This paper presents a quantitative imaging method and software technology to predict the risk and assess the severity of respiratory diseases in premature babies by fusing information from multiple sources: non-invasive low-radiation chest X-ray (CXR) imaging and clinical parameters. Prematurity is the largest single cause of death in children under five in the world. Lower respiratory tract infections (LRTI) are the top cause of hospitalization and mortality in prematurity. However, there is no objective clinical marker to predict and prevent severe LRTI in the 15 million babies born prematurely every year worldwide. Traditionally, imaging biomarkers of lung disease from computed tomography have been successfully used in adults, but they entail heightened risks for children due to cumulative radiation and the need for sedation. The proposed technology is the first approach that uses low-radiation CXR imaging to predict hospitalization due to LRTI in prematurity. The method uses deep learning to quantify heterogeneous patterns (air trapping and irregular opacities) in the chest, which are combined with clinical parameters to predict the risk of LRTI. Our preliminary results obtained using a data obtained from ten premature subjects with LRTI showed high correlation between our imaging biomarkers and the rehospitalization of these subjects R2=0.98).
Representation learning through deep learning (DL) architecture has shown tremendous potential for identification, local-
ization, and texture classification in various medical imaging modalities. However, DL applications to segmentation of
objects especially to deformable objects are rather limited and mostly restricted to pixel classification. In this work, we
propose marginal shape deep learning (MaShDL), a framework that extends the application of DL to deformable shape
segmentation by using deep classifiers to estimate the shape parameters. MaShDL combines the strength of statistical
shape models with the automated feature learning architecture of DL. Unlike the iterative shape parameters estimation
approach of classical shape models that often leads to a local minima, the proposed framework is robust to local minima
optimization and illumination changes. Furthermore, since the direct application of DL framework to a multi-parameter
estimation problem results in a very high complexity, our framework provides an excellent run-time performance solution
by independently learning shape parameter classifiers in marginal eigenspaces in the decreasing order of variation. We
evaluated MaShDL for segmenting the lung field from 314 normal and abnormal pediatric chest radiographs and obtained
a mean Dice similarity coefficient of 0:927 using only the four highest modes of variation (compared to 0:888 with classical
ASM1 (p-value=0:01) using same configuration). To the best of our knowledge this is the first demonstration of using DL
framework for parametrized shape learning for the delineation of deformable objects.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.