Precision of computed tomography volumetry in living donor liver transplantation: current evidence, limitations, and future directions

The remarkable study “Optimizing accuracy: a comparative analysis of preoperative liver volumetry in living donor liver transplantation from a surgeon’s perspective - a retrospective cohort study” by Choi and Kim (1) provides valuable insights into this evolving field. Building upon their contribution, we offer a focused commentary addressing current evidence, limitations, and future directions of computed tomography (CT) volumetry in living donor liver transplantation (LDLT).

Accurate graft volume estimation is essential in LDLT because graft-recipient size mismatch can lead to small-for-size syndrome (SFSS), graft dysfunction, or graft loss. CT volumetry has long been considered the standard method for estimating graft size preoperatively due to its wide availability, rapid acquisition, and high spatial resolution (2). With advances in segmentation algorithms, software automation, and imaging quality, CT volumetry has gained improved precision and reproducibility. However, discrepancies remain between estimated graft volume (EGV) and actual graft weight (AGW), largely due to biological and technical variables—including steatosis, vascular collapse, dehydration, and segmentation error (3,4).

Accuracy of CT volumetry

Recent high-quality clinical studies demonstrate that CT volumetry can predict graft weight with high reliability, with error rates typically ranging from 3–8% using modern semi-automated segmentation systems (5). A large cohort analysis reported that accurate graft estimation correlated strongly with reduced SFSS incidence and improved postoperative survival (6). CT volumetry has therefore been endorsed as the preferred preoperative method by major international guidelines, including the International Liver Transplantation Society (ILTS) and the European Association for the Study of the Liver (EASL) (7). Despite its accuracy, CT volumetry tends to overestimate AGW due to hepatic venous collapse and dehydration occurring after graft retrieval and cold perfusion (8).

The discrepancy between CT-EGV and AGW is influenced by both physiological and procedural factors. While post-retrieval dehydration and venous collapse during cold perfusion partially explain overestimation, the definition of the reference graft volume itself warrants careful consideration. From a functional standpoint, in vivo volumetric assessment of the perfused future graft may represent the most clinically relevant reference, as it reflects the graft condition at the time of recipient implantation. In contrast, post-perfusion measurements reflect ex vivo changes and should be interpreted accordingly when evaluating volumetry accuracy.

To compensate, correction coefficients ranging between 0.82 and 0.95 have been proposed as observed similarly in the data reported in the present study, although values vary depending on donor body habitus and hepatic pathology (1,9).

Technical and methodological considerations

Segmentation methodology is a major determinant of volumetric precision. Manual segmentation offers high anatomical fidelity but suffers from interobserver variability and substantial time cost (10). Semi-automated segmentation software reduces variability, while fully automated artificial intelligence (AI)-based segmentation approaches now produce exceptionally consistent results. A recent deep learning-based volumetry study demonstrated close agreement with those measured by the radiologist manually, with a small bias (i.e., −1% to 0.6% of the measured volumetric indices for all indices) and measurement error (i.e., <5.2% of the measured volumetric indices for all indices (11). Optimal protocol parameters also influence measurement reliability. Portal venous phase CT is widely recommended because it provides uniform parenchymal enhancement and clear vascular boundary definition (7,11). Slice thickness ≤1.25 mm has further been associated with superior volumetric accuracy and detection of subtle anatomical variations (12).

Considerations in interpreting CT-based volume estimation

Despite its high anatomical precision, CT volumetry primarily provides quantitative structural information and does not, when used in isolation, fully reflect functional liver reserve. In contemporary clinical practice, volumetric assessment is therefore interpreted in conjunction with functional evaluation to optimize preoperative decision-making. Hepatic steatosis affects tissue density and may contribute to volume-weight mismatch; livers with >30% macrovesicular steatosis can exhibit reduced functional hepatocyte mass despite preserved volume, thereby increasing the risk of SFSS even when graft size meets conventional volumetric thresholds (13).

Another potential source of volumetric bias arises from segmental congestion caused by unanticipated venous outflow impairment, particularly in partial grafts. Such congestion may lead to discrepancies between EGV and effective functional graft mass, underscoring the limitation of purely volumetric approaches and the need for complementary functional assessment.

