Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

General General

Learning Spatial Attention for Face Super-Resolution.

In IEEE transactions on image processing : a publication of the IEEE Signal Processing Society

General image super-resolution techniques have difficulties in recovering detailed face structures when applying to low resolution face images. Recent deep learning based methods tailored for face images have achieved improved performance by jointly trained with additional task such as face parsing and landmark prediction. However, multi-task learning requires extra manually labeled data. Besides, most of the existing works can only generate relatively low resolution face images (e.g., 128×128), and their applications are therefore limited. In this paper, we introduce a novel SPatial Attention Residual Network (SPARNet) built on our newly proposed Face Attention Units (FAUs) for face super-resolution. Specifically, we introduce a spatial attention mechanism to the vanilla residual blocks. This enables the convolutional layers to adaptively bootstrap features related to the key face structures and pay less attention to those less feature-rich regions. This makes the training more effective and efficient as the key face structures only account for a very small portion of the face image. Visualization of the attention maps shows that our spatial attention network can capture the key face structures well even for very low resolution faces (e.g., 16×16). Quantitative comparisons on various kinds of metrics (including PSNR, SSIM, identity similarity, and landmark detection) demonstrate the superiority of our method over current state-of-the-arts. We further extend SPARNet with multi-scale discriminators, named as SPARNetHD, to produce high resolution results (i.e., 512×512). We show that SPARNetHD trained with synthetic data cannot only produce high quality and high resolution outputs for synthetically degraded face images, but also show good generalization ability to real world low quality face images.

Chen Chaofeng, Gong Dihong, Wang Hao, Li Zhifeng, Wong Kwan-Yee K

2020-Dec-14

General General

Insights into Algorithms for Separable Nonlinear Least Squares Problems.

In IEEE transactions on image processing : a publication of the IEEE Signal Processing Society

Separable nonlinear least squares (SNLLS) problems have attracted interest in a wide range of research fields such as machine learning, computer vision, and signal processing. During the past few decades, several algorithms, including the joint optimization algorithm, alternated least squares (ALS) algorithm, embedded point iterations (EPI) algorithm, and variable projection (VP) algorithms, have been employed for solving SNLLS problems in the literature. The VP approach has been proven to be quite valuable for SNLLS problems and the EPI method has been successful in solving many computer vision tasks. However, no clear explanations about the intrinsic relationships of these algorithms have been provided in the literature. In this paper, we give some insights into these algorithms for SNLLS problems. We derive the relationships among different forms of the VP algorithms, EPI algorithm and ALS algorithm. In addition, the convergence and robustness of some algorithms are investigated. Moreover, the analysis of the VP algorithm generates a negative answer to Kaufman's conjecture. Numerical experiments on the image restoration task, fitting the time series data using the radial basis function network based autoregressive (RBF-AR) model, and bundle adjustment are given to compare the performance of different algorithms.

Chen Guang-Yong, Gan Min, Wang Shu-Qiang, Philip Chena C L

2020-Dec-14

General General

Predictive Models on the Rise, But Do They Work for Health Care?

In IEEE pulse

Predictive models are designed to remove some of the subjectivity inherent in medical decision-making and to automate certain health-related services with the idea of improving the accuracy of diagnosis, providing personalized treatment options, and streamlining the health care industry overall. More and more of these models using approaches including machine learning are showing up for use in doctor's offices and hospitals, as well as in telemedicine applications, which have become prevalent with the growing demand for online alternatives to office visits.

Mertz Leslie

General General

Identifying Patient Phenotype Cohorts Using Prehospital Electronic Health Record Data.

