Prediction of Biliary Complications After Human Liver Transplantation Using Hyperspectral Imaging and Convolutional Neural Networks: A Proof-of-concept Study.
Journal
Transplantation
ISSN: 1534-6080
Titre abrégé: Transplantation
Pays: United States
ID NLM: 0132144
Informations de publication
Date de publication:
18 Aug 2023
18 Aug 2023
Historique:
pubmed:
18
8
2023
medline:
18
8
2023
entrez:
18
8
2023
Statut:
aheadofprint
Résumé
Biliary complications (BCs) negatively impact the outcome after liver transplantation. We herein tested whether hyperspectral imaging (HSI) generated data from bile ducts (BD) on reperfusion and machine learning techniques for data readout may serve as a novel approach for predicting BC. Tissue-specific data from 136 HSI liver images were integrated into a convolutional neural network (CNN). Fourteen patients undergoing liver transplantation after normothermic machine preservation served as a validation cohort. Assessment of oxygen saturation, organ hemoglobin, and tissue water levels through HSI was performed after completing the biliary anastomosis. Resected BD segments were analyzed by immunohistochemistry and real-time confocal microscopy. Immunohistochemistry and real-time confocal microscopy revealed mild (grade I: 1%-40%) BD damage in 8 patients and moderate (grade II: 40%-80%) injury in 1 patient. Donor and recipient data alone had no predictive capacity toward BC. Deep learning-based analysis of HSI data resulted in >90% accuracy of automated detection of BD. The CNN-based analysis yielded a correct classification in 72% and 69% for BC/no BC. The combination of HSI with donor and recipient factors showed 94% accuracy in predicting BC. Deep learning-based modeling using CNN of HSI-based tissue property data represents a noninvasive technique for predicting postoperative BC.
Sections du résumé
BACKGROUND
BACKGROUND
Biliary complications (BCs) negatively impact the outcome after liver transplantation. We herein tested whether hyperspectral imaging (HSI) generated data from bile ducts (BD) on reperfusion and machine learning techniques for data readout may serve as a novel approach for predicting BC.
METHODS
METHODS
Tissue-specific data from 136 HSI liver images were integrated into a convolutional neural network (CNN). Fourteen patients undergoing liver transplantation after normothermic machine preservation served as a validation cohort. Assessment of oxygen saturation, organ hemoglobin, and tissue water levels through HSI was performed after completing the biliary anastomosis. Resected BD segments were analyzed by immunohistochemistry and real-time confocal microscopy.
RESULTS
RESULTS
Immunohistochemistry and real-time confocal microscopy revealed mild (grade I: 1%-40%) BD damage in 8 patients and moderate (grade II: 40%-80%) injury in 1 patient. Donor and recipient data alone had no predictive capacity toward BC. Deep learning-based analysis of HSI data resulted in >90% accuracy of automated detection of BD. The CNN-based analysis yielded a correct classification in 72% and 69% for BC/no BC. The combination of HSI with donor and recipient factors showed 94% accuracy in predicting BC.
CONCLUSIONS
CONCLUSIONS
Deep learning-based modeling using CNN of HSI-based tissue property data represents a noninvasive technique for predicting postoperative BC.
Identifiants
pubmed: 37592397
doi: 10.1097/TP.0000000000004757
pii: 00007890-990000000-00515
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Subventions
Organisme : Gabriel Salzner Stiftung
Informations de copyright
Copyright © 2023 Wolters Kluwer Health, Inc. All rights reserved.
Déclaration de conflit d'intérêts
The authors declare no conflicts of interest.
Références
Kochhar G, Parungao JM, Hanouneh IA, et al. Biliary complications following liver transplantation. World J Gastroenterol. 2013;19:2841–2846.
Manay P, Seth A, Jackson K, et al. Biliary complications after liver transplantation in the united states: changing trends and economic implications. Transplantation. 2023;107:e127–e138.
Weissenbacher A, Bogensperger C, Oberhuber R, et al. Perfusate enzymes and platelets indicate early allograft dysfunction after transplantation of normothermically preserved livers. Transplantation. 2021;106:792–805.
Fodor M, Zoller H, Oberhuber R, et al. The need to update endpoints and outcome analysis in the rapidly changing field of liver transplantation. Transplantation. 2021;106:938–949.
Fodor M, Cardini B, Peter W, et al. Static cold storage compared with normothermic machine perfusion of the liver and effect on ischaemic-type biliary lesions after transplantation: a propensity score-matched study. Br J Surg. 2021;108:1082–1089.
Fodor M, Woerdehoff A, Peter W, et al. Reassessment of relevance and predictive value of parameters indicating early graft dysfunction in liver transplantation: AST is a weak, but bilirubin and INR strong predictors of mortality. Front Surg. 2021;8:693288.
Cardini B, Fodor M, Hermann M, et al. Live confocal imaging as a novel tool to assess liver quality: insights from a murine model. Transplantation. 2020;104:2528–2537.