While CT volumetry itself is not designed to directly assess hepatocyte metabolic capacity, biliary function, or microvascular perfusion, these functional aspects are increasingly evaluated using complementary modalities. Future advances integrating volumetric data with functional imaging and quantitative liver function tests are expected to further enhance the accuracy and clinical relevance of preoperative liver assessment.

Emerging technologies and future perspective

Recent advances in functional imaging and AI are reshaping the paradigm of liver graft assessment from purely volumetric estimation toward integrated quantitative evaluation. Magnetic resonance imaging-proton density fat fraction (MRI-PDFF) enables accurate, noninvasive quantification of donor steatosis and has demonstrated superiority over CT attenuation-based assessments (14).

Beyond single-modality improvements, the emerging novelty lies in the integration of multimodal data. AI-driven platforms combining CT-based anatomical segmentation, MRI-derived fat quantification, and radiomics-based tissue characterization have shown improved predictive performance for small-for-size syndrome compared with CT volumetry alone (15). Importantly, these approaches shift preoperative assessment from static volume measurement to patient-specific risk stratification. In addition to MRI-based functional imaging, hepatobiliary scintigraphy has also been investigated for functional liver assessment, particularly for segmental function estimation (16,17). Although promising, its spatial resolution and limited availability have thus far restricted widespread application in routine living donor evaluation. Further integration with high-resolution imaging modalities may enhance its clinical utility.

Future LDLT planning is expected to adopt stepwise hybrid frameworks in which volumetric data provide structural feasibility, functional imaging refines graft quality assessment, and AI-based modeling integrates these parameters into individualized decision-support systems. Such an approach may represent a conceptual transition from “quantitative anatomy” to “quantitative physiology”, enabling more precise graft selection and potentially improving postoperative outcomes (Figure 1).

Figure 1 Hybrid graft assessment framework in living donor liver transplantation. This schematic illustrates a hybrid approach to graft assessment integrating anatomical, functional, and computational parameters. CT volumetry provides structural assessment of graft volume and vascular anatomy, while functional imaging modalities, including MRI-PDFF and hepatobiliary scintigraphy, enable evaluation of graft quality through steatosis quantification and regional liver function. Artificial intelligence-based integration of multimodal imaging data supports individualized risk assessment for small-for-size syndrome and postoperative graft dysfunction. AI, artificial intelligence; CT, computed tomography; MRI-PDFF, magnetic resonance imaging-proton density fat fraction.

However, several practical barriers must be acknowledged before widespread clinical adoption of MRI- and AI-based hybrid assessment models. MRI availability remains limited in some transplant centers, and prolonged acquisition time may pose logistical challenges in time-sensitive donor evaluations. In addition, AI-driven platforms require dedicated infrastructure, technical expertise, and regulatory validation, which may increase initial implementation costs. Workflow integration into routine preoperative planning also represents a nontrivial hurdle. Therefore, stepwise adoption strategies that balance technological advancement with clinical feasibility will be essential for successful translation into daily practice.

Future validation of multimodal prediction models should be pursued through prospective, multicenter studies, enabling standardized assessment of predictive accuracy, generalizability, and clinical impact across diverse donor populations and institutional settings. Such study designs will be critical to establish robust evidence beyond retrospective analyses and to define clinically meaningful thresholds for hybrid graft assessment.

Conclusions

CT volumetry remains the foundation of graft size estimation in LDLT and provides excellent anatomical precision when using standardized imaging protocols and automated segmentation. However, persistent discrepancies between estimated volume and actual graft function highlight inherent limitations. As the field advances, integration of AI-driven segmentation and functional assessment tools such as MRI-PDFF will refine graft evaluation beyond pure anatomical volume. The future direction of transplant planning favors a hybrid framework combining structural, physiological, and computational metrics to enhance donor safety and optimize post-transplant outcomes.

Acknowledgments

None.

Footnote

Provenance and Peer Review: This article was commissioned by the editorial office, HepatoBiliary Surgery and Nutrition. The article has undergone external peer review.