In Prehospital emergency care : official journal of the National Association of EMS Physicians and the National Association of State EMS Directors

Objective: Emergency medical services (EMS) provide critical interventions for patients with acute illness and injury and are important in implementing prehospital emergency care research. Retrospective, manual patient record review, the current reference-standard for identifying patient cohorts, requires significant time and financial investment. We developed automated classification models to identify eligible patients for prehospital clinical trials using EMS clinical notes and compared model performance to manual review. Methods: With eligibility criteria for an ongoing prehospital study of chest pain patients, we used EMS clinical notes (n = 1208) to manually classify patients as eligible, ineligible, and indeterminate. We randomly split these same records into training and test sets to develop and evaluate machine-learning (ML) algorithms using natural language processing (NLP) for feature (variable) selection. We compared models to the manual classification to calculate sensitivity, specificity, accuracy, positive predictive value, and F1 measure. We measured clinical expert time to perform review for manual and automated methods. Results: ML models' sensitivity, specificity, accuracy, positive predictive value, and F1 measure ranged from 0.93 to 0.98. Compared to manual classification (N = 363 records), the automated method excluded 90.9% of records as ineligible and leaving only 33 records for manual review. Conclusions: Our ML derived approach demonstrates the feasibility of developing a high-performing, automated classification system using EMS clinical notes to streamline the identification of a specific cardiac patient cohort. This efficient approach can be leveraged to facilitate prehospital patient-trial matching, patient phenotyping (i.e. influenza-like illness), and create prehospital patient registries.

Stemerman Rachel, Bunning Thomas, Grover Joseph, Kitzmiller Rebecca, Patel Mehul D

2020-Dec-14

machine learning, natural language processing, patient phenotype, prehospital

General General

Disentangling temporal dynamics in attention bias from measurement error: A state-space modeling approach.

In Journal of abnormal psychology

Temporal dynamics in attention bias (AB) have gained increasing attention in recent years. It has been proposed that AB is variable over trials within a single test session of the dot-probe task, and that the variability in AB is more predictive of psychopathology than the traditional mean AB score. More important, one of the dynamics indices has shown better reliability than the traditional mean AB score. However, it has been also suggested that the dynamics indices are unable to uncouple random measurement error from true variability in AB, which questions the estimation precision of the dynamics indices. To clarify and overcome this issue, the current article introduces a state-space modeling (SSM) approach to estimate trial-level AB more accurately by filtering random measurement error. The estimation error of the extant dynamics indices versus SSM were evaluated by computer simulations with different parameter settings for the temporal variability and between-person variance in AB. Throughout the simulations, SSM showed robustly lower estimation error than the extant dynamics indices. We also applied these indices to real data sets, which revealed that the dynamics indices overestimate within-person variability relative to SSM. Here SSM indicated less temporal dynamics in AB than previously proposed. These findings suggest that SSM might be a better alternative to estimate trial level AB than the extant dynamics indices. However, it is still unclear whether AB has meaningful in-session variability that is predictive of psychopathology. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Takano Keisuke, Taylor Charles T, Wittekind Charlotte E, Sakamoto Jiro, Ehring Thomas

2020-Dec-14

General General

The Emerging Hazard of AI-Related Health Care Discrimination.

In The Hastings Center report

Artificial intelligence holds great promise for improved health-care outcomes. But it also poses substantial new hazards, including algorithmic discrimination. For example, an algorithm used to identify candidates for beneficial "high risk care management" programs routinely failed to select racial minorities. Furthermore, some algorithms deliberately adjust for race in ways that divert resources away from minority patients. To illustrate, algorithms have underestimated African Americans' risks of kidney stones and death from heart failure. Algorithmic discrimination can violate Title VI of the Civil Rights Act and Section 1557 of the Affordable Care Act when it unjustifiably disadvantages underserved populations. This article urges that both legal and technical tools be deployed to promote AI fairness. Plaintiffs should be able to assert disparate impact claims in health-care litigation, and Congress should enact an Algorithmic Accountability Act. In addition, fairness should be a key element in designing, implementing, validating, and employing AI.

Hoffman Sharona

2020-Dec-14

algorithmic fairness, artificial intelligence, civil rights, discrimination, disparate impact