Your browser doesn't support javascript.
Show: 20 | 50 | 100
Results 1 - 3 de 3
Add filters

Year range
Proc Mach Learn Res ; 146:159-170, 2021.
Article in English | PubMed | ID: covidwho-1772436


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.

PubMed; 2021.
Preprint in English | PubMed | ID: ppcovidwho-8846


Under-representation of certain populations, based on gender, race/ethnicity, and age, in data collection for predictive modeling may yield less-accurate predictions for the under-represented 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. Methods to achieve fairness in the machine learning literature typically build a single prediction model subject to some fairness criteria in a manner that encourages fair prediction performances 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 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 demonstrate the properties of the proposed JFM estimates. Next, we presented the key asymptotic properties for the JFM parameter estimates. We examined the efficacy of the JFM approach in achieving prediction performances and parities, in comparison with the Single Fairness Model, group-separate model, and group-ignorant model through extensive simulations. Finally, we demonstrated the utility of the JFM method in the motivating example to obtain fair risk predictions for under-represented older patients diagnosed with coronavirus disease 2019 (COVID-19).

Multiple Sclerosis Journal ; 26(3 SUPPL):555, 2020.
Article in English | EMBASE | ID: covidwho-1067121


Background: Neurological disability progression occurs across the spectrum of people living with multiple sclerosis (PwMS). Currently, no treatments exist that substantially modify the course of clinical progression in MS, one of the greatest unmet needs in clinical practice. Characterizing the determinants of clinical progression is essential for the development of novel therapeutic agents and treatment approaches that target progression in PwMS. Objectives: The overarching aim of CanProCo is to evaluate a wide spectrum of factors associated with the onset and rate of disease progression in MS, and to describe how these factors interact with one another to influence progression. Methods: CanProCo is a prospective, observational cohort study aiming to recruit 1000 individuals with radiologically-isolated syndrome (RIS), relapsing-remitting MS (RRMS), and primary-progressive MS (PPMS) within 10-15 years of disease onset, and 50 healthy controls (HCs) from five large academic MS centers in Canada. Participants undergo detailed clinical evaluations annually. A subset of participants enrolled within 5-10 years of disease onset (n=500) also have blood, cerebrospinal fluid, and MRIs collected facilitating study of biological measures (e.g. single-cell RNAsequencing[ scRNASeq]), MRI-based microstructural assessment, participant characteristics (self-reported, performance-based, clinician- assessed, health-system based), and environmental factors as determinants contributing to the differential progression in MS. Results: Recruitment commenced in April/May 2019 and n=536 patients have been recruited to date (RRMS=457, PPMS=35, RIS=25, HC=19). Baseline age, sex distribution, and Expanded Disability Status Scale (EDSS) scores (median, range) of each subgroup are: RRMS=38 years, 73% female, EDSS=1.5 (0-6.0);PPMS=52 years, 40% female, EDSS=4.0 (1.5-6.5);RIS=41 years, 68% female, EDSS=0 (0-3.0);HC=37 years, 63% female. Recruitment has surpassed the 50% target but has been paused due to the COVID-19 pandemic. scRNASeq on frozen blood samples has been validated. Conclusions: Halting the progression of MS is a fundamental clinical need to improve the lives of PwMS. Achieving this requires leveraging transdisciplinary approaches to better characterize mechanisms underlying clinical progression. CanProCo is the first prospective cohort study aiming to characterize these determinants to inform the development and implementation of efficacious and effective interventions.