scholarly journals Super-Learning of an Optimal Dynamic Treatment Rule

2016 ◽  
Vol 12 (1) ◽  
pp. 305-332 ◽  
Author(s):  
Alexander R. Luedtke ◽  
Mark J. van der Laan
Keyword(s):  
Cross Validation ◽  
Convex Combination ◽  
A Priori ◽  
Sequential Loss ◽  
Adaptive Estimators ◽  
Dynamic Treatment ◽  
Optimal Dynamic ◽  
The Mean ◽  
Data Adaptive ◽  
Time Point

Abstract We consider the estimation of an optimal dynamic two time-point treatment rule defined as the rule that maximizes the mean outcome under the dynamic treatment, where the candidate rules are restricted to depend only on a user-supplied subset of the baseline and intermediate covariates. This estimation problem is addressed in a statistical model for the data distribution that is nonparametric, beyond possible knowledge about the treatment and censoring mechanisms. We propose data adaptive estimators of this optimal dynamic regime which are defined by sequential loss-based learning under both the blip function and weighted classification frameworks. Rather than a priori selecting an estimation framework and algorithm, we propose combining estimators from both frameworks using a super-learning based cross-validation selector that seeks to minimize an appropriate cross-validated risk. The resulting selector is guaranteed to asymptotically perform as well as the best convex combination of candidate algorithms in terms of loss-based dissimilarity under conditions. We offer simulation results to support our theoretical findings.

scholarly journals Targeted Learning of the Mean Outcome under an Optimal Dynamic Treatment Rule

2015 ◽  
Vol 3 (1) ◽  
pp. 61-95 ◽  
Author(s):  
Mark J. van der Laan ◽  
Alexander R. Luedtke
Keyword(s):  
Statistical Inference ◽  
Empirical Process ◽  
Process Conditions ◽  
Order Difference ◽  
Dynamic Treatment ◽  
Optimal Dynamic ◽  
The Mean ◽  
Data Adaptive ◽  
Time Dependent Covariates ◽  
Asymptotic Linearity

AbstractWe consider estimation of and inference for the mean outcome under the optimal dynamic two time-point treatment rule defined as the rule that maximizes the mean outcome under the dynamic treatment, where the candidate rules are restricted to depend only on a user-supplied subset of the baseline and intermediate covariates. This estimation problem is addressed in a statistical model for the data distribution that is nonparametric beyond possible knowledge about the treatment and censoring mechanism. This contrasts from the current literature that relies on parametric assumptions. We establish that the mean of the counterfactual outcome under the optimal dynamic treatment is a pathwise differentiable parameter under conditions, and develop a targeted minimum loss-based estimator (TMLE) of this target parameter. We establish asymptotic linearity and statistical inference for this estimator under specified conditions. In a sequentially randomized trial the statistical inference relies upon a second-order difference between the estimator of the optimal dynamic treatment and the optimal dynamic treatment to be asymptotically negligible, which may be a problematic condition when the rule is based on multivariate time-dependent covariates. To avoid this condition, we also develop TMLEs and statistical inference for data adaptive target parameters that are defined in terms of the mean outcome under the estimate of the optimal dynamic treatment. In particular, we develop a novel cross-validated TMLE approach that provides asymptotic inference under minimal conditions, avoiding the need for any empirical process conditions. We offer simulation results to support our theoretical findings.

scholarly journals Influence Re-weighted G-Estimation

2016 ◽  
Vol 12 (1) ◽  
pp. 157-177
Author(s):  
Benjamin Rich ◽  
Erica E. M. Moodie ◽  
David A. Stephens
Keyword(s):  
Individualized Medicine ◽  
Treatment Strategies ◽  
Treatment Regimens ◽  
Dynamic Treatment Regime ◽  
Dynamic Treatment ◽  
Optimal Dynamic ◽  
Data Adaptive ◽  
Doubly Robust ◽  
The Cost ◽  
The Impact

