence of the modified policy iteration method for Markov decision processes with the total expected discounted reward. In the new policy iteration an additional operator is applied to the iterate generated by Markov operator, resulting in a bigger improvement in each iteration.

Virtual depot approximation for the transshipment problem

In the transshipment problem a number of retailers facing stochastic demand must place orders before the demand is known, but can transship inventory once the demand is realized. Recently developed simulation-based algorithm [1] provides near-optimal solutions, but can only handle small-to-medium problems. We develop an approximation-based approach where all transshipments are routed through a virtual depot. This allows us to reduce the solution time by orders of magnitude, while maintaining high solution accuracy, making it possible to solve realistic-size problems to optimality.