In data collection for predictive modeling, under-representation of certain groups, based on gender, race/ethnicity, or age, may yield less-accurate predictions for these groups. Recently, this issue of fairness in predictions has attracted significant attention, as data-driven models are increasingly utilized to perform crucial decision-making tasks. Existing methods to achieve fairness in the machine learning literature typically build a single prediction model in a manner that encourages fair prediction performance for all groups. These approaches have two major limitations: i) fairness is often achieved by compromising accuracy for some groups; ii) the underlying relationship between dependent and independent variables may not be the same across groups. We propose a Joint Fairness Model (JFM) approach for logistic regression models for binary outcomes that estimates group-specific classifiers using a joint modeling objective function that incorporates fairness criteria for prediction. We introduce an Accelerated Smoothing Proximal Gradient Algorithm to solve the convex objective function, and present the key asymptotic properties of the JFM estimates. Through simulations, we demonstrate the efficacy of the JFM in achieving good prediction performance and across-group parity, in comparison with the single fairness model, group-separate model, and group-ignorant model, especially when the minority group's sample size is small. Finally, we demonstrate the utility of the JFM method in a real-world example to obtain fair risk predictions for under-represented older patients diagnosed with coronavirus disease 2019 (COVID-19).

Applying standard machine learning approaches for classification can produce unequal results across different demographic groups. When then used in real-world settings, these inequities can have negative societal impacts. This has motivated the development of various approaches to fair classification with machine learning models in recent years. In this paper, we consider the problem of modifying the predictions of a blackbox machine learning classifier in order to achieve fairness in a multiclass setting. To accomplish this, we extend the 'post-processing' approach in Hardt et al. 2016, which focuses on fairness for binary classification, to the setting of fair multiclass classification. We explore when our approach produces both fair and accurate predictions through systematic synthetic experiments and also evaluate discrimination-fairness tradeoffs on several publicly available real-world application datasets. We find that overall, our approach produces minor drops in accuracy and enforces fairness when the number of individuals in the dataset is high relative to the number of classes and protected groups.

The widespread availability of high-dimensional electronic healthcare record (EHR) datasets has led to significant interest in using such data to derive clinical insights and make risk predictions. More specifically, techniques from machine learning are being increasingly applied to the problem of dynamic survival analysis, where updated time-to-event risk predictions are learned as a function of the full covariate trajectory from EHR datasets. EHR data presents unique challenges in the context of dynamic survival analysis, involving a variety of decisions about data representation, modeling, interpretability, and clinically meaningful evaluation. In this paper we propose a new approach to dynamic survival analysis which addresses some of these challenges. Our modeling approach is based on learning a global parametric distribution to represent population characteristics and then dynamically locating individuals on the time-axis of this distribution conditioned on their histories. For evaluation we also propose a new version of the dynamic C-Index for clinically meaningful evaluation of dynamic survival models. To validate our approach we conduct dynamic risk prediction on three real-world datasets, involving COVID-19 severe outcomes, cardiovascular disease (CVD) onset, and primary biliary cirrhosis (PBC) time-to-transplant. We find that our proposed modeling approach is competitive with other well-known statistical and machine learning approaches for dynamic risk prediction, while offering potential advantages in terms of interepretability of predictions at the individual level.

Dynamic survival analysis is a variant of traditional survival analysis where time-to-event predictions are updated as new information arrives about an individual over time. In this paper we propose a new approach to dynamic survival analysis based on learning a global parametric distribution, followed by individualization via truncating and renormalizing that distribution at different locations over time. We combine this approach with a likelihood-based loss that includes predictions at every time step within an individual's history, rather than just including one term per individual. The combination of this loss and model results in an interpretable approach to dynamic survival, requiring less fine tuning than existing methods, while still achieving good predictive performance. We evaluate the approach on the problem of predicting hospital mortality for a dataset with over 6900 COVID-19 patients.

Abstract Motivation Diagnosis of leukemia relies on accurate identification of leukemic cell populations. Flow cytometry (FCM) is a primary diagnostic assay routinely used in clinical practice. The assay workflows consist of multiple manual gating steps performed by technicians, followed by interpretation by hematopathologists. Challenges to this process include technical variability in manual gating, difficulty in identification of the atypical leukemic cells, and the growing number of antigens used for diagnosis. Methods Instead of conducting ad hoc analysis of individual samples, our proposed computational approach leverages preexisting clinical FCM data to improve robustness of the computational identification of leukemic cells. Instead of separating cell population identification and sample classification into two steps, our machine learning classification method optimizes them simultaneously, producing gating locations that are recognizable to hematopathologists. Results Our study consists of 10-color FCM data from blood or bone marrow samples of 129 random subjects for chronic lymphocytic leukemia (CLL) diagnosis. Our initial automated gating analysis rendered an accuracy of 89% matched the diagnosis of the hematopathologist. In the remaining cases with discrepant results, the misclassified CLL cells had atypical molecular phenotypes, making them difficult to identify. In our improved pipeline, we demonstrate that these atypical types of CLL cells can be clearly captured using a non-linear embedding dimensionality reduction step. Conclusion The results demonstrate the power of a novel computational analysis pipeline for improving the identification of aberrant leukemia cells for precision diagnosis.