Abstract Individualized medicine is an area that is growing, both in clinical and statistical settings, where in the latter, personalized treatment strategies are often referred to as dynamic treatment regimens. Estimation of the optimal dynamic treatment regime has focused primarily on semi-parametric approaches, some of which are said to be doubly robust in that they give rise to consistent estimators provided at least one of two models is correctly specified. In particular, the locally efficient doubly robust g-estimation is robust to misspecification of the treatment-free outcome model so long as the propensity model is specified correctly, at the cost of an increase in variability. In this paper, we propose data-adaptive weighting schemes that serve to decrease the impact of influential points and thus stabilize the estimator. In doing so, we provide a doubly robust g-estimator that is also robust in the sense of Hampel (15).

scholarly journals Postoperative analgesia effects of sulfentanyl plus dexmedetomidine in patients received VATS

Pteridines ◽  
2020 ◽  
Vol 31 (1) ◽  
pp. 55-60
Author(s):  
Haoyu Jiang ◽  
Ying Zheng ◽  
Chang Liu ◽  
Ying Bao
Keyword(s):  
Statistical Difference ◽  
Control Group ◽  
Control Groups ◽  
Significant Statistical Difference ◽  
Post Surgery ◽  
Intravenous Analgesia ◽  
The Mean ◽  
Nsclc Patients ◽  
And Control ◽  
Time Point

AbstractBackground To evaluate sulfentanyl combined with dexmedetomidine hydrochloride on postoperative analgesia in patients who received video-assisted thoracic surgery (VATS) and its effects on serum norepinephrine (NE), dopamine (DA), 5-hydroxytryptamine (5-HT), and prostaglandin (PGE2).Material and Methods Ninety-nine non-small cell lung cancer (NSCLC) patients who received VATS were included in the study. All the patients received intravenous inhalation compound anesthesia. Of the 99 cases, 49 subjects (control group) received sulfentanyl for patient controlled intravenous analgesia (PICA) and other 50 cases (experiment group) received sulfentanyl combined with dexmedetomidine hydrochloride for PICA after operation of VATS. The analgesic effects of the two groups were evaluated according to Visual Analogue Scales (VAS) and the Bruggrmann Comfort Scale (BCS). The serum pain mediator of NE, DA, 5-HT, and PGE2 were examined and compared between the two groups in the first 24 h post-surgery.Results The VAS scores for the experiment group were significant lower than that of control group on the time points of 8, 16, and 24 h post-surgery (pall<0.05), and the BCS scores of the experiment group in the time points of 8, 16, and 24 h were significantly higher than that of controls (p<0.05). However, the VAS and BCS scores were not statistical differently in the time point of 1, 2, and 4 h post-surgery (pall>0.05). The mean sulfentanyl dosage was 63.01 ± 5.14 μg and 67.12 ± 6.91 μg for the experiment and control groups respectively with significant statistical difference (p<0.05). The mean analgesic pump pressing times were 4.30 ± 1.31 and 5.31 ± 1.46 for experiment and control groups respectively with significant statistical difference (p<0.05). The serum NE, DA, 5-HT, and PGE2 levels were significantly lower in the experimental group compared to that of control group in the time point of 12 h post-surgery (pall<0.05). The side effects of nausea, vomiting, delirium, rash, and hypotension atrial fibrillation were not statistically different between the two groups (pall>0.05).Conclusion Patient controlled intravenous analgesia of sulfentanyl combined with dexmedetomidine hydrochloride was effective in reducing the VAS score and serum pain mediators in NSCLC patients who received VAST.

scholarly journals Pupil Dilation with Intracameral Epinephrine Hydrochloride during Phacoemulsification and Intraocular Lens Implantation

2016 ◽  
Vol 2016 ◽  
pp. 1-5 ◽  
Author(s):  
A-Yong Yu ◽  
Hua Guo ◽  
Qin-Mei Wang ◽  
Fang-Jun Bao ◽  
Jing-Hai Huang
Keyword(s):  
Intraocular Lens ◽  
Pupil Diameter ◽  
Pupil Dilation ◽  
Surgical Performance ◽  
Corneal Incision ◽  
Iol Implantation ◽  
The Mean ◽  
Contralateral Eye ◽  
To Receive ◽  
Time Point

