PMID- 31602089
OWN - NLM
STAT- PubMed-not-MEDLINE
LR  - 20200928
IS  - 0277-786X (Print)
IS  - 1996-756X (Electronic)
IS  - 0277-786X (Linking)
VI  - 10949
DP  - 2019 Mar
TI  - Distributed deep learning for robust multi-site segmentation of CT imaging after 
      traumatic brain injury.
LID - 109490A [pii]
LID - 10.1117/12.2511997 [doi]
AB  - Machine learning models are becoming commonplace in the domain of medical 
      imaging, and with these methods comes an ever-increasing need for more data. 
      However, to preserve patient anonymity it is frequently impractical or prohibited 
      to transfer protected health information (PHI) between institutions. 
      Additionally, due to the nature of some studies, there may not be a large public 
      dataset available on which to train models. To address this conundrum, we analyze 
      the efficacy of transferring the model itself in lieu of data between different 
      sites. By doing so we accomplish two goals: 1) the model gains access to training 
      on a larger dataset that it could not normally obtain and 2) the model better 
      generalizes, having trained on data from separate locations. In this paper, we 
      implement multi-site learning with disparate datasets from the National 
      Institutes of Health (NIH) and Vanderbilt University Medical Center (VUMC) 
      without compromising PHI. Three neural networks are trained to convergence on a 
      computed tomography (CT) brain hematoma segmentation task: one only with NIH 
      data, one only with VUMC data, and one multi-site model alternating between NIH 
      and VUMC data. Resultant lesion masks with the multi-site model attain an average 
      Dice similarity coefficient of 0.64 and the automatically segmented hematoma 
      volumes correlate to those done manually with a Pearson correlation coefficient 
      of 0.87, corresponding to an 8% and 5% improvement, respectively, over the 
      single-site model counterparts.
FAU - Remedios, Samuel
AU  - Remedios S
AD  - Center for Neuroscience and Regenerative Medicine, Henry Jackson Foundation.
AD  - Radiology and Imaging Sciences, Clinical Center, National Institute of Health.
AD  - Department of Computer Science, Middle Tennessee State University.
AD  - Department of Electrical Engineering, Vanderbilt University.
FAU - Roy, Snehashis
AU  - Roy S
AD  - Center for Neuroscience and Regenerative Medicine, Henry Jackson Foundation.
AD  - Radiology and Imaging Sciences, Clinical Center, National Institute of Health.
FAU - Blaber, Justin
AU  - Blaber J
AD  - Department of Electrical Engineering, Vanderbilt University.
FAU - Bermudez, Camilo
AU  - Bermudez C
AD  - Department of Biomedical Engineering, Vanderbilt University.
FAU - Nath, Vishwesh
AU  - Nath V
AD  - Department of Computer Science, Vanderbilt University.
FAU - Patel, Mayur B
AU  - Patel MB
AD  - Departments of Surgery, Neurosurgery, Hearing & Speech Sciences; Center for 
      Health Services Research, Vanderbilt Brain Institute; Critical Illness, Brain 
      Dysfunction, and Survivorship Center, Vanderbilt University Medical Center; VA 
      Tennessee Valley Healthcare System, Department of Veterans Affairs Medical 
      Center.
FAU - Butman, John A
AU  - Butman JA
AD  - Radiology and Imaging Sciences, Clinical Center, National Institute of Health.
FAU - Landman, Bennett A
AU  - Landman BA
AD  - Department of Electrical Engineering, Vanderbilt University.
AD  - Department of Biomedical Engineering, Vanderbilt University.
AD  - Department of Computer Science, Vanderbilt University.
FAU - Pham, Dzung L
AU  - Pham DL
AD  - Center for Neuroscience and Regenerative Medicine, Henry Jackson Foundation.
AD  - Radiology and Imaging Sciences, Clinical Center, National Institute of Health.
LA  - eng
GR  - R01 EB017230/EB/NIBIB NIH HHS/United States
GR  - R01 GM120484/GM/NIGMS NIH HHS/United States
PT  - Journal Article
PL  - United States
TA  - Proc SPIE Int Soc Opt Eng
JT  - Proceedings of SPIE--the International Society for Optical Engineering
JID - 101524122
PMC - PMC6786776
MID - NIHMS1009837
OTO - NOTNLM
OT  - computed tomography (CT)
OT  - deep learning
OT  - distributed
OT  - hematoma
OT  - lesion
OT  - multi-site
OT  - neural network
OT  - segmentation
EDAT- 2019/10/12 06:00
MHDA- 2019/10/12 06:01
PMCR- 2019/10/10
CRDT- 2019/10/12 06:00
PHST- 2019/10/12 06:00 [entrez]
PHST- 2019/10/12 06:00 [pubmed]
PHST- 2019/10/12 06:01 [medline]
PHST- 2019/10/10 00:00 [pmc-release]
AID - 109490A [pii]
AID - 10.1117/12.2511997 [doi]
PST - ppublish
SO  - Proc SPIE Int Soc Opt Eng. 2019 Mar;10949:109490A. doi: 10.1117/12.2511997.