Peer Review File: Available at https://hbsn.amegroups.com/article/view/10.21037/hbsn-2025-1-927/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://hbsn.amegroups.com/article/view/10.21037/hbsn-2025-1-927/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Choi E, Kim SH. Optimizing accuracy: a comparative analysis of preoperative liver volumetry in living donor liver transplantation from a surgeon's perspective - a retrospective cohort study. Int J Surg 2025;111:8149-58. [Crossref] [PubMed]
2. Yang X, Park S, Lee S, et al. Estimation of right lobe graft weight for living donor liver transplantation using deep learning-based fully automatic computed tomographic volumetry. Sci Rep 2023;13:17746. [Crossref] [PubMed]
3. Kalshabay Y, Zholdybay Z, Di Martino M, et al. CT volume analysis in living donor liver transplantation: accuracy of three different approaches. Insights Imaging 2023;14:82. [Crossref] [PubMed]
4. Machry M, Ferreira LF, Lucchese AM, et al. Liver volumetric and anatomic assessment in living donor liver transplantation: The role of modern imaging and artificial intelligence. World J Transplant 2023;13:290-8. [Crossref] [PubMed]
5. Çelik H, Odaman H, Altay C, et al. Manual and semi-automated computed tomography volumetry significantly overestimates the right liver lobe graft weight: a single-center study with adult living liver donors. Diagn Interv Radiol 2024;30:3-8. [Crossref] [PubMed]
6. Do H, Baik J, Gwon MS, et al. Accuracy and efficiency of artificial Intelligence–Assisted three-dimensional liver volumetry in living donor evaluation based on real world prospective data. HPB 2026;28:236-44. [Crossref] [PubMed]
7. Chadha R, Sakai T, Rajakumar A, et al. Anesthesia and Critical Care for the Prediction and Prevention for Small-for-size Syndrome: Guidelines from the ILTS-iLDLT-LTSI Consensus Conference. Transplantation 2023;107:2216-25. [Crossref] [PubMed]
8. Goja S, Yadav SK, Yadav A, et al. Accuracy of preoperative CT liver volumetry in living donor hepatectomy and its clinical implications. Hepatobiliary Surg Nutr 2018;7:167-74. [Crossref] [PubMed]
9. Karlo C, Reiner CS, Stolzmann P, et al. CT- and MRI-based volumetry of resected liver specimen: comparison to intraoperative volume and weight measurements and calculation of conversion factors. Eur J Radiol 2010;75:e107-11. [Crossref] [PubMed]
10. Baiguissova D, Kalshabay Y, Martino MD, et al. Semi-automatic and automatic segmentation in the preoperative assessment of graft volume before living donor liver transplantation. Eur J Radiol 2025;191:112367. [Crossref] [PubMed]
11. Ahn Y, Yoon JS, Lee SS, et al. Deep Learning Algorithm for Automated Segmentation and Volume Measurement of the Liver and Spleen Using Portal Venous Phase Computed Tomography Images. Korean J Radiol 2020;21:987-97. [Crossref] [PubMed]
12. Hori M, Suzuki K, Epstein ML, et al. Computed tomography liver volumetry using 3-dimensional image data in living donor liver transplantation: effects of the slice thickness on the volume calculation. Liver Transpl 2011;17:1427-36. [Crossref] [PubMed]
13. Spitzer AL, Lao OB, Dick AA, et al. The biopsied donor liver: incorporating macrosteatosis into high-risk donor assessment. Liver Transpl 2010;16:874-84. [Crossref] [PubMed]
14. Qi Q, Weinstock AK, Chupetlovska K, et al. Magnetic resonance imaging-derived proton density fat fraction (MRI-PDFF) is a viable alternative to liver biopsy for steatosis quantification in living liver donor transplantation. Clin Transplant 2021;35:e14339. [Crossref] [PubMed]
15. Gross M, Huber S, Arora S, et al. Automated MRI liver segmentation for anatomical segmentation, liver volumetry, and the extraction of radiomics. Eur Radiol 2024;34:5056-65. [Crossref] [PubMed]
16. Serenari M, Ravaioli M, Cescon M. Hepatobiliary Scintigraphy to Increase the Safety in Right Lobe Living Donor Liver Transplantation: Not Just a Matter of Age and Remnant Volume. Transplantation 2021;105:e63. [Crossref] [PubMed]
17. Serenari M, Pettinato C, Bonatti C, et al. Hepatobiliary Scintigraphy in the Preoperative Evaluation of Potential Living Liver Donors. Transplant Proc 2019;51:167-70. [Crossref] [PubMed]