Objective. To investigate mydriatic effect of intracamerally injected epinephrine hydrochloride during phacoemulsification and intraocular lens (IOL) implantation.Methods. Eighteen cataract patients for bilateral phacoemulsification were enrolled. To dilate pupil, one eye was randomly selected to receive intracamerally 1 mL epinephrine hydrochloride 0.001% for 1 minute after corneal incision (intracameral group), and the contralateral eye received 3 drops of compound tropicamide 0.5% and phenylephrine 0.5% at 5-minute intervals 30 minutes before surgery (topical group). Pupil diameters were measured before corneal incision, before ophthalmic viscoelastic device (OVD) injection, after OVD injection, before IOL implantation, and at the end of surgery.Results. At each time point, the mean pupil diameter in the intracameral group was2.20±0.08,5.09±0.20,6.76±0.19,6.48±0.18, and5.97±0.24 mm, respectively, and in the topical group it was7.98±0.15,7.98±0.15,8.53±0.14,8.27±0.16, and7.93±0.20 mm, respectively. The topical group consistently had larger mydriatic effects than the intracameral group (P<0.05). The onset of mydriatic effect was rapid in the intracameral group. There was no difference in surgical performance or other parameters between groups.Conclusions. Intracameral epinephrine hydrochloride appears to be an alternative to the mydriatic modalities for phacoemulsification and IOL implantation. In comparison with topical mydriatics, intracameral epinephrine hydrochloride offers easier preoperative preparation, more rapid pupil dilation, and comparable surgical performance.

Maternal attendance patterns of Steller sea lions (Eumetopias jubatus) from stable and declining populations in Alaska

2003 ◽  
Vol 81 (2) ◽  
pp. 340-348 ◽  
Author(s):  
Linda L Milette ◽  
Andrew W Trites
Keyword(s):  
Prey Availability ◽  
A Priori ◽  
Population Decline ◽  
Perinatal Period ◽  
Eumetopias Jubatus ◽  
Steller Sea Lions ◽  
Sea Lions ◽  
First Feeding ◽  
Attendance Patterns ◽  
The Mean

Maternal attendance patterns of Alaskan Steller sea lions (Eumetopias jubatus) were compared during the summer breeding seasons in 1994 and 1995 at Sugarloaf Island (a declining population) and Lowrie Island (a stable population). Our goal was to determine whether there were differences in maternal attendance between the two populations that were consistent with the hypothesis that lactating Steller sea lions in the area of decline were food-limited during summer. Our a priori expectations were based on well-documented behavioural responses of otariids to reduced prey availability. We found that foraging trips were significantly shorter in the area of population decline, counter to initial predictions. The mean length of foraging trips in the declining area was 19.5 h compared with 24.9 h in the stable area. In contrast, the mean perinatal period (time between parturition and first feeding trip) was significantly longer in the area of decline (9.9 versus 7.9 days), again countering initial predictions. The mean length of shore visits for the declining population was also significantly longer (27.0 h compared with 22.6 h where the population was stable). For both populations, the mean time that mothers foraged increased as pups grew older, whereas the time that they spent on shore with their pups became shorter. Behavioural observations of maternal attendance patterns are inconsistent with the hypothesis that lactating Steller sea lions from the declining population had difficulty obtaining prey during summer.

scholarly journals Discussion of Identification, Estimation and Approximation of Risk under Interventions that Depend on the Natural Value of Treatment Using Observational Data, by Jessica Young, Miguel Hernán, and James Robins

2014 ◽  
Vol 3 (1) ◽  
Author(s):  
Mark J. van der Laan ◽  
Alexander R. Luedtke ◽  
Iván Díaz
Keyword(s):  
Statistical Model ◽  
Observational Data ◽  
Robust Estimation ◽  
Statistical Estimation ◽  
Estimation Problem ◽  
Target Parameter ◽  
Adaptive Parameters ◽  
The Mean ◽  
Data Adaptive ◽  
Statistical Estimation Problem