Cardini B, Oberhuber R, Fodor M, et al. Clinical implementation of prolonged liver preservation and monitoring through normothermic machine perfusion in liver transplantation. Transplantation. 2020;104:1917–1928.
Neuberger J, Callaghan C. Organ utilization—the next hurdle in transplantation? Transpl Int. 2020;33:1597–1609.
Ivanics T, Abreu P, De Martin E, et al. Changing trends in liver transplantation: challenges and solutions. Transplantation. 2021;105:743–756.
Ivanics T, Shwaartz C, Claasen MPAW, et al. Trends in indications and outcomes of liver transplantation in Canada: a multicenter retrospective study. Transpl Int. 2021;34:1444–1454.
Oniscu GC, Mehew J, Butler AJ, et al. Improved organ utilization and better transplant outcomes with in situ normothermic regional perfusion in controlled donation after circulatory death. Transplantation. 2023;107:438–448.
de Jong IEM, Bodewes SB, van Leeuwen OB, et al. Restoration of bile duct injury of donor livers during ex situ normothermic machine perfusion. Transplantation. 2023;107:e161–e172.
Ramos P, Williams P, Salinas J, et al. Abdominal organ preservation solutions in the age of machine perfusion. Transplantation. 2023;107:326–340.
van Rijn R, Schurink IJ, de Vries Y, et al.; DHOPE-DCD Trial Investigators. Hypothermic machine perfusion in liver transplantation—a randomized trial. N Engl J Med. 2021;384:1391–1401.
Diaspective Vision GmbH G. Hyperspectral imaging. 2021.
Fodor M, Hofmann J, Lanser L, et al. Hyperspectral imaging and machine perfusion in solid organ transplantation: clinical potentials of combining two novel technologies. J Clin Med. 2021;10:3838.
Holmer A, Marotz J, Wahl P, et al. Hyperspectral imaging in perfusion and wound diagnostics—methods and algorithms for the determination of tissue parameters. Biomed Tech (Berl). 2018;63:547–556.
Lu G, Fei B. Medical hyperspectral imaging: a review. J Biomed Opt. 2014;19:10901.
Moulla Y, Buchloh DC, Köhler H, et al. Hyperspectral Imaging (HSI)—a new tool to estimate the perfusion of upper abdominal organs during pancreatoduodenectomy. Cancers (Basel). 2021;13:2846.
Mühle R, Ernst H, Sobottka SB, et al. Workflow and hardware for intraoperative hyperspectral data acquisition in neurosurgery. Biomed Tech (Berl). [Epub ahead of print. July 25, 2020]. doi:10.1515/bmt-2019-0333.
doi: 10.1515/bmt-2019-0333
Mühle R, Markgraf W, Hilsmann A, et al. Comparison of different spectral cameras for image-guided organ transplantation. J Biomed Opt. 2021;26:076007.
Sucher R, Scheuermann U, Rademacher S, . Intraoperative reperfusion assessment of human pancreas allografts using hyperspectral imaging (HSI). Hepatobiliary Surg Nutr. 2022;11:67–77.
Sucher R, Athanasios A, Köhler H, et al. Hyperspectral imaging (HSI) in anatomic left liver resection. Int J Surg Case Rep. 2019;62:108–111.
Florian T, Wenke M, Marian G, et al. Hyperspectral imaging for monitoring oxygen saturation levels during normothermic kidney perfusion. J Sens Sens Syst. 2016;5:313–318.
Sucher E, Sucher R, Guice H, et al. Hyperspectral evaluation of the human liver during major resection. Annal Surg Open. 2022;3:e169.
Dietrich M, Özdemir B, Gruneberg D, et al. Hyperspectral imaging for the evaluation of microcirculatory tissue oxygenation and perfusion quality in haemorrhagic shock: a porcine study. Biomedicines. 2021;9:1829.
Sucher R, Wagner T, Köhler H, et al. Hyperspectral imaging (HSI) of human kidney allografts. Ann Surg. 2020;276:e48–e55.
Chalopin C, Nickel F, Pfahl A, et al. Künstliche Intelligenz und hyperspektrale Bildgebung zur bildgestützten Assistenz in der minimal-invasiven Chirurgie [Artificial intelligence and hyperspectral imaging for image-guided assistance in minimally invasive surgery]. Chirurgie (Heidelb). 2022;93:940–947.
Fodor M, Lanser L, Hofmann J, et al. Hyperspectral imaging as a tool for viability assessment during normothermic machine perfusion of human livers: a proof of concept pilot study. Transpl Int. 2022;35:10355.
Signoroni A, Savardi M, Baronio A, et al. Deep learning meets hyperspectral image analysis: a multidisciplinary review. J Imaging. 2019;5:52.
Litjens G, Kooi T, Bejnordi BE, et al. A survey on deep learning in medical image analysis. Med Image Anal. 2017;42:60–88.
Sommer F, Sun B, Fischer J, et al. Hyperspectral imaging during normothermic machine perfusion—a functional classification of ex vivo kidneys based on convolutional neural networks. Biomedicines. 2022;10:397.
