Benchmarking Deep Learning for Future Liver Remnant Segmentation in Colorectal Liver Metastasis

Anthony T. Wu; Arghavan Rezvani; Kela Liu; Roozbeh Houshyar; Pooya Khosravi; Whitney Li; Xiaohui Xie

cs.LGApr 9, 2026

Benchmarking Deep Learning for Future Liver Remnant Segmentation in Colorectal Liver Metastasis

Authors: Anthony T. Wu, Arghavan Rezvani, Kela Liu, Roozbeh Houshyar, Pooya Khosravi, Whitney Li, Xiaohui Xie

Organizations: University of California, Irvine Donald Bren School of Information and Computer Science · University of California, Irvine School of Medicine

Abstract

Accurate segmentation of the future liver remnant (FLR) is critical for surgical planning in colorectal liver metastases (CRLM) to prevent fatal post-hepatectomy liver failure. However, this segmentation task is technically challenging due to complex resection boundaries, convoluted hepatic vasculature and diffuse metastatic lesions. A primary bottleneck in developing automated AI tools has been the lack of high-fidelity, validated data. We address this gap by manually refining all 197 volumes from the public CRLM-CT-Seg dataset, creating the first open-source, validated benchmark for this task. We then establish the first segmentation baselines, comparing cascaded (Liver->CRLM->FLR) and end-to-end (E2E) strategies using nnU-Net, SwinUNETR, and STU-Net. We find a cascaded nnU-Net achieves the best final FLR segmentation Dice (0.767), while the pretrained STU-Net provides superior CRLM segmentation (0.620 Dice) and is significantly more robust to cascaded errors. This work provides the first validated benchmark and a reproducible framework to accelerate research in AI-assisted surgical planning.

Figures & tables

Figure 1: Study Overview. (a) Dataset refinement (radiologist-confirmed) and data split. An example correction of an erroneous CRLM label in the lung (blue) is shown. Labels: LRS (red), FLR (green). (b) Model development comparing a 3-stage Cascaded pipeline (top) and an End-to-End (E2E) strategy (bottom). For the cascade, solid lines = training (ground-truth inputs); dotted lines = inference (predicted inputs).

Figure 2: Refinement of liver border. (A) Axial CT (W/L 370/70 HU) shows two lesions (arrows). (B) Original segmentation contains patch-like artifacts. (C) Refined manual segmentation removes voids and ensures precise liver contours.

Figure 3: Refinement of imprecise CRLM segmentation. (A) Axial CT (W/L 178/88 HU) shows two lesions (arrows). (B) Original segmentation fails to capture lesion L1 and coarsely over-segments L2. (C) Refined manual segmentation provides precise contours for both.

		Pipelined Tasks			E2E Task
Model		Liver (CT)	CRLM (L+CT)	FLR (L+T+CT)	FLR (CT)
nnU-Net	D	$0.951\pm 0.009$	$0.689\pm 0.041$	$\mathbf{0.834\pm 0.023}$	$0.755\pm 0.027$
	P	$0.940\pm 0.012$	$0.740\pm 0.057$	$\mathbf{0.822\pm 0.025}$	$0.677\pm 0.042$
	R	$0.965\pm 0.008$	$0.703\pm 0.053$	$0.903\pm 0.029$	$\mathbf{0.923\pm 0.010}$
Swin- UNETR	D	$0.960\pm 0.012$	$0.573\pm 0.055$	$0.756\pm 0.037$	$0.737\pm 0.032$
	P	$0.959\pm 0.017$	$\mathbf{0.784\pm 0.051}$	$0.732\pm 0.042$	$\mathbf{0.717\pm 0.053}$

Table 1: Model 5-fold cross-validation performance. All metrics are macro-averaged on a per-case basis. Abbreviations: D (Dice), P (Precision), R (Recall), L (Liver), T (CRLM Tumor), E2E (End-to-End).

		Pipelined Tasks			Cascaded Tasks		E2E Task
Model		Liver (CT)	CRLM (L+CT)	FLR (L+T+CT)	CRLM (L+CT)	FLR (L+T+CT)	FLR (CT)
nnU-Net	D	$0.944$	$0.588$	$0.815$	$0.525$	0.767	$0.757$
	P	$0.944$	$0.605$	0.815	$0.631$	0.768	$0.684$
	R	$0.945$	0.641	$0.883$	$0.520$	$0.831$	0.915
Swin- UNETR	D	$0.969$	$0.445$	$0.742$	$0.379$	$0.694$	$0.754$
	P	$0.971$	$0.635$	$0.727$	$0.549$	$0.639$	0.738

Table 2: Test set performance. All metrics are macro-averaged on a per-case basis. Abbreviations: D (Dice), P (Precision), R (Recall), L (Liver), T (CRLM Tumor), E2E (End-to-End).

