{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,17]],"date-time":"2026-08-17T15:43:06Z","timestamp":1786981386059,"version":"build-2736575974"},"reference-count":33,"publisher":"Springer Science and Business Media LLC","issue":"4","license":[{"start":{"date-parts":[[2024,3,26]],"date-time":"2024-03-26T00:00:00Z","timestamp":1711411200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2024,3,26]],"date-time":"2024-03-26T00:00:00Z","timestamp":1711411200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100003246","name":"Nederlandse Organisatie voor Wetenschappelijk Onderzoek","doi-asserted-by":"crossref","award":["NWA.1160.18.038"],"award-info":[{"award-number":["NWA.1160.18.038"]}],"id":[{"id":"10.13039\/501100003246","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["J Pharmacokinet Pharmacodyn"],"published-print":{"date-parts":[[2024,8]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Conventional pharmacokinetic (PK) models contain several useful inductive biases guiding model convergence to more realistic predictions of drug concentrations. Implementing similar biases in standard neural networks can be challenging, but might be fundamental for model robustness and predictive performance. In this study, we build on the deep compartment model (DCM) architecture by introducing constraints that guide the model to explore more physiologically realistic solutions. Using a simulation study, we show that constraints improve robustness in sparse data settings. Additionally, predicted concentration\u2013time curves took on more realistic shapes compared to unconstrained models. Next, we propose the use of multi-branch networks, where each covariate can be connected to specific PK parameters, to\u00a0reduce the propensity of models to\u00a0learn spurious effects. Another benefit of this architecture is that covariate effects are isolated, enabling model interpretability through the visualization of learned functions. We show that all models were sensitive to learning false effects when trained in the presence of unimportant covariates, indicating the importance of selecting an appropriate set of covariates to link to the PK parameters. Finally, we compared the predictive performance of the constrained models to previous relevant population PK models on a real-world data set of 69 haemophilia A patients. Here, constrained models obtained higher accuracy compared to the standard DCM, with the multi-branch network outperforming previous PK models. We conclude that physiological-based constraints can improve model robustness. We describe an interpretable architecture which aids model trust, which will be key for the adoption of machine learning-based models in clinical practice.<\/jats:p>","DOI":"10.1007\/s10928-024-09906-x","type":"journal-article","created":{"date-parts":[[2024,3,26]],"date-time":"2024-03-26T16:02:01Z","timestamp":1711468921000},"page":"355-366","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":7,"title":["On inductive biases for the robust and interpretable prediction of drug concentrations using deep compartment models"],"prefix":"10.1007","volume":"51","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-0719-0831","authenticated-orcid":false,"given":"Alexander","family":"Janssen","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Frank C.","family":"Bennis","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marjon H.","family":"Cnossen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ron A. A.","family":"Math\u00f4t","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"name":"for the OPTI-CLOT Study Group and SYMPHONY Consortium","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2024,3,26]]},"reference":[{"key":"9906_CR1","first-page":"612","volume":"119","author":"S Bj\u00f6rkman","year":"2012","unstructured":"Bj\u00f6rkman S, Oh M, Spotts G et al (2012) Population pharmacokinetics of recombinant factor VIII: the relationships of pharmacokinetics to age and body weight. Blood J Am Soc Hematol 119:612\u2013618","journal-title":"Blood J Am Soc Hematol"},{"key":"9906_CR2","doi-asserted-by":"publisher","first-page":"326","DOI":"10.1097\/FTD.0000000000000004","volume":"36","author":"NA Lankheet","year":"2014","unstructured":"Lankheet NA, Knapen LM, Schellens JH et al (2014) Plasma concentrations of tyrosine kinase inhibitors imatinib, erlotinib, and