Segmentation of the Cerebellar Peduncles Using a Random Forest Classifier and a Multi-object Geometric Deformable Model: Application to Spinocerebellar Ataxia Type 6

Chuyang Ye, Zhen Yang, Sarah H. Ying, Jerry Ladd Prince

Research output: Contribution to journalArticle

Abstract

The cerebellar peduncles, comprising the superior cerebellar peduncles (SCPs), the middle cerebellar peduncle (MCP), and the inferior cerebellar peduncles (ICPs), are white matter tracts that connect the cerebellum to other parts of the central nervous system. Methods for automatic segmentation and quantification of the cerebellar peduncles are needed for objectively and efficiently studying their structure and function. Diffusion tensor imaging (DTI) provides key information to support this goal, but it remains challenging because the tensors change dramatically in the decussation of the SCPs (dSCP), the region where the SCPs cross. This paper presents an automatic method for segmenting the cerebellar peduncles, including the dSCP. The method uses volumetric segmentation concepts based on extracted DTI features. The dSCP and noncrossing portions of the peduncles are modeled as separate objects, and are initially classified using a random forest classifier together with the DTI features. To obtain geometrically correct results, a multi-object geometric deformable model is used to refine the random forest classification. The method was evaluated using a leave-one-out cross-validation on five control subjects and four patients with spinocerebellar ataxia type 6 (SCA6). It was then used to evaluate group differences in the peduncles in a population of 32 controls and 11 SCA6 patients. In the SCA6 group, we have observed significant decreases in the volumes of the dSCP and the ICPs and significant increases in the mean diffusivity in the noncrossing SCPs, the MCP, and the ICPs. These results are consistent with a degeneration of the cerebellar peduncles in SCA6 patients.

Original languageEnglish (US)
Pages (from-to)367-381
Number of pages15
JournalNeuroinformatics
Volume13
Issue number3
DOIs
StatePublished - Mar 7 2015

Fingerprint

Diffusion tensor imaging
Spinocerebellar Ataxias
Diffusion Tensor Imaging
Classifiers
Neurology
Tensors
Population Control
Cerebellum
Central Nervous System
Middle Cerebellar Peduncle

Keywords

  • Cerebellar peduncles
  • Multi-object geometric deformable model
  • Random forest classifier
  • Spinocerebellar ataxia type 6

ASJC Scopus subject areas

  • Neuroscience(all)
  • Information Systems
  • Software

Cite this

Segmentation of the Cerebellar Peduncles Using a Random Forest Classifier and a Multi-object Geometric Deformable Model : Application to Spinocerebellar Ataxia Type 6. / Ye, Chuyang; Yang, Zhen; Ying, Sarah H.; Prince, Jerry Ladd.

In: Neuroinformatics, Vol. 13, No. 3, 07.03.2015, p. 367-381.

Research output: Contribution to journalArticle

@article{a11d1585b1ca4b90af8dd72f7d904f15,
title = "Segmentation of the Cerebellar Peduncles Using a Random Forest Classifier and a Multi-object Geometric Deformable Model: Application to Spinocerebellar Ataxia Type 6",
abstract = "The cerebellar peduncles, comprising the superior cerebellar peduncles (SCPs), the middle cerebellar peduncle (MCP), and the inferior cerebellar peduncles (ICPs), are white matter tracts that connect the cerebellum to other parts of the central nervous system. Methods for automatic segmentation and quantification of the cerebellar peduncles are needed for objectively and efficiently studying their structure and function. Diffusion tensor imaging (DTI) provides key information to support this goal, but it remains challenging because the tensors change dramatically in the decussation of the SCPs (dSCP), the region where the SCPs cross. This paper presents an automatic method for segmenting the cerebellar peduncles, including the dSCP. The method uses volumetric segmentation concepts based on extracted DTI features. The dSCP and noncrossing portions of the peduncles are modeled as separate objects, and are initially classified using a random forest classifier together with the DTI features. To obtain geometrically correct results, a multi-object geometric deformable model is used to refine the random forest classification. The method was evaluated using a leave-one-out cross-validation on five control subjects and four patients with spinocerebellar ataxia type 6 (SCA6). It was then used to evaluate group differences in the peduncles in a population of 32 controls and 11 SCA6 patients. In the SCA6 group, we have observed significant decreases in the volumes of the dSCP and the ICPs and significant increases in the mean diffusivity in the noncrossing SCPs, the MCP, and the ICPs. These results are consistent with a degeneration of the cerebellar peduncles in SCA6 patients.",
keywords = "Cerebellar peduncles, Multi-object geometric deformable model, Random forest classifier, Spinocerebellar ataxia type 6",
author = "Chuyang Ye and Zhen Yang and Ying, {Sarah H.} and Prince, {Jerry Ladd}",
year = "2015",
month = "3",
day = "7",
doi = "10.1007/s12021-015-9264-7",
language = "English (US)",
volume = "13",
pages = "367--381",
journal = "Neuroinformatics",
issn = "1539-2791",
publisher = "Humana Press",
number = "3",