AbstractYoung, Hernán, and Robins consider the mean outcome under a dynamic intervention that may rely on the natural value of treatment. They first identify this value with a statistical target parameter, and then show that this statistical target parameter can also be identified with a causal parameter which gives the mean outcome under a stochastic intervention. The authors then describe estimation strategies for these quantities. Here we augment the authors’ insightful discussion by sharing our experiences in situations where two causal questions lead to the same statistical estimand, or the newer problem that arises in the study of data adaptive parameters, where two statistical estimands can lead to the same estimation problem. Given a statistical estimation problem, we encourage others to always use a robust estimation framework where the data generating distribution truly belongs to the statistical model. We close with a discussion of a framework which has these properties.

scholarly journals A Novel Method for the Separation of Overlapping Pollen Species for Automated Detection and Classification

2016 ◽  
Vol 2016 ◽  
pp. 1-12 ◽  
Author(s):  
Santiago Tello-Mijares ◽  
Francisco Flores
Keyword(s):  
Multilayer Perceptron ◽  
Cross Validation ◽  
Mean Shift ◽  
Automated Detection ◽  
Automated Identification ◽  
Fibonacci Series ◽  
Microscopic Images ◽  
Pollen Species ◽  
Novel Method ◽  
The Mean

The identification of pollen in an automated way will accelerate different tasks and applications of palynology to aid in, among others, climate change studies, medical allergies calendar, and forensic science. The aim of this paper is to develop a system that automatically captures a hundred microscopic images of pollen and classifies them into the 12 different species from Lagunera Region, Mexico. Many times, the pollen is overlapping on the microscopic images, which increases the difficulty for its automated identification and classification. This paper focuses on a method to segment the overlapping pollen. First, the proposed method segments the overlapping pollen. Second, the method separates the pollen based on the mean shift process (100% segmentation) and erosion by H-minima based on the Fibonacci series. Thus, pollen is characterized by its shape, color, and texture for training and evaluating the performance of three classification techniques: random tree forest, multilayer perceptron, and Bayes net. Using the newly developed system, we obtained segmentation results of 100% and classification on top of 96.2% and 96.1% in recall and precision using multilayer perceptron in twofold cross validation.

A Geometric Model of the Beating Heart

2007 ◽  
Vol 46 (03) ◽  
pp. 282-286 ◽  
Author(s):  
C. Lorenz ◽  
J. von Berg
Keyword(s):  
Computed Tomography ◽  
A Priori ◽  
Anatomical Landmarks ◽  
Motion Model ◽  
A Priori Information ◽  
Average Accuracy ◽  
Tomography Image ◽  
The Mean ◽  
Priori Information ◽  
Computed Tomography Image

Summary Objectives : A comprehensive model of the human heart that covers multiple surfaces, like those of the four chambers and the attached vessels, is presented. It also contains the coronary arteries and a set of 25 anatomical landmarks. The statistical model is intended to provide a priori information for automated diagnostic and interventional procedures. Methods : The end-diastolic phase of the model was adapted to fit 27 clinical multi-slice computed tomography images, thus reflecting the anatomical variability to be observed in that sample. A mean cardiac motion model was also calculated from a set of eleven multi-phase computed tomography image sets. A number of experiments were performed to determine the accuracy of model-based predictions done on unseen cardiac images. Results : Using an additional deformable surface technique, the model allows for determination of all chambers and the attached vessels on the basis of given anatomical landmarks with an average accuracy of 1.1 mm. After such an individualization of the model by surface adaptation the centerlines of the three main coronary arteries may be estimated with an average accuracy of 5.2 mm. The mean motion model was used to estimate the cardiac phase of an unknown multislice computed tomography image. Conclusion : The mean shape model of the human heart as presented here complements automated image analysis methods with the required a priori information about anatomical constraints to make them work fast and robustly.

Sign in / Sign up

Export Citation Format

Share Document