Felli E, Al-Taher M, Collins T, et al. Automatic liver viability scoring with deep learning and hyperspectral imaging. Diagnostics (Basel). 2021;11:1527.
Kim HE, Cosa-Linan A, Santhanam N, et al. Transfer learning for medical image classification: a literature review. BMC Med Imaging. 2022;22:69.
Ebbehoj A, Thunbo M, Andersen OE, et al. Transfer learning for non-image data in clinical research: a scoping review. PLOS Digit Health. 2022;1:e0000014.
Brooke-Smith M, Figueras J, Ullah S, et al. Prospective evaluation of the International Study Group for Liver Surgery definition of bile leak after a liver resection and the role of routine operative drainage: an international multicentre study. HPB (Oxford). 2015;17:46–51.
Matton APM, de Vries Y, Burlage LC, et al. Biliary bicarbonate, pH, and glucose are suitable biomarkers of biliary viability during ex situ normothermic machine perfusion of human donor livers. Transplantation. 2019;103:1405–1413.
op den Dries S, Westerkamp AC, Karimian N, et al. Injury to peribiliary glands and vascular plexus before liver transplantation predicts formation of non-anastomotic biliary strictures. J Hepatol. 2014;60:1172–1179.
Senter-Zapata M, Khan AS, Subramanian T, et al. Patient and graft survival: biliary complications after liver transplantation. J Am Coll Surg. 2018;226:484–494.
Muller X, Schlegel A, Kron P, et al. Novel real-time prediction of liver graft function during hypothermic oxygenated machine perfusion before liver transplantation. Ann Surg. 2019;270:783–790.
Schlegel A, Kron P, Graf R, et al. Hypothermic oxygenated perfusion (HOPE) downregulates the immune response in a rat model of liver transplantation. Ann Surg. 2014;260:931–937.
Schlegel A, Dutkowski P. Impact of machine perfusion on biliary complications after liver transplantation. Int J Mol Sci . 2018;19:3567.
Schlegel A, Muller X, Kalisvaart M, et al. Outcomes of DCD liver transplantation using organs treated by hypothermic oxygenated perfusion before implantation. J Hepatol. 2019;70:50–57.
Schlegel A, Porte R, Dutkowski P. Protective mechanisms and current clinical evidence of hypothermic oxygenated machine perfusion (HOPE) in preventing post-transplant cholangiopathy. J Hepatol. 2022;76:1330–1347.
Schlegel A, Dutkowski P. Role of hypothermic machine perfusion in liver transplantation. Transpl Int. 2015;28:677–689.
Watson CJE, Kosmoliaptsis V, Pley C, et al. Observations on the ex situ perfusion of livers for transplantation. Am J Transplant. 2018;18:2005–2020.
Watson CJE, Jochmans I. From “Gut feeling” to objectivity: machine preservation of the liver as a tool to assess organ viability. Curr Transplant Rep. 2018;5:72–81.
Watson CJE, Hunt F, Messer S, et al. In situ normothermic perfusion of livers in controlled circulatory death donation may prevent ischemic cholangiopathy and improve graft survival. Am J Transplant. 2019;19:1745–1758.
Dietrich M, Marx S, von der Forst M, et al. Hyperspectral imaging for perioperative monitoring of microcirculatory tissue oxygenation and tissue water content in pancreatic surgery—an observational clinical pilot study. Perioper Med (Lond). 2021;10:42.
Dietrich M, Marx S, Bruckner T, et al. Bedside hyperspectral imaging for the evaluation of microcirculatory alterations in perioperative intensive care medicine: a study protocol for an observational clinical pilot study (HySpI-ICU). BMJ Open. 2020;10:e035742.
Dietrich M, Özdemir B, Gruneberg D, et al. Hyperspectral imaging for the evaluation of microcirculatory tissue oxygenation and perfusion quality in haemorrhagic shock: a porcine study. Biomedicines. 2021;9:1829.
Daeschlein G, Langner I, Wild T, et al. Hyperspectral imaging as a novel diagnostic tool in microcirculation of wounds. Clin Hemorheol Microcirc. 2017;67:467–474.
Shi HY, Lee KT, Lee HH, et al. Comparison of artificial neural network and logistic regression models for predicting in-hospital mortality after primary liver cancer surgery. PLoS One. 2012;7:e35781.
Tokuyasu T, Iwashita Y, Matsunobu Y, et al. Development of an artificial intelligence system using deep learning to indicate anatomical landmarks during laparoscopic cholecystectomy. Surg Endosc. 2021;35:1651–1658.
Meng L, Zhang Q, Bu S. Two-stage liver and tumor segmentation algorithm based on convolutional neural network. Diagnostics (Basel). 2021;11:1806.
Fabelo H, Halicek M, Ortega S, et al. Surgical aid visualization system for glioblastoma tumor identification based on deep learning and in-vivo hyperspectral images of human patients. Proc SPIE Int Soc Opt Eng. 2019;10951:1095110.