Abstract

Objective and Impact Statement. We propose an automated method of predicting Normal Pressure Hydrocephalus (NPH) from CT scans. A deep convolutional network segments regions of interest from the scans. These regions are then combined with MRI information to predict NPH. To our knowledge, this is the first method which automatically predicts NPH from CT scans and incorporates diffusion tractography information for prediction. Introduction. Due to their low cost and high versatility, CT scans are often used in NPH diagnosis. No well-defined and effective protocol currently exists for analysis of CT scans for NPH. Evans' index, an approximation of the ventricle to brain volume using one 2D image slice, has been proposed but is not robust. The proposed approach is an effective way to quantify regions of interest and offers a computational method for predicting NPH. Methods. We propose a novel method to predict NPH by combining regions of interest segmented from CT scans with connectome data to compute features which capture the impact of enlarged ventricles by excluding fiber tracts passing through these regions. The segmentation and network features are used to train a model for NPH prediction. Results. Our method outperforms the current state-of-the-art by 9 precision points and 29 recall points. Our segmentation model outperforms the current state-of-the-art in segmenting the ventricle, gray-white matter, and subarachnoid space in CT scans. Conclusion. Our experimental results demonstrate that fast and accurate volumetric segmentation of CT brain scans can help improve the NPH diagnosis process, and network properties can increase NPH prediction accuracy.

Highlights

  • This paper describes a novel and robust method for segmenting Computed Tomography (CT) scans of the brain and predicting possible Normal Pressure Hydrocephalus (NPH) from the segmentation

  • (1) We have developed a robust, data-driven CT segmentation method which is derived from data from NPH and age-matched normal patients which can aid in NPH diagnosis

  • A 90 × 90 connectivity matrix computed using the Automated Anatomical Labeling 2 (AAL2) atlas (Tzourio-Mazoyer et al [21]) in MNI152 space is constructed for each set of fibers, and the network properties of the matrix are used as additional features to a shallow, fully connected network which uses the features from the last downsampling layer of the UNet to differentiate NPH and non-NPH subjects

Read more

Summary

Introduction

This paper describes a novel and robust method for segmenting Computed Tomography (CT) scans of the brain and predicting possible Normal Pressure Hydrocephalus (NPH) from the segmentation. Our proposed method is the first to predict NPH by first fully segmenting the subarachnoid space, lateral ventricles, and gray-white matter of the brain from a CT scan. (2) we combine diffusion tractography information with the segmentation in a novel way to improve NPH prediction (3) The creation of a novel and valuable dataset of detailed and full manual segmentations of CT brain scans This data will be made publicly available and can further benefit the medical research community. To incorporate additional white matter structural (diffusion tractography) information into the NPH prediction process, diffusion MRI of normal subjects between the ages of 75 and 85 are collected from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). In addition to the CT data, diffusion MRI of 11 normal subjects between the ages of 75 and 85 from the Alzheimer’s Disease Neuroimaging Initia-

Approach
Diffusion Tractography Analysis
NPH Prediction Module
Experiments and Results
Method
Discussion and Future
Conflicts of Interest
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call