Learning based segmentation of CT brain images: Application to postoperative hydrocephalic scans

Venkateswararao Cherukuri, Peter Ssenyonga, Benjamin C. Warf, Abhaya V. Kulkarni, Vishal Monga, Steven J. Schiff

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Objective: Hydrocephalus is a medical condition in which there is an abnormal accumulation of cerebrospinal fluid (CSF) in the brain. Segmentation of brain imagery into brain tissue and CSF [before and after surgery, i.e., preoperative (pre-op) versus postoperative (post-op)] plays a crucial role in evaluating surgical treatment. Segmentation of pre-op images is often a relatively straightforward problem and has been well researched. However, segmenting post-op computational tomographic (CT) scans becomes more challenging due to distorted anatomy and subdural hematoma collections pressing on the brain. Most intensity- and feature-based segmentation methods fail to separate subdurals from brain and CSF as subdural geometry varies greatly across different patients and their intensity varies with time. We combat this problem by a learning approach that treats segmentation as supervised classification at the pixel level, i.e., a training set of CT scans with labeled pixel identities is employed. Methods: Our contributions include: 1) a dictionary learning framework that learns class (segment) specific dictionaries that can efficiently represent test samples from the same class while poorly represent corresponding samples from other classes; 2) quantification of associated computation and memory footprint; and 3) a customized training and test procedure for segmenting post-op hydrocephalic CT images. Results: Experiments performed on infant CT brain images acquired from the CURE Children's Hospital of Uganda reveal the success of our method against the state-of-the-art alternatives. We also demonstrate that the proposed algorithm is computationally less burdensome and exhibits a graceful degradation against a number of training samples, enhancing its deployment potential.

Original languageEnglish (US)
Pages (from-to)1871-1884
Number of pages14
JournalIEEE Transactions on Biomedical Engineering
Volume65
Issue number8
DOIs
StatePublished - Aug 2018

Fingerprint

Brain
Cerebrospinal fluid
Glossaries
Pixels
Surgery
Tissue
Data storage equipment
Degradation
Geometry
Experiments

All Science Journal Classification (ASJC) codes

  • Biomedical Engineering

Cite this

Cherukuri, Venkateswararao ; Ssenyonga, Peter ; Warf, Benjamin C. ; Kulkarni, Abhaya V. ; Monga, Vishal ; Schiff, Steven J. / Learning based segmentation of CT brain images : Application to postoperative hydrocephalic scans. In: IEEE Transactions on Biomedical Engineering. 2018 ; Vol. 65, No. 8. pp. 1871-1884.
@article{dcf1e515a7b84d469f2bd868e1c40446,
title = "Learning based segmentation of CT brain images: Application to postoperative hydrocephalic scans",
abstract = "Objective: Hydrocephalus is a medical condition in which there is an abnormal accumulation of cerebrospinal fluid (CSF) in the brain. Segmentation of brain imagery into brain tissue and CSF [before and after surgery, i.e., preoperative (pre-op) versus postoperative (post-op)] plays a crucial role in evaluating surgical treatment. Segmentation of pre-op images is often a relatively straightforward problem and has been well researched. However, segmenting post-op computational tomographic (CT) scans becomes more challenging due to distorted anatomy and subdural hematoma collections pressing on the brain. Most intensity- and feature-based segmentation methods fail to separate subdurals from brain and CSF as subdural geometry varies greatly across different patients and their intensity varies with time. We combat this problem by a learning approach that treats segmentation as supervised classification at the pixel level, i.e., a training set of CT scans with labeled pixel identities is employed. Methods: Our contributions include: 1) a dictionary learning framework that learns class (segment) specific dictionaries that can efficiently represent test samples from the same class while poorly represent corresponding samples from other classes; 2) quantification of associated computation and memory footprint; and 3) a customized training and test procedure for segmenting post-op hydrocephalic CT images. Results: Experiments performed on infant CT brain images acquired from the CURE Children's Hospital of Uganda reveal the success of our method against the state-of-the-art alternatives. We also demonstrate that the proposed algorithm is computationally less burdensome and exhibits a graceful degradation against a number of training samples, enhancing its deployment potential.",
author = "Venkateswararao Cherukuri and Peter Ssenyonga and Warf, {Benjamin C.} and Kulkarni, {Abhaya V.} and Vishal Monga and Schiff, {Steven J.}",
year = "2018",
month = "8",
doi = "10.1109/TBME.2017.2783305",
language = "English (US)",
volume = "65",
pages = "1871--1884",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "8",

}

Learning based segmentation of CT brain images : Application to postoperative hydrocephalic scans. / Cherukuri, Venkateswararao; Ssenyonga, Peter; Warf, Benjamin C.; Kulkarni, Abhaya V.; Monga, Vishal; Schiff, Steven J.

