Active Learning Pipeline for Brain Mapping in a High Performance Computing Environment

Adam Michaleas; Lars A. Gjesteby; Michael Snyder; David Chavez; Meagan Ash; Matthew A. Melton; Damon G. Lamb; Sara N. Burke; Kevin J. Otto; Lee Kamentsky; Webster Guan; Kwanghun Chung; Laura J. Brattain

doi:10.1109/HPEC43674.2020.9286225

Active Learning Pipeline for Brain Mapping in a High Performance Computing Environment

Adam Michaleas
, Lars A. Gjesteby
, Michael Snyder
, David Chavez
, Meagan Ash
, Matthew A. Melton
, Damon G. Lamb
, Sara N. Burke
, Kevin J. Otto
, Lee Kamentsky
, Webster Guan
, Kwanghun Chung
, Laura J. Brattain

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

7 Scopus citations

Abstract

This paper describes a scalable active learning pipeline prototype for large-scale brain mapping that leverages high performance computing power. It enables high-throughput evaluation of algorithm results, which, after human review, are used for iterative machine learning model training. Image processing and machine learning are performed in a batch layer. Benchmark testing of image processing using pMATLAB shows that a 100x increase in throughput (10,000%) can be achieved while total processing time only increases by 9% on Xeon-G6 CPUs and by 22% on Xeon-E5 CPUs, indicating robust scalability. The images and algorithm results are provided through a serving layer to a browser-based user interface for interactive review. This pipeline has the potential to greatly reduce the manual annotation burden and improve the overall performance of machine learning-based brain mapping.

Original language	English (US)
Title of host publication	2020 IEEE High Performance Extreme Computing Conference, HPEC 2020
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781728192192
DOIs	https://doi.org/10.1109/HPEC43674.2020.9286225
State	Published - Sep 22 2020
Externally published	Yes
Event	2020 IEEE High Performance Extreme Computing Conference, HPEC 2020 - Virtual, Waltham, United States Duration: Sep 21 2020 → Sep 25 2020

Publication series

Name	2020 IEEE High Performance Extreme Computing Conference, HPEC 2020

Conference

Conference	2020 IEEE High Performance Extreme Computing Conference, HPEC 2020
Country/Territory	United States
City	Virtual, Waltham
Period	9/21/20 → 9/25/20

Keywords

Active learning
axon tracing
brain mapping
high performance computing
neuron segmentation

ASJC Scopus subject areas

Artificial Intelligence
Computational Theory and Mathematics
Computer Networks and Communications
Computer Science Applications
Hardware and Architecture

Access to Document

10.1109/HPEC43674.2020.9286225

Cite this

Michaleas, A., Gjesteby, L. A., Snyder, M., Chavez, D., Ash, M., Melton, M. A., Lamb, D. G., Burke, S. N., Otto, K. J., Kamentsky, L., Guan, W., Chung, K., & Brattain, L. J. (2020). Active Learning Pipeline for Brain Mapping in a High Performance Computing Environment. In 2020 IEEE High Performance Extreme Computing Conference, HPEC 2020 Article 9286225 (2020 IEEE High Performance Extreme Computing Conference, HPEC 2020). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/HPEC43674.2020.9286225

Active Learning Pipeline for Brain Mapping in a High Performance Computing Environment. / Michaleas, Adam; Gjesteby, Lars A.; Snyder, Michael et al.
2020 IEEE High Performance Extreme Computing Conference, HPEC 2020. Institute of Electrical and Electronics Engineers Inc., 2020. 9286225 (2020 IEEE High Performance Extreme Computing Conference, HPEC 2020).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Michaleas, A, Gjesteby, LA, Snyder, M, Chavez, D, Ash, M, Melton, MA, Lamb, DG, Burke, SN, Otto, KJ, Kamentsky, L, Guan, W, Chung, K & Brattain, LJ 2020, Active Learning Pipeline for Brain Mapping in a High Performance Computing Environment. in 2020 IEEE High Performance Extreme Computing Conference, HPEC 2020., 9286225, 2020 IEEE High Performance Extreme Computing Conference, HPEC 2020, Institute of Electrical and Electronics Engineers Inc., 2020 IEEE High Performance Extreme Computing Conference, HPEC 2020, Virtual, Waltham, United States, 9/21/20. https://doi.org/10.1109/HPEC43674.2020.9286225

Michaleas A, Gjesteby LA, Snyder M, Chavez D, Ash M, Melton MA et al. Active Learning Pipeline for Brain Mapping in a High Performance Computing Environment. In 2020 IEEE High Performance Extreme Computing Conference, HPEC 2020. Institute of Electrical and Electronics Engineers Inc. 2020. 9286225. (2020 IEEE High Performance Extreme Computing Conference, HPEC 2020). doi: 10.1109/HPEC43674.2020.9286225