}

TY - JOUR

T1 - Segmentation of the Cerebellar Peduncles Using a Random Forest Classifier and a Multi-object Geometric Deformable Model

T2 - Application to Spinocerebellar Ataxia Type 6

AU - Ye, Chuyang

AU - Yang, Zhen

AU - Ying, Sarah H.

AU - Prince, Jerry Ladd

PY - 2015/3/7

Y1 - 2015/3/7

N2 - The cerebellar peduncles, comprising the superior cerebellar peduncles (SCPs), the middle cerebellar peduncle (MCP), and the inferior cerebellar peduncles (ICPs), are white matter tracts that connect the cerebellum to other parts of the central nervous system. Methods for automatic segmentation and quantification of the cerebellar peduncles are needed for objectively and efficiently studying their structure and function. Diffusion tensor imaging (DTI) provides key information to support this goal, but it remains challenging because the tensors change dramatically in the decussation of the SCPs (dSCP), the region where the SCPs cross. This paper presents an automatic method for segmenting the cerebellar peduncles, including the dSCP. The method uses volumetric segmentation concepts based on extracted DTI features. The dSCP and noncrossing portions of the peduncles are modeled as separate objects, and are initially classified using a random forest classifier together with the DTI features. To obtain geometrically correct results, a multi-object geometric deformable model is used to refine the random forest classification. The method was evaluated using a leave-one-out cross-validation on five control subjects and four patients with spinocerebellar ataxia type 6 (SCA6). It was then used to evaluate group differences in the peduncles in a population of 32 controls and 11 SCA6 patients. In the SCA6 group, we have observed significant decreases in the volumes of the dSCP and the ICPs and significant increases in the mean diffusivity in the noncrossing SCPs, the MCP, and the ICPs. These results are consistent with a degeneration of the cerebellar peduncles in SCA6 patients.

AB - The cerebellar peduncles, comprising the superior cerebellar peduncles (SCPs), the middle cerebellar peduncle (MCP), and the inferior cerebellar peduncles (ICPs), are white matter tracts that connect the cerebellum to other parts of the central nervous system. Methods for automatic segmentation and quantification of the cerebellar peduncles are needed for objectively and efficiently studying their structure and function. Diffusion tensor imaging (DTI) provides key information to support this goal, but it remains challenging because the tensors change dramatically in the decussation of the SCPs (dSCP), the region where the SCPs cross. This paper presents an automatic method for segmenting the cerebellar peduncles, including the dSCP. The method uses volumetric segmentation concepts based on extracted DTI features. The dSCP and noncrossing portions of the peduncles are modeled as separate objects, and are initially classified using a random forest classifier together with the DTI features. To obtain geometrically correct results, a multi-object geometric deformable model is used to refine the random forest classification. The method was evaluated using a leave-one-out cross-validation on five control subjects and four patients with spinocerebellar ataxia type 6 (SCA6). It was then used to evaluate group differences in the peduncles in a population of 32 controls and 11 SCA6 patients. In the SCA6 group, we have observed significant decreases in the volumes of the dSCP and the ICPs and significant increases in the mean diffusivity in the noncrossing SCPs, the MCP, and the ICPs. These results are consistent with a degeneration of the cerebellar peduncles in SCA6 patients.

KW - Cerebellar peduncles

KW - Multi-object geometric deformable model

KW - Random forest classifier

KW - Spinocerebellar ataxia type 6

UR - http://www.scopus.com/inward/record.url?scp=84931008462&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84931008462&partnerID=8YFLogxK

U2 - 10.1007/s12021-015-9264-7

DO - 10.1007/s12021-015-9264-7

M3 - Article

C2 - 25749985

AN - SCOPUS:84931008462

VL - 13

SP - 367

EP - 381

JO - Neuroinformatics

JF - Neuroinformatics

SN - 1539-2791

IS - 3

ER -