Learning optimal biomarker-guided treatment policy for chronic disorders.

Electroencephalogram (EEG) provides noninvasive measures of brain activity and is found to be valuable for the diagnosis of some chronic disorders. Specifically, pre-treatment EEG signals in the alpha and theta frequency bands have demonstrated some association with antidepressant response, which is well-known to have a low response rate. We aim to design an integrated pipeline that improves the response rate of patients with major depressive disorder by developing a treatment policy guided by the resting state pre-treatment EEG recordings and other treatment effects modifiers. First, we design an innovative automatic site-specific EEG preprocessing pipeline to extract features with stronger signals than raw data. We then estimate the conditional average treatment effect (CATE) using causal forests and use a doubly robust technique to improve efficiency in the estimation of the average treatment effect. We present evidence of heterogeneity in the treatment effect and the modifying power of the EEG features, as well as a significant average treatment effect, a result that cannot be obtained with conventional methods. Finally, we employ an efficient policy learning algorithm to learn an optimal depth-2 treatment assignment decision tree and compare its performance with Q-Learning and outcome-weighted learning via simulation studies and an application to a large multi-site, double-blind, randomized controlled clinical trial, EMBARC.

A Semiparametric Inverse Reinforcement Learning Approach to Characterize Decision Making for Mental Disorders.

Major depressive disorder (MDD) is one of the leading causes of disability-adjusted life years. Emerging evidence indicates the presence of reward processing abnormalities in MDD. An important scientific question is whether the abnormalities are due to reduced sensitivity to received rewards or reduced learning ability. Motivated by the probabilistic reward task (PRT) experiment in the EMBARC study, we propose a semiparametric inverse reinforcement learning (RL) approach to characterize the reward-based decision-making of MDD patients. The model assumes that a subject's decision-making process is updated based on a reward prediction error weighted by the subject-specific learning rate. To account for the fact that one favors a decision leading to a potentially high reward, but this decision process is not necessarily linear, we model reward sensitivity with a non-decreasing and nonlinear function. For inference, we estimate the latter via approximation by I-splines and then maximize the joint conditional log-likelihood. We show that the resulting estimators are consistent and asymptotically normal. Through extensive simulation studies, we demonstrate that under different reward-generating distributions, the semiparametric inverse RL outperforms the parametric inverse RL. We apply the proposed method to EMBARC and find that MDD and control groups have similar learning rates but different reward sensitivity functions. There is strong statistical evidence that reward sensitivity functions have nonlinear forms. Using additional brain imaging data in the same study, we find that both reward sensitivity and learning rate are associated with brain activities in the negative affect circuitry under an emotional conflict task.

High-Throughput Field Plant Phenotyping: A Self-Supervised Sequential CNN Method to Segment Overlapping Plants.

High-throughput plant phenotyping-the use of imaging and remote sensing to record plant growth dynamics-is becoming more widely used. The first step in this process is typically plant segmentation, which requires a well-labeled training dataset to enable accurate segmentation of overlapping plants. However, preparing such training data is both time and labor intensive. To solve this problem, we propose a plant image processing pipeline using a self-supervised sequential convolutional neural network method for in-field phenotyping systems. This first step uses plant pixels from greenhouse images to segment nonoverlapping in-field plants in an early growth stage and then applies the segmentation results from those early-stage images as training data for the separation of plants at later growth stages. The proposed pipeline is efficient and self-supervising in the sense that no human-labeled data are needed. We then combine this approach with functional principal components analysis to reveal the relationship between the growth dynamics of plants and genotypes. We show that the proposed pipeline can accurately separate the pixels of foreground plants and estimate their heights when foreground and background plants overlap and can thus be used to efficiently assess the impact of treatments and genotypes on plant growth in a field environment by computer vision techniques. This approach should be useful for answering important scientific questions in the area of high-throughput phenotyping.

KAT4IA: K-Means Assisted Training for Image Analysis of Field-Grown Plant Phenotypes.

High-throughput phenotyping enables the efficient collection of plant trait data at scale. One example involves using imaging systems over key phases of a crop growing season. Although the resulting images provide rich data for statistical analyses of plant phenotypes, image processing for trait extraction is required as a prerequisite. Current methods for trait extraction are mainly based on supervised learning with human labeled data or semisupervised learning with a mixture of human labeled data and unsupervised data. Unfortunately, preparing a sufficiently large training data is both time and labor-intensive. We describe a self-supervised pipeline (KAT4IA) that uses K-means clustering on greenhouse images to construct training data for extracting and analyzing plant traits from an image-based field phenotyping system. The KAT4IA pipeline includes these main steps: self-supervised training set construction, plant segmentation from images of field-grown plants, automatic separation of target plants, calculation of plant traits, and functional curve fitting of the extracted traits. To deal with the challenge of separating target plants from noisy backgrounds in field images, we describe a novel approach using row-cuts and column-cuts on images segmented by transform domain neural network learning, which utilizes plant pixels identified from greenhouse images to train a segmentation model for field images. This approach is efficient and does not require human intervention. Our results show that KAT4IA is able to accurately extract plant pixels and estimate plant heights.

Metabolomic Profiling of Nicotiana Spp. Nectars Indicate That Pollinator Feeding Preference Is a Stronger Determinant Than Plant Phylogenetics in Shaping Nectar Diversity.

Floral nectar is a rich secretion produced by the nectary gland and is offered as reward to attract pollinators leading to improved seed set. Nectars are composed of a complex mixture of sugars, amino acids, proteins, vitamins, lipids, organic and inorganic acids. This composition is influenced by several factors, including floral morphology, mechanism of nectar secretion, time of flowering, and visitation by pollinators. The objective of this study was to determine the contributions of flowering time, plant phylogeny, and pollinator selection on nectar composition in Nicotiana. The main classes of nectar metabolites (sugars and amino acids) were quantified using gas chromatography/mass spectrometric analytical platforms to identify differences among fifteen Nicotiana species representing day- and night-flowering plants from ten sections of the genus that are visited by five different primary pollinators. The nectar metabolomes of different Nicotiana species can predict the feeding preferences of the target pollinator(s) of each species, and the nectar sugars (i.e., glucose, fructose, and sucrose) are a distinguishing feature of Nicotiana species phylogeny. Moreover, comparative statistical analysis indicate that pollinators are a stronger determinant of nectar composition than plant phylogeny.