@inproceedings{0f0b543fb90f470d90b7faeb18d223cd,

title = "Active Learning Pipeline for Brain Mapping in a High Performance Computing Environment",

abstract = "This paper describes a scalable active learning pipeline prototype for large-scale brain mapping that leverages high performance computing power. It enables high-throughput evaluation of algorithm results, which, after human review, are used for iterative machine learning model training. Image processing and machine learning are performed in a batch layer. Benchmark testing of image processing using pMATLAB shows that a 100x increase in throughput (10,000\%) can be achieved while total processing time only increases by 9\% on Xeon-G6 CPUs and by 22\% on Xeon-E5 CPUs, indicating robust scalability. The images and algorithm results are provided through a serving layer to a browser-based user interface for interactive review. This pipeline has the potential to greatly reduce the manual annotation burden and improve the overall performance of machine learning-based brain mapping.",

keywords = "Active learning, axon tracing, brain mapping, high performance computing, neuron segmentation",

author = "Adam Michaleas and Gjesteby, \{Lars A.\} and Michael Snyder and David Chavez and Meagan Ash and Melton, \{Matthew A.\} and Lamb, \{Damon G.\} and Burke, \{Sara N.\} and Otto, \{Kevin J.\} and Lee Kamentsky and Webster Guan and Kwanghun Chung and Brattain, \{Laura J.\}",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 2020 IEEE High Performance Extreme Computing Conference, HPEC 2020 ; Conference date: 21-09-2020 Through 25-09-2020",

year = "2020",

month = sep,

day = "22",

doi = "10.1109/HPEC43674.2020.9286225",

language = "English (US)",

series = "2020 IEEE High Performance Extreme Computing Conference, HPEC 2020",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2020 IEEE High Performance Extreme Computing Conference, HPEC 2020",

}

TY - GEN

T1 - Active Learning Pipeline for Brain Mapping in a High Performance Computing Environment

AU - Michaleas, Adam

AU - Gjesteby, Lars A.

AU - Snyder, Michael

AU - Chavez, David

AU - Ash, Meagan

AU - Melton, Matthew A.

AU - Lamb, Damon G.

AU - Burke, Sara N.

AU - Otto, Kevin J.

AU - Kamentsky, Lee

AU - Guan, Webster

AU - Chung, Kwanghun

AU - Brattain, Laura J.

PY - 2020/9/22

Y1 - 2020/9/22

N2 - This paper describes a scalable active learning pipeline prototype for large-scale brain mapping that leverages high performance computing power. It enables high-throughput evaluation of algorithm results, which, after human review, are used for iterative machine learning model training. Image processing and machine learning are performed in a batch layer. Benchmark testing of image processing using pMATLAB shows that a 100x increase in throughput (10,000%) can be achieved while total processing time only increases by 9% on Xeon-G6 CPUs and by 22% on Xeon-E5 CPUs, indicating robust scalability. The images and algorithm results are provided through a serving layer to a browser-based user interface for interactive review. This pipeline has the potential to greatly reduce the manual annotation burden and improve the overall performance of machine learning-based brain mapping.

AB - This paper describes a scalable active learning pipeline prototype for large-scale brain mapping that leverages high performance computing power. It enables high-throughput evaluation of algorithm results, which, after human review, are used for iterative machine learning model training. Image processing and machine learning are performed in a batch layer. Benchmark testing of image processing using pMATLAB shows that a 100x increase in throughput (10,000%) can be achieved while total processing time only increases by 9% on Xeon-G6 CPUs and by 22% on Xeon-E5 CPUs, indicating robust scalability. The images and algorithm results are provided through a serving layer to a browser-based user interface for interactive review. This pipeline has the potential to greatly reduce the manual annotation burden and improve the overall performance of machine learning-based brain mapping.

KW - Active learning

KW - axon tracing

KW - brain mapping

KW - high performance computing

KW - neuron segmentation

UR - https://www.scopus.com/pages/publications/85099356409

UR - https://www.scopus.com/pages/publications/85099356409#tab=citedBy

U2 - 10.1109/HPEC43674.2020.9286225

DO - 10.1109/HPEC43674.2020.9286225

M3 - Conference contribution

AN - SCOPUS:85099356409

T3 - 2020 IEEE High Performance Extreme Computing Conference, HPEC 2020

BT - 2020 IEEE High Performance Extreme Computing Conference, HPEC 2020

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2020 IEEE High Performance Extreme Computing Conference, HPEC 2020

Y2 - 21 September 2020 through 25 September 2020

ER -

Active Learning Pipeline for Brain Mapping in a High Performance Computing Environment

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this