sunitinib in routine clinical outpatient cancer care. Ther Drug Monit 36:326\u2013334","journal-title":"Ther Drug Monit"},{"key":"9906_CR3","doi-asserted-by":"publisher","first-page":"1072","DOI":"10.1093\/cid\/ciu027","volume":"58","author":"JA Roberts","year":"2014","unstructured":"Roberts JA, Paul SK, Akova M et al (2014) DALI: defining antibiotic levels in intensive care unit patients: are current \u03b2-lactam antibiotic doses sufficient for critically ill patients? Clin Infect Dis 58:1072\u20131083","journal-title":"Clin Infect Dis"},{"key":"9906_CR4","doi-asserted-by":"publisher","first-page":"e492","DOI":"10.1016\/S2352-3026(21)00135-6","volume":"8","author":"I van Moort","year":"2021","unstructured":"van Moort I, Preijers T, Bukkems LH et al (2021) Perioperative pharmacokinetic-guided factor VIII concentrate dosing in haemophilia (OPTI-CLOT trial): an open-label, multicentre, randomised, controlled trial. Lancet Haematol 8:e492\u2013e502","journal-title":"Lancet Haematol"},{"key":"9906_CR5","doi-asserted-by":"publisher","first-page":"185","DOI":"10.1146\/annurev.pa.32.040192.001153","volume":"32","author":"LB Sheiner","year":"1992","unstructured":"Sheiner LB, Ludden T (1992) Population pharmacokinetics\/dynamics. Annu Rev Pharmacol Toxicol 32:185\u2013209","journal-title":"Annu Rev Pharmacol Toxicol"},{"key":"9906_CR6","doi-asserted-by":"publisher","first-page":"249","DOI":"10.1016\/S0169-409X(01)00118-1","volume":"48","author":"M Holz","year":"2001","unstructured":"Holz M, Fahr A (2001) Compartment modeling. Adv Drug Deliv Rev 48:249\u2013264","journal-title":"Adv Drug Deliv Rev"},{"key":"9906_CR7","doi-asserted-by":"publisher","first-page":"1814","DOI":"10.3390\/pharmaceutics14091814","volume":"14","author":"A Janssen","year":"2022","unstructured":"Janssen A, Bennis FC, Math\u00f4t RA (2022) Adoption of machine learning in pharmacometrics: an overview of recent implementations and their considerations. Pharmaceutics 14:1814","journal-title":"Pharmaceutics"},{"key":"9906_CR8","doi-asserted-by":"publisher","DOI":"10.1016\/j.isci.2021.102804","volume":"24","author":"J Lu","year":"2021","unstructured":"Lu J, Deng K, Zhang X et al (2021) Neural-ODE for pharmacokinetics modeling and its advantage to alternative machine learning models in predicting new dosing regimens. Iscience 24:102804","journal-title":"Iscience"},{"key":"9906_CR9","doi-asserted-by":"publisher","first-page":"934","DOI":"10.1002\/psp4.12808","volume":"11","author":"A Janssen","year":"2022","unstructured":"Janssen A, Leebeek FW, Cnossen MH et al (2022) Deep compartment models: a deep learning approach for the reliable prediction of time-series data in pharmacokinetic modeling. Pharmacomet Syst Pharmacol 11:934\u2013945","journal-title":"Pharmacomet Syst Pharmacol"},{"key":"9906_CR10","doi-asserted-by":"publisher","DOI":"10.21203\/rs.3.rs-2428689\/v1","author":"DS Br\u00e4m","year":"2023","unstructured":"Br\u00e4m DS, Nahum U, Schropp J et al (2023) Neural ODEs in pharmacokinetics: concepts and applications. J Pharmacokinet Pharmacodyn. https:\/\/doi.org\/10.21203\/rs.3.rs-2428689\/v1","journal-title":"J Pharmacokinet Pharmacodyn"},{"key":"9906_CR11","first-page":"11364","volume":"34","author":"Z Qian","year":"2021","unstructured":"Qian Z, Zame W, Fleuren L et al (2021) Integrating expert ODEs into neural ODEs: pharmacology and disease progression. Adv Neural Inf Process Syst 34:11364\u201311383","journal-title":"Adv Neural Inf Process Syst"},{"key":"9906_CR12","doi-asserted-by":"publisher","first-page":"223","DOI":"10.1172\/JCI109120","volume":"62","author":"J Over","year":"1978","unstructured":"Over J, Sixma J, Bruine M et al (1978) Survival of 125iodine-labeled factor VIII in normals and patients with classic hemophilia. Observations on the heterogeneity of human factor VIII. J Clin Investig 62:223\u2013234","journal-title":"J Clin Investig"},{"key":"9906_CR13","doi-asserted-by":"publisher","first-page":"289","DOI":"10.1159\/000217222","volume":"26","author":"DA Noe","year":"1996","unstructured":"Noe DA (1996) A mathematical model of coagulation factor VIII kinetics. Pathophysiol Haemost Thromb 26:289\u2013303","journal-title":"Pathophysiol Haemost Thromb"},{"key":"9906_CR14","doi-asserted-by":"publisher","first-page":"385","DOI":"10.2165\/00003088-199222050-00005","volume":"22","author":"S