Evaluating nnU-Net for early ischemic change segmentation on
non-contrast computed tomography in patients with Acute
Ischemic Stroke

J 2022

Evaluating nnU-Net for early ischemic change segmentation on non-contrast computed tomography in patients with Acute Ischemic Stroke

EL-HARIRI, Houssam, Luis A Souto Maior NETO, Petra CIMFLOVÁ, Fouzi BALA, Rotem GOLAN et. al.

Basic information

Original name

Evaluating nnU-Net for early ischemic change segmentation on non-contrast computed tomography in patients with Acute Ischemic Stroke

Authors

EL-HARIRI, Houssam (guarantor), Luis A Souto Maior NETO, Petra CIMFLOVÁ (203 Czech Republic, belonging to the institution), Fouzi BALA, Rotem GOLAN, Alireza SOJOUDI, Chris DUSZYNSKI, Ibukun ELEBUTE, Seyed Hossein MOUSAVI, Wu QIU and Bijoy K MENON

Edition

Computers in Biology and Medicine, OXFORD, PERGAMON-ELSEVIER SCIENCE LTD, 2022, 0010-4825

Other information

Language

English

Type of outcome

Článek v odborném periodiku

Field of Study

30224 Radiology, nuclear medicine and medical imaging

Country of publisher

United Kingdom of Great Britain and Northern Ireland

Confidentiality degree

není předmětem státního či obchodního tajemství

References:

URL

Impact factor

Impact factor: 7.700

RIV identification code

RIV/00216224:14110/22:00128306

Organization unit

Faculty of Medicine

DOI

http://dx.doi.org/10.1016/j.compbiomed.2021.105033

UT WoS

000788076100001

Keywords in English

Machine learning; Deep learning; Computer vision; Segmentation; Neurovascular imaging; Computed tomography; Acute ischemic stroke; Brain lesion

Abstract

V originále

Identifying the presence and extent of early ischemic changes (EIC) on Non-Contrast Computed Tomography (NCCT) is key to diagnosing and making time-sensitive treatment decisions in patients that present with Acute Ischemic Stroke (AIS). Segmenting EIC on NCCT is however a challenging task. In this study, we investigated a 3D CNN based on nnU-Net, a self-adapting CNN technique that has become the state-of-the-art in medical image segmentation, for segmenting EIC in NCCT of AIS patients. We trained and tested this model on a sizeable and heterogenous dataset of 534 patients, split into 438 for training and validation and 96 for testing. On this test set, we additionally assessed the inter-rater performance by comparing the proposed approach against two reference segmentation annotations by expert neuroradiologist readers, using this as the benchmark against which to compare our model. In terms of spatial agreement, we report median Dice Similarity Coefficients (DSCs) of 39.8% for the model vs. Reader-1, 39.4% for the model vs. Reader-2, and 55.6% for Reader-2 vs. Reader-1. In terms of lesion volume agreement, we report Intraclass Correlation Coefficients (ICCs) of 83.4% for model vs. Reader-1, 80.4% for model vs. Reader-2, and 94.8% for Reader-2 vs. Reader-1. Based on these results, we conclude that our model performs well relative to expert human performance and therefore may be useful as a decision-aid for clinicians.

Citovat

EL-HARIRI, Houssam, Luis A Souto Maior NETO, Petra CIMFLOVÁ, Fouzi BALA, Rotem GOLAN, Alireza SOJOUDI, Chris DUSZYNSKI, Ibukun ELEBUTE, Seyed Hossein MOUSAVI, Wu QIU and Bijoy K MENON. Evaluating nnU-Net for early ischemic change segmentation on non-contrast computed tomography in patients with Acute Ischemic Stroke. Computers in Biology and Medicine. OXFORD: PERGAMON-ELSEVIER SCIENCE LTD, 2022, vol. 141, February 2022, p. 1-7. ISSN 0010-4825. Available from: https://dx.doi.org/10.1016/j.compbiomed.2021.105033.

