SegResNet Based Reciprocal Transformation for BONBID-HIE Lesion Segmentation

M. Arda Aydın; Elvin Abdinli; Gozde Unal

doi:10.1007/978-3-031-71626-3_6

SegResNet Based Reciprocal Transformation for BONBID-HIE Lesion Segmentation

M. Arda Aydın^*, Elvin Abdinli, Gozde Unal

^*Bu çalışma için yazışmadan sorumlu yazar

Yapay Zeka ve Veri Mühendisliği Bölümü

Araştırma sonucu: Kitap/Rapor/Konferans Bildirisinde Bölüm › Konferans katkısı › bilirkişi

Özet

Hypoxic Ischemic Encephalopathy (HIE) is a brain disease that affects thousands of neonates every year. Accurate and immediate diagnoses of HIE are crucial elements for clinical treatment and can reduce the disease’s fatality. Magnetic resonance images (MRIs) are effectively used in the clinical treatment of HIE hence, correct segmentation of lesion regions is quite beneficial. However, segmenting HIE lesions is a challenging task due to their nature (i.e., more diffuse and smaller than other lesions). In this paper we propose a novel mathematical operation, “Reciprocal Transformation”, which transforms the data from one distribution to a more suitable distribution to train a deep learning model better. Combined with the concatenation operation, we observe a significant improvement in our validation dataset. The proposed method achieved a mean dice score of 48.26 in the test set of the challenge. Codes are available at: https://github.com/m-arda-aydn/SegResNet-based-Reciprocal-Transformation.

Orijinal dil	İngilizce
Ana bilgisayar yayını başlığı	AI for Brain Lesion Detection and Trauma Video Action Recognition - 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge, Held in Conjunction with MICCAI 2023, Proceedings
Editörler	Rina Bao, Ellen Grant, Yangming Ou, Andrew Kirkpatrick, Juan Wachs
Yayınlayan	Springer Science and Business Media Deutschland GmbH
Sayfalar	39-44
Sayfa sayısı	6
ISBN (Basılı)	9783031716256
DOI'lar	https://doi.org/10.1007/978-3-031-71626-3_6
Yayın durumu	Yayınlandı - 2025
Etkinlik	1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023 - Vancouver, Canada Süre: 12 Eki 2023 → 16 Eki 2023

Yayın serisi

Adı	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Hacim	14567 LNCS
ISSN (Basılı)	0302-9743
ISSN (Elektronik)	1611-3349

???event.eventtypes.event.conference???

???event.eventtypes.event.conference???	1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023
Ülke/Bölge	Canada
Şehir	Vancouver
Periyot	12/10/23 → 16/10/23

Bibliyografik not

Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

Belgeye Erişim

10.1007/978-3-031-71626-3_6

Diğer Dosyalar ve Linkler

Link to publication in Scopus

Alıntı Yap

Aydın, M. A., Abdinli, E., & Unal, G. (2025). SegResNet Based Reciprocal Transformation for BONBID-HIE Lesion Segmentation. In R. Bao, E. Grant, Y. Ou, A. Kirkpatrick, & J. Wachs (Eds.), AI for Brain Lesion Detection and Trauma Video Action Recognition - 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge, Held in Conjunction with MICCAI 2023, Proceedings (pp. 39-44). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Hac. 14567 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-71626-3_6

Aydın, M. Arda ; Abdinli, Elvin ; Unal, Gozde. / SegResNet Based Reciprocal Transformation for BONBID-HIE Lesion Segmentation. AI for Brain Lesion Detection and Trauma Video Action Recognition - 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge, Held in Conjunction with MICCAI 2023, Proceedings. editör / Rina Bao ; Ellen Grant ; Yangming Ou ; Andrew Kirkpatrick ; Juan Wachs. Springer Science and Business Media Deutschland GmbH, 2025. pp. 39-44 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{3a9dea75965e43378df08b7ace036e10,

title = "SegResNet Based Reciprocal Transformation for BONBID-HIE Lesion Segmentation",

abstract = "Hypoxic Ischemic Encephalopathy (HIE) is a brain disease that affects thousands of neonates every year. Accurate and immediate diagnoses of HIE are crucial elements for clinical treatment and can reduce the disease{\textquoteright}s fatality. Magnetic resonance images (MRIs) are effectively used in the clinical treatment of HIE hence, correct segmentation of lesion regions is quite beneficial. However, segmenting HIE lesions is a challenging task due to their nature (i.e., more diffuse and smaller than other lesions). In this paper we propose a novel mathematical operation, “Reciprocal Transformation”, which transforms the data from one distribution to a more suitable distribution to train a deep learning model better. Combined with the concatenation operation, we observe a significant improvement in our validation dataset. The proposed method achieved a mean dice score of 48.26 in the test set of the challenge. Codes are available at: https://github.com/m-arda-aydn/SegResNet-based-Reciprocal-Transformation.",

keywords = "3D Image Segmentation, Hypoxic Ischemic Encephalopathy (HIE), Magnetic Resonance Imaging (MRI)",

author = "Aydın, {M. Arda} and Elvin Abdinli and Gozde Unal",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.; 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023 ; Conference date: 12-10-2023 Through 16-10-2023",

