Feature Extraction and Classification to Diagnose Hypoxic-Ischemic Encephalopathy Patients by Using Susceptibility-Weighted MRI Images

Abstract

Abstract. In this paper. a method is presented to enable automatic classification of the degree of abnormality of susceptibility-weighted images (SWI) acquired from babies with hypoxic-ischemic encephalopathy (HIE), in order to more accurately predict eventual cognitive and motor outcomes in these infants. SWI images highlight the cerebral venous vasculature and can reflect abnormalities in blood flow and oxygenation, which may be linked to adverse outcomes. A qualitative score based on magnetic resonance imaging (MRI) analyses is assigned to SWIs by specialists to determine the severity of abnormality in an HIE patient. The method allows the detection of image ridges, representing the vessels in SWIs, and the histogram of the ridges grey scales. A curve with only four parameters is fitted to the histograms. These parameters are then used to estimate the SWI abnormality score. The images are classified by using a kNNand multiple SVM classifiers based on the parameters of the fitting curves. The algorithm is tested on an SWI-MRI dataset consisting of 10 healthy infants and 48 infants with HIE with a range of SWI abnormality scores between 1 and 7. The accuracy of classifying babies with HIE vs. those without (ie: healthy controls) using our algorithm with a leave-one-out strategy is measured as 91.38%. Our method is fast and could increase the prognostic value of these scans, thereby improving management of the condition, as well as elucidating the disease mechanisms of HIE.

Ridge Detection and Analysis of Susceptibility-Weighted Magnetic Resonance Imaging in Neonatal Hypoxic-Ischaemic Encephalopathy

“…Susceptibility-weighted images are very sensitive to the detection of vascular extraneous blood products and hypoxic ischemia [11]. We used an active contour model to segment the brain and to remove the background from the SW images to enable us to focus more on the blood vessels in brain [14] as shown in Figure 2(b).…”

“…Ridge Eigenvalues. Ridge point is the point where the intensity image has a local minimum in the direction where the gradient changes the most [11]. The second derivative information can be derived from the Hessian matrix for the local intensities in the neighborhood of a pixel on the ridge.…”

“…In [10], a machine learning tool is developed to predict outcomes for neonatal HIE patients by building their Apparent Diffusion Coefficient (ADC) map measurements. In [11], the set of Hessian eigenvalues of the ridges in the SWIs is used to correctly diagnose infants for HIE. They were divided into two categories based on the radiological report of the scan [12] -10 healthy infants and 48 infants with HIE, and the best classification accuracy was reported in [11] to be 91.83%.…”

“…In [11], the set of Hessian eigenvalues of the ridges in the SWIs is used to correctly diagnose infants for HIE. They were divided into two categories based on the radiological report of the scan [12] -10 healthy infants and 48 infants with HIE, and the best classification accuracy was reported in [11] to be 91.83%. In [13], an extended three-dimensional local binary patterns (LBP) is developed to distinguish the oxygenated images of three-dimensional SW images of infants with HIE, thought to be related to the oxygenation levels of the blood -a possible marker of hypoxic ischemic injury.…”

“…The best classification accuracy reported in [13] is 89.9%. However, there are two issues related to the classification accuracies presented in [11] [13]. 1) The classification accuracies are calculated in [11] and [13] by using Kitamura scores evaluated by radiologists.…”