Medical Image Segmentation

Medical image segmentation delineates organs, lesions, and other structures in scans to support measurement, diagnosis, treatment planning, and research.

See all 788 papers

An Automated Radiomics Framework for Postoperative Survival Prediction in Colorectal Liver Metastases using Preoperative MRI
Mar 10, 2026Muhammad Alberb, Jianan Chen, Hossam El-rewaidy +5Cancer Survival Prediction Radiomics
Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada · Physical Sciences Platform, Sunnybrook Research Institute, Toronto, ON, Canada · UCL Cancer Institute, University College London, London, UK +5
Spectral Adapters for Segment Anything Model-based Segmentation of Colorectal Liver Metastases in Computed Tomography
Sep 10, 2026Ramtin Mojtahedi, Mohammad Hamghalam, Jacob J. Peoples +6Promptable Image Segmentation Vision Foundation Model Adaptation
School of Computing, Queen’s University, Kingston, ON, Canada · Toronto General Hospital Research Institute, University Health Network, 200 Elizabeth Street, Toronto, ON M5G 2C4, Canada · Department of Electrical Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran +6
TriALS: Triphasic-Aided Liver Lesion Segmentation Benchmark in Non-Contrast CT
May 15, 2026Marawan Elbatel, Mohamed Ghonim, Jiaji Mao +62Image Segmentation Medical Image Segmentation

Mar 10, 2026J/K move · Enter open · S save

An Automated Radiomics Framework for Postoperative Survival Prediction in Colorectal Liver Metastases using Preoperative MRI

Muhammad Alberb, Jianan Chen, Hossam El-rewaidy, Paul Karanicolas+4

Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada · Physical Sciences Platform, Sunnybrook Research Institute, Toronto, ON, Canada · UCL Cancer Institute, University College London, London, UK+5

Cancer Survival Prediction Radiomics Survival Analysis Medical Image Segmentation Semi-Supervised Medical Image Segmentation

Open paper AskPDF arXiv

While colorectal liver metastasis (CRLM) is potentially curable via hepatectomy, patient outcomes remain highly heterogeneous. Postoperative survival prediction is necessary to avoid non-beneficial surgeries and guide personalized therapy. In this study, we present an automated AI-based framework for postoperative CRLM survival prediction using pre- and post-contrast MRI. We performed a retrospective study of 227 CRLM patients who had gadoxetate-enhanced MRI prior to curative-intent hepatectomy between 2013 and 2020. We developed a survival prediction framework comprising an anatomy-aware segmentation pipeline followed by a radiomics pipeline. The segmentation pipeline learns liver, CRLMs, and spleen segmentation from partially-annotated data, leveraging promptable foundation models to generate pseudo-labels. To support this pipeline, we propose SAMONAI, a prompt propagation algorithm that extends Segment Anything Model to 3D point-based segmentation. Predicted pre- and post-contrast segmentations are then fed into our radiomics pipeline, which extracts per-tumor features and predicts survival using SurvAMINN, an autoencoder-based multiple instance neural network for time-to-event survival prediction. SurvAMINN jointly learns dimensionality reduction and survival prediction from right-censored data, emphasizing high-risk metastases. We compared our framework against established methods and biomarkers using univariate and multivariate Cox regression. Our segmentation pipeline achieves median Dice scores of 0.96 (liver) and 0.93 (spleen), driving a CRLM segmentation Dice score of 0.78 and a detection F1-score of 0.79. Accurate segmentation enables our radiomics pipeline to achieve a survival prediction C-index of 0.69. Our results show the potential of integrating segmentation algorithms with radiomics-based survival analysis to deliver accurate and automated CRLM outcome prediction.

Abstract

Medical Image Segmentation

Explore similar work

An Automated Radiomics Framework for Postoperative Survival Prediction in Colorectal Liver Metastases using Preoperative MRI

Spectral Adapters for Segment Anything Model-based Segmentation of Colorectal Liver Metastases in Computed Tomography

TriALS: Triphasic-Aided Liver Lesion Segmentation Benchmark in Non-Contrast CT