Cross-Professional Evaluation of 3D Visualization of Liver Malignancies in the Decade of AI and Automatic Segmentation: A Benefit for Multidisciplinary Teams and Tumor Board Decisions?

Overview

Journal Cureus

Specialty Biomedical Engineering

Date 2024 Nov 25

PMID 39583479

Authors

Christian Heiliger

Dorian Andrade

Lucas Etzel

Daniel Roessler

Vanessa F Schmidt

Florian Boesch

Jens Ricke

Jens Werner

Konrad Karcz

Olga Solyanik

Affiliations

Soon will be listed here.

Abstract

Purpose To investigate whether automatic 3D visualization of computed tomography (CT) data sets with singular liver tumor compared to 2D images could foster a broader understanding of tumor localization and resectability in the liver within a multidisciplinary team and might therefore be a useful tool in multidisciplinary decision-making. Material and methods The study was configured as a web-based questionnaire. Physicians of all levels of medical training from surgery, radiology, and gastroenterology departments were recruited. A total of seven cases with singular liver tumor CT images with adequate quality were selected. Automatic 3D segmentation was performed using Universal Atlas (Release 5.0) as part of the Brainlab of Elements software suite (Brainlab AG, Munich, Germany). All cases were randomly presented in a 2D and 3D manner. After every case-presentation, multiple choice (single answer) questions concerning tumor extent and resectability were asked. The questions as well as the answers defined to be correct, were evaluated by two senior consultants from the radiology and surgery department. The primary outcome parameters were the correctness of answers stratified for medical specialty and for the level of medical training. The secondary outcome was the time needed for the evaluation of seven liver cases using 2D versus 3D images. Six additional questions were tailored to evaluate the subjective value of the 3D visualization. Results A total of 92 participants participated in the study, 31.5% of them were abdominal surgeons, 34.8% gastroenterologists, and 33.7% radiologists. Based on the level of medical training, 66 were residents (71.7%) and 26 consultants (28.3%). Only radiologists answered more questions correctly using 2D imaging compared to the 3D method (p = 0.006). There was no statistically significant difference between correctly answered questions when using 2D vs. 3D visualization in the gastroenterologist and surgeon groups (p > 0.05). The resident subgroup showed no statistically significant difference when using the 2D vs. 3D images (p > 0.05), the consultant subgroup answered more questions correctly using 2D imaging (p = 0.009). Physicians with elementary experience of liver pathology also showed no difference in 2D vs. 3D (p = 0.332), physicians with proficient experience of liver pathology answered more questions correctly using 2D imaging (p = 0.010). The median time taken for the evaluation of the seven liver cases was only significantly faster for the gastroenterologist group (p = 0.006) using the 3D analysis (median: 9.1 minutes) than the 2D analysis (median: 10.7 minutes). Over 80% of the participants found the 3D presentation to be a helpful additional tool for the clinical routine according to the subjective questionnaire. Conclusion In this study 3D visualization of liver tumors was evaluated as helpful within a multidisciplinary team of radiologists, surgeons, and gastroenterologists. However, significantly superior results in the understanding of liver anatomy could not be demonstrated by means of 3D visualization. It may be that more immersive technologies such as augmented reality or virtual reality will lead to a superior understanding compared to conventional presentation of information in 2D cross-sectional images.

References

Gotra A, Sivakumaran L, Chartrand G, Vu K, Vandenbroucke-Menu F, Kauffmann C . Liver segmentation: indications, techniques and future directions. Insights Imaging. 2017; 8(4):377-392. PMC: 5519497. DOI: 10.1007/s13244-017-0558-1. View

Nickel F, Hendrie J, Bruckner T, Kowalewski K, Kenngott H, Muller-Stich B . Successful learning of surgical liver anatomy in a computer-based teaching module. Int J Comput Assist Radiol Surg. 2016; 11(12):2295-2301. DOI: 10.1007/s11548-016-1354-y. View

Drefs M, Schoenberg M, Schiergens T, Albertsmeier M, Andrassy J, Angele M . Chances, challenges and outcomes of patients with complex bilobar colorectal liver metastases approached by a combined multimodal therapy with liver resection and local ablative therapy. J Surg Oncol. 2021; 123(7):1578-1591. DOI: 10.1002/jso.26443. View

Takamoto T, Sano K, Hashimoto T, Ichida A, Shimada K, Maruyama Y . Practical Contribution of Virtual Hepatectomy for Colorectal Liver Metastases: a Propensity-Matched Analysis of Clinical Outcome. J Gastrointest Surg. 2018; 22(12):2037-2044. DOI: 10.1007/s11605-018-3860-4. View

Mise Y, Hasegawa K, Satou S, Shindoh J, Miki K, Akamatsu N . How Has Virtual Hepatectomy Changed the Practice of Liver Surgery?: Experience of 1194 Virtual Hepatectomy Before Liver Resection and Living Donor Liver Transplantation. Ann Surg. 2017; 268(1):127-133. DOI: 10.1097/SLA.0000000000002213. View