year = "2025",

doi = "10.1007/978-3-031-71626-3_6",

language = "English",

isbn = "9783031716256",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "39--44",

editor = "Rina Bao and Ellen Grant and Yangming Ou and Andrew Kirkpatrick and Juan Wachs",

booktitle = "AI for Brain Lesion Detection and Trauma Video Action Recognition - 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge, Held in Conjunction with MICCAI 2023, Proceedings",

address = "Germany",

}

Aydın, MA, Abdinli, E & Unal, G 2025, SegResNet Based Reciprocal Transformation for BONBID-HIE Lesion Segmentation. in R Bao, E Grant, Y Ou, A Kirkpatrick & J Wachs (eds), AI for Brain Lesion Detection and Trauma Video Action Recognition - 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge, Held in Conjunction with MICCAI 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), hac. 14567 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 39-44, 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023, Vancouver, Canada, 12/10/23. https://doi.org/10.1007/978-3-031-71626-3_6

SegResNet Based Reciprocal Transformation for BONBID-HIE Lesion Segmentation. / Aydın, M. Arda; Abdinli, Elvin; Unal, Gozde.
AI for Brain Lesion Detection and Trauma Video Action Recognition - 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge, Held in Conjunction with MICCAI 2023, Proceedings. ed. / Rina Bao; Ellen Grant; Yangming Ou; Andrew Kirkpatrick; Juan Wachs. Springer Science and Business Media Deutschland GmbH, 2025. p. 39-44 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Hac. 14567 LNCS).

Araştırma sonucu: Kitap/Rapor/Konferans Bildirisinde Bölüm › Konferans katkısı › bilirkişi

TY - GEN

T1 - SegResNet Based Reciprocal Transformation for BONBID-HIE Lesion Segmentation

AU - Aydın, M. Arda

AU - Abdinli, Elvin

AU - Unal, Gozde

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

PY - 2025

Y1 - 2025

N2 - Hypoxic Ischemic Encephalopathy (HIE) is a brain disease that affects thousands of neonates every year. Accurate and immediate diagnoses of HIE are crucial elements for clinical treatment and can reduce the disease’s fatality. Magnetic resonance images (MRIs) are effectively used in the clinical treatment of HIE hence, correct segmentation of lesion regions is quite beneficial. However, segmenting HIE lesions is a challenging task due to their nature (i.e., more diffuse and smaller than other lesions). In this paper we propose a novel mathematical operation, “Reciprocal Transformation”, which transforms the data from one distribution to a more suitable distribution to train a deep learning model better. Combined with the concatenation operation, we observe a significant improvement in our validation dataset. The proposed method achieved a mean dice score of 48.26 in the test set of the challenge. Codes are available at: https://github.com/m-arda-aydn/SegResNet-based-Reciprocal-Transformation.

AB - Hypoxic Ischemic Encephalopathy (HIE) is a brain disease that affects thousands of neonates every year. Accurate and immediate diagnoses of HIE are crucial elements for clinical treatment and can reduce the disease’s fatality. Magnetic resonance images (MRIs) are effectively used in the clinical treatment of HIE hence, correct segmentation of lesion regions is quite beneficial. However, segmenting HIE lesions is a challenging task due to their nature (i.e., more diffuse and smaller than other lesions). In this paper we propose a novel mathematical operation, “Reciprocal Transformation”, which transforms the data from one distribution to a more suitable distribution to train a deep learning model better. Combined with the concatenation operation, we observe a significant improvement in our validation dataset. The proposed method achieved a mean dice score of 48.26 in the test set of the challenge. Codes are available at: https://github.com/m-arda-aydn/SegResNet-based-Reciprocal-Transformation.

KW - 3D Image Segmentation

KW - Hypoxic Ischemic Encephalopathy (HIE)

KW - Magnetic Resonance Imaging (MRI)

UR - http://www.scopus.com/inward/record.url?scp=85208424097&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-71626-3_6

DO - 10.1007/978-3-031-71626-3_6

M3 - Conference contribution

AN - SCOPUS:85208424097

SN - 9783031716256

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 39

EP - 44

BT - AI for Brain Lesion Detection and Trauma Video Action Recognition - 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge, Held in Conjunction with MICCAI 2023, Proceedings

A2 - Bao, Rina

A2 - Grant, Ellen

A2 - Ou, Yangming

A2 - Kirkpatrick, Andrew

A2 - Wachs, Juan

PB - Springer Science and Business Media Deutschland GmbH

T2 - 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge Held in Conjunction with 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023

Y2 - 12 October 2023 through 16 October 2023

ER -

Aydın MA, Abdinli E, Unal G. SegResNet Based Reciprocal Transformation for BONBID-HIE Lesion Segmentation. In Bao R, Grant E, Ou Y, Kirkpatrick A, Wachs J, editörler, AI for Brain Lesion Detection and Trauma Video Action Recognition - 1st BONBID-HIE Lesion Segmentation Challenge and 1st Trauma Thompson Challenge, Held in Conjunction with MICCAI 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2025. p. 39-44. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-71626-3_6