Enhanced Small Liver Lesion Detection and Segmentation Using a Size-Focused Multi-model Approach in CT Scans

Abdullah F. Al-Battal; Van Ha Tang; Steven Q.H. Truong; Truong Q. Nguyen; Cheolhong An

doi:10.1007/978-3-031-73284-3_32

Enhanced Small Liver Lesion Detection and Segmentation Using a Size-Focused Multi-model Approach in CT Scans

Abdullah F. Al-Battal^*
, Van Ha Tang
, Steven Q.H. Truong
, Truong Q. Nguyen
, Cheolhong An

^*Corresponding author for this work

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

Abstract

This paper presents a novel approach to enhance the detection and segmentation of small liver lesions in computed tomography (CT) scans using a size-focused multi-model framework. Current state-of-the-art segmentation models, primarily based on the UNet architecture, often exhibit inferior performance on small lesions due to severe class and size imbalances. We introduce a model architecture incorporating a configurable attention mechanism within the model’s skip connections and a lesion selection algorithm that compares predictions from multiple models, including a general lesion segmentation model and a small lesion-focused model, selecting the most suitable prediction. The approach was evaluated on a clinical 3-phase CT dataset and the public LiTS dataset. Results show improvements in overall lesion segmentation performance by 1.5% and 1.9% for the clinical and LiTS datasets, respectively. Additionally, the detection of small lesions improved by 4.4% and 1.8% for both datasets, respectively.

Original language	English
Title of host publication	Machine Learning in Medical Imaging - 15th International Workshop, MLMI 2024, Held in Conjunction with MICCAI 2024, Proceedings
Editors	Xuanang Xu, Zhiming Cui, Kaicong Sun, Islem Rekik, Xi Ouyang
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	320-330
Number of pages	11
ISBN (Print)	9783031732836
DOIs	https://doi.org/10.1007/978-3-031-73284-3_32
State	Published - 2025
Event	15th International Workshop on Machine Learning in Medical Imaging, MLMI 2024 was held in conjunction with the 27th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024 - Marrakesh, Morocco Duration: 6 Oct 2024 → 6 Oct 2024

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	15241 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	15th International Workshop on Machine Learning in Medical Imaging, MLMI 2024 was held in conjunction with the 27th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024
Country/Territory	Morocco
City	Marrakesh
Period	6/10/24 → 6/10/24

Bibliographical note

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

Keywords

Convolutional neural networks
Liver lesion segmentation
Medical image segmentation
Small lesion segmentation

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-031-73284-3_32

Cite this

Al-Battal, A. F., Tang, V. H., Truong, S. Q. H., Nguyen, T. Q., & An, C. (2025). Enhanced Small Liver Lesion Detection and Segmentation Using a Size-Focused Multi-model Approach in CT Scans. In X. Xu, Z. Cui, K. Sun, I. Rekik, & X. Ouyang (Eds.), Machine Learning in Medical Imaging - 15th International Workshop, MLMI 2024, Held in Conjunction with MICCAI 2024, Proceedings (pp. 320-330). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 15241 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-73284-3_32

Al-Battal, Abdullah F. ; Tang, Van Ha ; Truong, Steven Q.H. et al. / Enhanced Small Liver Lesion Detection and Segmentation Using a Size-Focused Multi-model Approach in CT Scans. Machine Learning in Medical Imaging - 15th International Workshop, MLMI 2024, Held in Conjunction with MICCAI 2024, Proceedings. editor / Xuanang Xu ; Zhiming Cui ; Kaicong Sun ; Islem Rekik ; Xi Ouyang. Springer Science and Business Media Deutschland GmbH, 2025. pp. 320-330 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{d2157a6ab28c4d708916f0ed1b0a8431,

title = "Enhanced Small Liver Lesion Detection and Segmentation Using a Size-Focused Multi-model Approach in CT Scans",

abstract = "This paper presents a novel approach to enhance the detection and segmentation of small liver lesions in computed tomography (CT) scans using a size-focused multi-model framework. Current state-of-the-art segmentation models, primarily based on the UNet architecture, often exhibit inferior performance on small lesions due to severe class and size imbalances. We introduce a model architecture incorporating a configurable attention mechanism within the model{\textquoteright}s skip connections and a lesion selection algorithm that compares predictions from multiple models, including a general lesion segmentation model and a small lesion-focused model, selecting the most suitable prediction. The approach was evaluated on a clinical 3-phase CT dataset and the public LiTS dataset. Results show improvements in overall lesion segmentation performance by 1.5\% and 1.9\% for the clinical and LiTS datasets, respectively. Additionally, the detection of small lesions improved by 4.4\% and 1.8\% for both datasets, respectively.",

keywords = "Convolutional neural networks, Liver lesion segmentation, Medical image segmentation, Small lesion segmentation",

author = "Al-Battal, \{Abdullah F.\} and Tang, \{Van Ha\} and Truong, \{Steven Q.H.\} and Nguyen, \{Truong Q.\} and Cheolhong An",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.; 15th International Workshop on Machine Learning in Medical Imaging, MLMI 2024 was held in conjunction with the 27th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024 ; Conference date: 06-10-2024 Through 06-10-2024",