High-throughput single-cell cytometry technologies have significantly improved our understanding of cellular phenotypes to support translational research and the clinical diagnosis of hematological and immunological diseases. However, subjective and ad hoc manual gating analysis does not adequately handle the increasing volume and heterogeneity of cytometry data for optimal diagnosis. Prior work has shown that machine learning can be applied to classify cytometry samples effectively. However, many of the machine learning classification results are either difficult to interpret without using characteristics of cell populations to make the classification, or suboptimal due to the use of inaccurate cell population characteristics derived from gating boundaries. To date, little has been done to optimize both the gating boundaries and the diagnostic accuracy simultaneously. In this work, we describe a fully discriminative machine learning approach that can simultaneously learn feature representations (e.g., combinations of coordinates of gating boundaries) and classifier parameters for optimizing clinical diagnosis from cytometry measurements. The approach starts from an initial gating position and then refines the position of the gating boundaries by gradient descent until a set of globally-optimized gates across different samples are achieved. The learning procedure is constrained by regularization terms encoding domain knowledge that encourage the algorithm to seek interpretable results. We evaluate the proposed approach using both simulated and real data, producing classification results on par with those generated via human expertise, in terms of both the positions of the gating boundaries and the diagnostic accuracy. © 2019 The Authors. Cytometry Part A published by Wiley Periodicals, Inc. on behalf of International Society for Advancement of Cytometry.

Individuals create and consume more diverse data about themselves today than any time in history. Sources of this data include wearable devices, images, social media, geo-spatial information and more. A tremendous opportunity rests within cross-modal data analysis that leverages existing domain knowledge methods to understand and guide human health. Especially in chronic diseases, current medical practice uses a combination of sparse hospital based biological metrics (blood tests, expensive imaging, etc.) to understand the evolving health status of an individual. Future health systems must integrate data created at the individual level to better understand health status perpetually, especially in a cybernetic framework. In this work we fuse multiple user created and open source data streams along with established biomedical domain knowledge to give two types of quantitative state estimates of cardiovascular health. First, we use wearable devices to calculate cardiorespiratory fitness (CRF), a known quantitative leading predictor of heart disease which is not routinely collected in clinical settings. Second, we estimate inherent genetic traits, living environmental risks, circadian rhythm, and biological metrics from a diverse dataset. Our experimental results on 24 subjects demonstrate how multi-modal data can provide personalized health insight. Understanding the dynamic nature of health status will pave the way for better health based recommendation engines, better clinical decision making and positive lifestyle changes.

Objectives To identify differences in letters of recommendation (LORs) of applicants to a single ophthalmology residency program by gender, race, academic performance, and match outcome. Design This was a retrospective analysis of LORS for 2,523 applicants (7,569 letters) to the University of California, Irvine ophthalmology residency program from 2011 to 2018. Methods Programming scripts were employed to determine the number of times 22 key words from four thematic categories (standout words, ability, grindstone, and compassion) appeared in LORs for each applicant. A chi-square test was performed to assess for possible differences in the presence of each key word by the following characteristics: gender, underrepresented minority (URM) status, Alpha Omega Alpha (AOA) membership, the United States Medical Licensing Exam (USMLE) Step 1 score, and match outcome. Linear regressions were created to determine the frequency at which words in each thematic category appeared according to the same baseline characteristics. Results In the LORs, females were more likely to be described as “empathetic” (p = 0.002), URMs were more likely to be described as “caring” (p = 0.002), high Step 1 scorers (≥240) were more likely to be described as “outstanding” (p = 0.002), and matched students were more likely to be described as “exceptional” (p = 0.001), “outstanding” (p < 0.001), and “superb” (p = 0.001). Standout words appeared more often in the LORs of AOA members, matched candidates, and high Step 1 scorers (p < 0.001 for all comparisons). “Competent” appeared more commonly in LORs for low Step 1 scorers (p < 0.001) and unmatched applicants (p = 0.001). Conclusion This study identifies differences in LORs by gender, URM status, and achievement including successful ophthalmology residency match. Females and URMs were more likely to be described as “empathetic” and “caring,” respectively; otherwise, we detected no gender or racial disparities in key word use in LORs. Candidates with high USMLE Step 1 scores or AOA membership had a higher frequency of standout words in their LORs. Whether they were truly more qualified in various dimensions or if they benefited from a halo effect bias warrants further investigation. There was a significant difference in the number of standout words in LORs between matched and unmatched applicants, suggesting that key word frequency may be a relevant metric for LOR appraisal.