Pelanis E, Kumar R, Aghayan D, Palomar R, Fretland A, Brun H . Use of mixed reality for improved spatial understanding of liver anatomy. Minim Invasive Ther Allied Technol. 2019; 29(3):154-160. DOI: 10.1080/13645706.2019.1616558. View

Beermann J, Tetzlaff R, Bruckner T, Schoebinger M, Muller-Stich B, Gutt C . Three-dimensional visualisation improves understanding of surgical liver anatomy. Med Educ. 2010; 44(9):936-940. DOI: 10.1111/j.1365-2923.2010.03742.x. View

Jurgaitis J, Paskonis M, Pivoriunas J, Martinaityte I, Juska A, Jurgaitiene R . The comparison of 2-dimensional with 3-dimensional hepatic visualization in the clinical hepatic anatomy education. Medicina (Kaunas). 2008; 44(6):428-38. View

Muller-Stich B, Lob N, Wald D, Bruckner T, Meinzer H, Kadmon M . Regular three-dimensional presentations improve in the identification of surgical liver anatomy - a randomized study. BMC Med Educ. 2013; 13:131. PMC: 3848999. DOI: 10.1186/1472-6920-13-131. View

10.

Fang X, Xu S, Wood B, Yan P . Deep learning-based liver segmentation for fusion-guided intervention. Int J Comput Assist Radiol Surg. 2020; 15(6):963-972. PMC: 7305971. DOI: 10.1007/s11548-020-02147-6. View

11.

Modest D, Denecke T, Pratschke J, Ricard I, Lang H, Bemelmans M . Surgical treatment options following chemotherapy plus cetuximab or bevacizumab in metastatic colorectal cancer-central evaluation of FIRE-3. Eur J Cancer. 2017; 88:77-86. DOI: 10.1016/j.ejca.2017.10.028. View

12.

Garlipp B, Gibbs P, Van Hazel G, Jeyarajah R, Martin R, Bruns C . Secondary technical resectability of colorectal cancer liver metastases after chemotherapy with or without selective internal radiotherapy in the randomized SIRFLOX trial. Br J Surg. 2019; 106(13):1837-1846. PMC: 6899564. DOI: 10.1002/bjs.11283. View

13.

Yeo C, MacDonald A, Ungi T, Lasso A, Jalink D, Zevin B . Utility of 3D Reconstruction of 2D Liver Computed Tomography/Magnetic Resonance Images as a Surgical Planning Tool for Residents in Liver Resection Surgery. J Surg Educ. 2017; 75(3):792-797. DOI: 10.1016/j.jsurg.2017.07.031. View

14.

Schiergens T, Einem J, Thomas M, Albertsmeier M, Giessen-Jung C, Dorsch M . Multidisciplinary treatment of colorectal liver metastases. Minerva Med. 2017; 108(6):527-546. DOI: 10.23736/S0026-4806.17.05371-X. View

15.

Heker M, Ben-Cohen A, Greenspan H . Hierarchical Fine-Tuning for joint Liver Lesion Segmentation and Lesion Classification in CT. Annu Int Conf IEEE Eng Med Biol Soc. 2020; 2019:895-898. DOI: 10.1109/EMBC.2019.8857127. View

16.

Rafiei S, Karimi N, Mirmahboub B, Najarian K, Felfeliyan B, Samavi S . Liver Segmentation in Abdominal CT Images Using Probabilistic Atlas and Adaptive 3D Region Growing. Annu Int Conf IEEE Eng Med Biol Soc. 2020; 2019:6310-6313. DOI: 10.1109/EMBC.2019.8857835. View

17.

Winkel D, Weikert T, Breit H, Chabin G, Gibson E, Heye T . Validation of a fully automated liver segmentation algorithm using multi-scale deep reinforcement learning and comparison versus manual segmentation. Eur J Radiol. 2020; 126:108918. DOI: 10.1016/j.ejrad.2020.108918. View

18.

Chen S, Zhong X, Hu S, Dorn S, Kachelriess M, Lell M . Automatic multi-organ segmentation in dual-energy CT (DECT) with dedicated 3D fully convolutional DECT networks. Med Phys. 2019; 47(2):552-562. DOI: 10.1002/mp.13950. View

19.

Tang X, Jafargholi Rangraz E, Coudyzer W, Bertels J, Robben D, Schramm G . Whole liver segmentation based on deep learning and manual adjustment for clinical use in SIRT. Eur J Nucl Med Mol Imaging. 2020; 47(12):2742-2752. DOI: 10.1007/s00259-020-04800-3. View

20.

Fang C, An J, Bruno A, Cai X, Fan J, Fujimoto J . Consensus recommendations of three-dimensional visualization for diagnosis and management of liver diseases. Hepatol Int. 2020; 14(4):437-453. PMC: 7366600. DOI: 10.1007/s12072-020-10052-y. View