Learning and combining image neighborhoods using random forests for neonatal brain disease classification

Abstract

It is challenging to characterize and classify normal and abnormal brain development during early childhood. To reduce the complexity of heterogeneous data population, manifold learning techniques are increasingly applied, which find a low-dimensional representation of the data, while preserving all relevant information. The neighborhood definition used for constructing manifold representations of the population is crucial for preserving the similarity structure and it is highly application dependent. The recently proposed neighborhood approximation forests learn a neighborhood structure in a dataset based on a user-defined distance. We propose a framework to learn multiple pairwise distances in a population of brain images and to combine them in an unsupervised manner optimally in a manifold learning step. Unlike other methods that only use a univariate distance measure, our method allows for a natural combination of multiple distances from heterogeneous sources. As a result, it yields a representation of the population that preserves the multiple distances. Furthermore, our method also selects the most predictive features associated with the distances. We evaluate our method in neonatal magnetic resonance images of three groups (term controls, patients affected by intrauterine growth restriction and mild isolated ventriculomegaly). We show that combining multiple distances related to the condition improves the overall characterization and classification of the three clinical groups compared to the use of single distances and classical unsupervised manifold learning.

The past, current, and future of neonatal intensive care units with artificial intelligence: a systematic review

“… - EHR (not included AI analysis) - Small sample size - Retrospective design Ball et al, 2015 90 Random Forest (RF) To compare whole-brain functional connectivity in preterm newborns at term-equivalent age with healthy term-born neonates in order to determine if preterm birth leads in particular changes to functional connectivity by term-equivalent age. 105 preterm infants and 26 term controls Both resting state functional MRI and T2-weighted Brain MRI 80% (accuracy) + Prospective + Connectivity differences between term and preterm brain - Not well-established model Smyser et al, 2016 88 Support vector machine (SVM)-multivariate pattern analysis (MVPA) To compare resting state-activity of preterm-born infants (Scanned at term equivalent postmenstrual age) to term infants 50 preterm infants (born at 23–29 weeks of gestation and without moderate–severe brain injury) 50 term-born control infants studied Functional MRI data + Clinical variables 84% (accuracy) + Prospective + GA at birth was used as an indicator of the degree of disruption of brain development + Optimal methods for rs-fMRI data acquisition and preprocessing for this population have not yet been rigorously defined - Small sample size Zimmer et al, 2017 93 NAF: Neighborhood approximation forest classifier of forests To reduce the complexity of heterogeneous data population, manifold learning techniques are applied, which find a low-dimensional representation of the data. 111 infants (NC, 70 subjects), affected by IUGR (27 subjects) or VM (14 subjects).…”

Digital Health Europe (DHE) Twinning on severe asthma—kick-off meeting report

“…More general machine learning techniques, such as manifold learning, can be used to make sense of large and heterogeneous datasets, such as the ones involved in clinical studies. Examples of these techniques and their application to medical datasets have been shown (44). This strategy seems particularly suited for the multifactorial study of rhinitis and asthma, and is being developed in our on-going research.…”

Multi-modal neuroimaging feature selection with consistent metric constraint for diagnosis of Alzheimer's disease

“…Sample similarity metrics learning via random forest has been used in a variety of applications, such as disease classification and image segmentation (Mitra et al, 2014). In addition, some recent studies have incorporated the computational similarity methods into medical imaging analysis (Zimmer et al, 2017). For example, Veronika et al [47] proposed a method for calculating the similarity via random forests and combining images to determine the classification of neonatal brain disease.…”