Abstract
Aberrant development of the human brain during the first year after birth is known to cause critical implications in later stages of life. In particular, neuropsychiatric disorders, such as attention deficit hyperactivity disorder (ADHD), have been linked with abnormal early development of the hippocampus. Despite its known importance, studying the hippocampus in infant subjects is very challenging due to the significantly smaller brain size, dynamically varying image contrast, and large across-subject variation. In this paper, we present a novel method for effective hippocampus segmentation by using a multi- atlas approach that integrates the complementary multimodal information from longitudinal T1 and T2 MR images. In particular, considering the highly heterogeneous nature of the longitudinal data, we propose to learn their common feature representations by using hierarchical multi-set kernel canonical correlation analysis (CCA). Specifically, we will learn (1) within-time-point common features by projecting different modality features of each time point to its own modality-free common space, and (2) across-time-point common features by mapping all time-point-specific common features to a global common space for all time points. These final features are then employed in patch matching across different modalities and time points for hippocampus segmentation, via label propagation and fusion. Experimental results demonstrate the improved performance of our method over the state-of-the-art methods.
| Original language | English |
|---|---|
| Title of host publication | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
| Publisher | Springer Verlag |
| Pages | 63-71 |
| Number of pages | 9 |
| Volume | 9351 |
| ISBN (Print) | 9783319245737 |
| DOIs | |
| Publication status | Published - 2015 |
| Externally published | Yes |
| Event | 18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015 - Munich, Germany Duration: 2015 Oct 5 → 2015 Oct 9 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 9351 |
| ISSN (Print) | 03029743 |
| ISSN (Electronic) | 16113349 |
Other
| Other | 18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015 |
|---|---|
| Country | Germany |
| City | Munich |
| Period | 15/10/5 → 15/10/9 |
Fingerprint
ASJC Scopus subject areas
- Computer Science(all)
- Theoretical Computer Science
Cite this
Segmentation of infant hippocampus using common feature representations learned for multimodal longitudinal data. / Guo, Yanrong; Wu, Guorong; Yap, Pew Thian; Jewells, Valerie; Lin, Weili; Shen, Dinggang.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9351 Springer Verlag, 2015. p. 63-71 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9351).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Segmentation of infant hippocampus using common feature representations learned for multimodal longitudinal data
AU - Guo, Yanrong
AU - Wu, Guorong
AU - Yap, Pew Thian
AU - Jewells, Valerie
AU - Lin, Weili
AU - Shen, Dinggang
PY - 2015
Y1 - 2015
N2 - Aberrant development of the human brain during the first year after birth is known to cause critical implications in later stages of life. In particular, neuropsychiatric disorders, such as attention deficit hyperactivity disorder (ADHD), have been linked with abnormal early development of the hippocampus. Despite its known importance, studying the hippocampus in infant subjects is very challenging due to the significantly smaller brain size, dynamically varying image contrast, and large across-subject variation. In this paper, we present a novel method for effective hippocampus segmentation by using a multi- atlas approach that integrates the complementary multimodal information from longitudinal T1 and T2 MR images. In particular, considering the highly heterogeneous nature of the longitudinal data, we propose to learn their common feature representations by using hierarchical multi-set kernel canonical correlation analysis (CCA). Specifically, we will learn (1) within-time-point common features by projecting different modality features of each time point to its own modality-free common space, and (2) across-time-point common features by mapping all time-point-specific common features to a global common space for all time points. These final features are then employed in patch matching across different modalities and time points for hippocampus segmentation, via label propagation and fusion. Experimental results demonstrate the improved performance of our method over the state-of-the-art methods.
AB - Aberrant development of the human brain during the first year after birth is known to cause critical implications in later stages of life. In particular, neuropsychiatric disorders, such as attention deficit hyperactivity disorder (ADHD), have been linked with abnormal early development of the hippocampus. Despite its known importance, studying the hippocampus in infant subjects is very challenging due to the significantly smaller brain size, dynamically varying image contrast, and large across-subject variation. In this paper, we present a novel method for effective hippocampus segmentation by using a multi- atlas approach that integrates the complementary multimodal information from longitudinal T1 and T2 MR images. In particular, considering the highly heterogeneous nature of the longitudinal data, we propose to learn their common feature representations by using hierarchical multi-set kernel canonical correlation analysis (CCA). Specifically, we will learn (1) within-time-point common features by projecting different modality features of each time point to its own modality-free common space, and (2) across-time-point common features by mapping all time-point-specific common features to a global common space for all time points. These final features are then employed in patch matching across different modalities and time points for hippocampus segmentation, via label propagation and fusion. Experimental results demonstrate the improved performance of our method over the state-of-the-art methods.
UR - http://www.scopus.com/inward/record.url?scp=84951793320&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84951793320&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-24574-4_8
DO - 10.1007/978-3-319-24574-4_8
M3 - Conference contribution
SN - 9783319245737
VL - 9351
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 63
EP - 71
BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PB - Springer Verlag
ER -