Prognostication and Risk Factors for Cystic Fibrosis via Automated Machine Learning

Abstract

Accurate prediction of survival for cystic fibrosis (CF) patients is instrumental in establishing the optimal timing for referring patients with terminal respiratory failure for lung transplantation (LT). Current practice considers referring patients for LT evaluation once the forced expiratory volume (FEV1) drops below 30% of its predicted nominal value. While FEV1 is indeed a strong predictor of CF-related mortality, we hypothesized that the survival behavior of CF patients exhibits a lot more heterogeneity. To this end, we developed an algorithmic framework, which we call AutoPrognosis, that leverages the power of machine learning to automate the process of constructing clinical prognostic models, and used it to build a prognostic model for CF using data from a contemporary cohort that involved 99% of the CF population in the UK. AutoPrognosis uses Bayesian optimization techniques to automate the process of configuring ensembles of machine learning pipelines, which involve imputation, feature processing, classification and calibration algorithms. Because it is automated, it can be used by clinical researchers to build prognostic models without the need for in-depth knowledge of machine learning. Our experiments revealed that the accuracy of the model learned by AutoPrognosis is superior to that of existing guidelines and other competing models.

Diagnostic value of spirometry vs impulse oscillometry: A comparative study in children with sickle cell disease

“…13 XGBoost is a well-known ML technique that utilizes boosting decision trees algorithm to build a prediction model and has been utilized to predict the outcome of cystic fibrosis based on spirometry measures. 14,15 Amaral et al 16…”

“…ML methods are more precise and accurate in terms of prediction abilities compared with traditional statistical methods, because complex intervariable interactions are taken into account in ML only . XGBoost is a well‐known ML technique that utilizes boosting decision trees algorithm to build a prediction model and has been utilized to predict the outcome of cystic fibrosis based on spirometry measures . Amaral et al demonstrated that boosting decision trees, trained by IOS parameters, can be successfully used to diagnose OAD in patients with asthma.…”

Gender and ethnic differences in publication of BMJ letters to the editor: an observational study using machine learning

“…It has also opened countless possibilities in medicine across the entire patient journey, from triage and improving attendance, to automated disease diagnosis, prognostication, management and even the discovery or repurposing of new medications. [39][40][41][42][43][44][45] Unanswered questions and future research This work highlights important associations in past data; however, more research is necessary to draw concrete conclusions regarding the reasons for these associations. For example, other confounders might be considered, including communication style, field of study of the rapid response author and of the article being responded to, and the locations of institutes of submitted pieces.…”

Trial Refresh: A Case for an Adaptive Platform Trial for Pulmonary Exacerbations of Cystic Fibrosis

“…For instance, patient-centric registries allow patients to securely register through online registration with configurable online informed consent (Bellgard et al, 2012; Napier et al, 2017). Others have demonstrated the potential value of rich and well-collected patient registry data for improving patient decision-making in CF beyond simple rule-based algorithms (Alaa and van der Schaar, 2018). Online clinical reporting forms and participant questionnaires that can be configured by coordinators without software development skills enables the digital health solution to dynamically adapt to requirements (Bellgard et al, 2014).…”