In: IEEE Transactions on Biomedical Engineering, Vol. 65, No. 8, 08.2018, p. 1871-1884.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Learning based segmentation of CT brain images

T2 - Application to postoperative hydrocephalic scans

AU - Cherukuri, Venkateswararao

AU - Ssenyonga, Peter

AU - Warf, Benjamin C.

AU - Kulkarni, Abhaya V.

AU - Monga, Vishal

AU - Schiff, Steven J.

PY - 2018/8

Y1 - 2018/8

N2 - Objective: Hydrocephalus is a medical condition in which there is an abnormal accumulation of cerebrospinal fluid (CSF) in the brain. Segmentation of brain imagery into brain tissue and CSF [before and after surgery, i.e., preoperative (pre-op) versus postoperative (post-op)] plays a crucial role in evaluating surgical treatment. Segmentation of pre-op images is often a relatively straightforward problem and has been well researched. However, segmenting post-op computational tomographic (CT) scans becomes more challenging due to distorted anatomy and subdural hematoma collections pressing on the brain. Most intensity- and feature-based segmentation methods fail to separate subdurals from brain and CSF as subdural geometry varies greatly across different patients and their intensity varies with time. We combat this problem by a learning approach that treats segmentation as supervised classification at the pixel level, i.e., a training set of CT scans with labeled pixel identities is employed. Methods: Our contributions include: 1) a dictionary learning framework that learns class (segment) specific dictionaries that can efficiently represent test samples from the same class while poorly represent corresponding samples from other classes; 2) quantification of associated computation and memory footprint; and 3) a customized training and test procedure for segmenting post-op hydrocephalic CT images. Results: Experiments performed on infant CT brain images acquired from the CURE Children's Hospital of Uganda reveal the success of our method against the state-of-the-art alternatives. We also demonstrate that the proposed algorithm is computationally less burdensome and exhibits a graceful degradation against a number of training samples, enhancing its deployment potential.

AB - Objective: Hydrocephalus is a medical condition in which there is an abnormal accumulation of cerebrospinal fluid (CSF) in the brain. Segmentation of brain imagery into brain tissue and CSF [before and after surgery, i.e., preoperative (pre-op) versus postoperative (post-op)] plays a crucial role in evaluating surgical treatment. Segmentation of pre-op images is often a relatively straightforward problem and has been well researched. However, segmenting post-op computational tomographic (CT) scans becomes more challenging due to distorted anatomy and subdural hematoma collections pressing on the brain. Most intensity- and feature-based segmentation methods fail to separate subdurals from brain and CSF as subdural geometry varies greatly across different patients and their intensity varies with time. We combat this problem by a learning approach that treats segmentation as supervised classification at the pixel level, i.e., a training set of CT scans with labeled pixel identities is employed. Methods: Our contributions include: 1) a dictionary learning framework that learns class (segment) specific dictionaries that can efficiently represent test samples from the same class while poorly represent corresponding samples from other classes; 2) quantification of associated computation and memory footprint; and 3) a customized training and test procedure for segmenting post-op hydrocephalic CT images. Results: Experiments performed on infant CT brain images acquired from the CURE Children's Hospital of Uganda reveal the success of our method against the state-of-the-art alternatives. We also demonstrate that the proposed algorithm is computationally less burdensome and exhibits a graceful degradation against a number of training samples, enhancing its deployment potential.

UR - http://www.scopus.com/inward/record.url?scp=85038828104&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85038828104&partnerID=8YFLogxK

U2 - 10.1109/TBME.2017.2783305

DO - 10.1109/TBME.2017.2783305

M3 - Article

C2 - 29989926

AN - SCOPUS:85038828104

VL - 65

SP - 1871

EP - 1884

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 8

ER -