@article{2249999,
   author = {ElandHariri, Houssam and Neto, Luis A Souto Maior and Cimflová, Petra and Bala, Fouzi and Golan, Rotem and Sojoudi, Alireza and Duszynski, Chris and Elebute, Ibukun and Mousavi, Seyed Hossein and Qiu, Wu and Menon, Bijoy K},
   article_location = {OXFORD},
   article_number = {February 2022},
   doi = {http://dx.doi.org/10.1016/j.compbiomed.2021.105033},
   keywords = {Machine learning; Deep learning; Computer vision; Segmentation; Neurovascular imaging; Computed tomography; Acute ischemic stroke; Brain lesion},
   language = {eng},
   issn = {0010-4825},
   journal = {Computers in Biology and Medicine},
   title = {Evaluating nnU-Net for early ischemic change segmentation on non-contrast computed tomography in patients with Acute Ischemic Stroke},
   url = {https://www.sciencedirect.com/science/article/pii/S0010482521008271?via%3Dihub},
   volume = {141},
   year = {2022}
}

TY  - JOUR
ID  - 2249999
AU  - El-Hariri, Houssam - Neto, Luis A Souto Maior - Cimflová, Petra - Bala, Fouzi - Golan, Rotem - Sojoudi, Alireza - Duszynski, Chris - Elebute, Ibukun - Mousavi, Seyed Hossein - Qiu, Wu - Menon, Bijoy K
PY  - 2022
TI  - Evaluating nnU-Net for early ischemic change segmentation on non-contrast computed tomography in patients with Acute Ischemic Stroke
JF  - Computers in Biology and Medicine
VL  - 141
IS  - February 2022
SP  - 1-7
EP  - 1-7
PB  - PERGAMON-ELSEVIER SCIENCE LTD
SN  - 00104825
KW  - Machine learning
KW  - Deep learning
KW  - Computer vision
KW  - Segmentation
KW  - Neurovascular imaging
KW  - Computed tomography
KW  - Acute ischemic stroke
KW  - Brain lesion
UR  - https://www.sciencedirect.com/science/article/pii/S0010482521008271?via%3Dihub
N2  - Identifying the presence and extent of early ischemic changes (EIC) on Non-Contrast Computed Tomography (NCCT) is key to diagnosing and making time-sensitive treatment decisions in patients that present with Acute Ischemic Stroke (AIS). Segmenting EIC on NCCT is however a challenging task. In this study, we investigated a 3D CNN based on nnU-Net, a self-adapting CNN technique that has become the state-of-the-art in medical image segmentation, for segmenting EIC in NCCT of AIS patients. We trained and tested this model on a sizeable and heterogenous dataset of 534 patients, split into 438 for training and validation and 96 for testing. On this test set, we additionally assessed the inter-rater performance by comparing the proposed approach against two reference segmentation annotations by expert neuroradiologist readers, using this as the benchmark against which to compare our model. In terms of spatial agreement, we report median Dice Similarity Coefficients (DSCs) of 39.8% for the model vs. Reader-1, 39.4% for the model vs. Reader-2, and 55.6% for Reader-2 vs. Reader-1. In terms of lesion volume agreement, we report Intraclass Correlation Coefficients (ICCs) of 83.4% for model vs. Reader-1, 80.4% for model vs. Reader-2, and 94.8% for Reader-2 vs. Reader-1. Based on these results, we conclude that our model performs well relative to expert human performance and therefore may be useful as a decision-aid for clinicians.
ER  -

EL-HARIRI, Houssam, Luis A Souto Maior NETO, Petra CIMFLOVÁ, Fouzi BALA, Rotem GOLAN, Alireza SOJOUDI, Chris DUSZYNSKI, Ibukun ELEBUTE, Seyed Hossein MOUSAVI, Wu QIU and Bijoy K MENON. Evaluating nnU-Net for early ischemic change segmentation on non-contrast computed tomography in patients with Acute Ischemic Stroke. \textit{Computers in Biology and Medicine}. OXFORD: PERGAMON-ELSEVIER SCIENCE LTD, 2022, vol.~141, February 2022, p.~1-7. ISSN~0010-4825. Available from: https://dx.doi.org/10.1016/j.compbiomed.2021.105033.

Detailed Information on Publication Record