year = "2025",

doi = "10.1007/978-3-031-73284-3\_32",

language = "English",

isbn = "9783031732836",

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 = "320--330",

editor = "Xuanang Xu and Zhiming Cui and Kaicong Sun and Islem Rekik and Xi Ouyang",

booktitle = "Machine Learning in Medical Imaging - 15th International Workshop, MLMI 2024, Held in Conjunction with MICCAI 2024, Proceedings",

address = "Germany",

}

Al-Battal, AF, Tang, VH, Truong, SQH, Nguyen, TQ & An, C 2025, Enhanced Small Liver Lesion Detection and Segmentation Using a Size-Focused Multi-model Approach in CT Scans. in X Xu, Z Cui, K Sun, I Rekik & X Ouyang (eds), Machine Learning in Medical Imaging - 15th International Workshop, MLMI 2024, Held in Conjunction with MICCAI 2024, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 15241 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 320-330, 15th International Workshop on Machine Learning in Medical Imaging, MLMI 2024 was held in conjunction with the 27th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024, Marrakesh, Morocco, 6/10/24. https://doi.org/10.1007/978-3-031-73284-3_32

Enhanced Small Liver Lesion Detection and Segmentation Using a Size-Focused Multi-model Approach in CT Scans. / Al-Battal, Abdullah F.; Tang, Van Ha; Truong, Steven Q.H. et al.
Machine Learning in Medical Imaging - 15th International Workshop, MLMI 2024, Held in Conjunction with MICCAI 2024, Proceedings. ed. / Xuanang Xu; Zhiming Cui; Kaicong Sun; Islem Rekik; Xi Ouyang. Springer Science and Business Media Deutschland GmbH, 2025. p. 320-330 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 15241 LNCS).

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

TY - GEN

T1 - Enhanced Small Liver Lesion Detection and Segmentation Using a Size-Focused Multi-model Approach in CT Scans

AU - Al-Battal, Abdullah F.

AU - Tang, Van Ha

AU - Truong, Steven Q.H.

AU - Nguyen, Truong Q.

AU - An, Cheolhong

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

PY - 2025

Y1 - 2025

N2 - This paper presents a novel approach to enhance the detection and segmentation of small liver lesions in computed tomography (CT) scans using a size-focused multi-model framework. Current state-of-the-art segmentation models, primarily based on the UNet architecture, often exhibit inferior performance on small lesions due to severe class and size imbalances. We introduce a model architecture incorporating a configurable attention mechanism within the model’s skip connections and a lesion selection algorithm that compares predictions from multiple models, including a general lesion segmentation model and a small lesion-focused model, selecting the most suitable prediction. The approach was evaluated on a clinical 3-phase CT dataset and the public LiTS dataset. Results show improvements in overall lesion segmentation performance by 1.5% and 1.9% for the clinical and LiTS datasets, respectively. Additionally, the detection of small lesions improved by 4.4% and 1.8% for both datasets, respectively.

AB - This paper presents a novel approach to enhance the detection and segmentation of small liver lesions in computed tomography (CT) scans using a size-focused multi-model framework. Current state-of-the-art segmentation models, primarily based on the UNet architecture, often exhibit inferior performance on small lesions due to severe class and size imbalances. We introduce a model architecture incorporating a configurable attention mechanism within the model’s skip connections and a lesion selection algorithm that compares predictions from multiple models, including a general lesion segmentation model and a small lesion-focused model, selecting the most suitable prediction. The approach was evaluated on a clinical 3-phase CT dataset and the public LiTS dataset. Results show improvements in overall lesion segmentation performance by 1.5% and 1.9% for the clinical and LiTS datasets, respectively. Additionally, the detection of small lesions improved by 4.4% and 1.8% for both datasets, respectively.

KW - Convolutional neural networks

KW - Liver lesion segmentation

KW - Medical image segmentation

KW - Small lesion segmentation

UR - https://www.scopus.com/pages/publications/85208276776

U2 - 10.1007/978-3-031-73284-3_32

DO - 10.1007/978-3-031-73284-3_32

M3 - Conference contribution

AN - SCOPUS:85208276776

SN - 9783031732836

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

SP - 320

EP - 330

BT - Machine Learning in Medical Imaging - 15th International Workshop, MLMI 2024, Held in Conjunction with MICCAI 2024, Proceedings

A2 - Xu, Xuanang

A2 - Cui, Zhiming

A2 - Sun, Kaicong

A2 - Rekik, Islem

A2 - Ouyang, Xi

PB - Springer Science and Business Media Deutschland GmbH

T2 - 15th International Workshop on Machine Learning in Medical Imaging, MLMI 2024 was held in conjunction with the 27th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024

Y2 - 6 October 2024 through 6 October 2024

ER -

Al-Battal AF, Tang VH, Truong SQH, Nguyen TQ, An C. Enhanced Small Liver Lesion Detection and Segmentation Using a Size-Focused Multi-model Approach in CT Scans. In Xu X, Cui Z, Sun K, Rekik I, Ouyang X, editors, Machine Learning in Medical Imaging - 15th International Workshop, MLMI 2024, Held in Conjunction with MICCAI 2024, Proceedings. Springer Science and Business Media Deutschland GmbH. 2025. p. 320-330. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-73284-3_32