Bj\u00f6rkman","year":"1992","unstructured":"Bj\u00f6rkman S, Carlsson M, Berntorp E, Stenberg P (1992) Pharmacokinetics of factor VIII in humans: obtaining clinically relevant data from comparative studies. Clin Pharmacokinet 22:385\u2013395","journal-title":"Clin Pharmacokinet"},{"key":"9906_CR15","doi-asserted-by":"publisher","first-page":"195","DOI":"10.1152\/jappl.1962.17.2.195","volume":"17","author":"MJ Yiengst","year":"1962","unstructured":"Yiengst MJ, Shock NW (1962) Blood and plasma volume in adult males. J Appl Physiol 17:195\u2013198","journal-title":"J Appl Physiol"},{"key":"9906_CR16","doi-asserted-by":"publisher","first-page":"1525","DOI":"10.1111\/bcp.13577","volume":"84","author":"XS Xu","year":"2018","unstructured":"Xu XS, Yuan M, Zhu H et al (2018) Full covariate modelling approach in population pharmacokinetics: understanding the underlying hypothesis tests and implications of multiplicity. Br J Clin Pharmacol 84:1525\u20131534","journal-title":"Br J Clin Pharmacol"},{"key":"9906_CR17","unstructured":"Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey Data. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 2009. https:\/\/wwwn.cdc.gov\/nchs\/nhanes\/continuousnhanes\/default.aspx?BeginYear=2009"},{"key":"9906_CR18","doi-asserted-by":"publisher","first-page":"245","DOI":"10.1007\/s40262-019-00809-6","volume":"59","author":"P Chelle","year":"2020","unstructured":"Chelle P, Yeung CH, Croteau SE et al (2020) Development and validation of a population-pharmacokinetic model for rurioctacog alfa pegol (adynovate): a report on behalf of the WAPPS-hemo investigators ad hoc subgroup. Clin Pharmacokinet 59:245\u2013256","journal-title":"Clin Pharmacokinet"},{"key":"9906_CR19","doi-asserted-by":"publisher","first-page":"1169","DOI":"10.1007\/s40262-015-0277-z","volume":"54","author":"HS Al-Sallami","year":"2015","unstructured":"Al-Sallami HS, Goulding A, Grant A et al (2015) Prediction of fat-free mass in children. Clin Pharmacokinet 54:1169\u20131178","journal-title":"Clin Pharmacokinet"},{"key":"9906_CR20","unstructured":"Ramachandran P, Zoph B, Le QV (2017) Searching for activation functions. arXiv:171005941"},{"key":"9906_CR21","unstructured":"Kingma DP, Ba J (2014) Adam: a method for stochastic optimization. arXiv:14126980."},{"key":"9906_CR22","doi-asserted-by":"publisher","first-page":"989","DOI":"10.1007\/s00228-009-0676-x","volume":"65","author":"S Bj\u00f6rkman","year":"2009","unstructured":"Bj\u00f6rkman S, Folkesson A, J\u00f6nsson S (2009) Pharmacokinetics and dose requirements of factor VIII over the age range 3\u201374 years: a population analysis based on 50 patients with long-term prophylactic treatment for haemophilia A. Eur J Clin Pharmacol 65:989\u2013998","journal-title":"Eur J Clin Pharmacol"},{"key":"9906_CR23","doi-asserted-by":"publisher","first-page":"411","DOI":"10.1007\/s10928-019-09634-7","volume":"46","author":"A McEneny-King","year":"2019","unstructured":"McEneny-King A, Chelle P, Foster G et al (2019) Development and evaluation of a generic population pharmacokinetic model for standard half-life factor VIII for use in dose individualization. J Pharmacokinet Pharmacodyn 46:411\u2013426","journal-title":"J Pharmacokinet Pharmacodyn"},{"key":"9906_CR24","doi-asserted-by":"publisher","first-page":"343","DOI":"10.1046\/j.1365-3148.2001.00315.x","volume":"11","author":"O\u2019donnell J, Laffan M","year":"2001","unstructured":"O\u2019donnell J, Laffan M (2001) The relationship between ABO histo-blood group, factor VIII and von willebrand factor. Transfus Med 11:343\u2013351","journal-title":"Transfus Med"},{"key":"9906_CR25","doi-asserted-by":"publisher","first-page":"1282800","DOI":"10.3389\/frai.2023.1282800","volume":"6","author":"Z He","year":"2023","unstructured":"He Z, Zhang R, Diallo G, Huang Z, Glicksberg BS (2023) Editorial: explainable artificial intelligence for critical healthcare applications. Front Artif Intell 6:1282800. https:\/\/doi.org\/10.3389\/frai.2023.1282800","journal-title":"Front Artif Intell"},{"key":"9906_CR26","doi-asserted-by":"crossref","unstructured":"Amann J, Blasimme A, Vayena E, Frey D, Madai VI; Precise4Q Consortium (2020) Explainability for artificial intelligence in healthcare: a multidisciplinary perspective. BMC Med Inform Decis Mak 20(1):310","DOI":"10.1186\/s12911-020-01332-6"},{"key":"9906_CR27","doi-asserted-by":"publisher","DOI":"10.1007\/s10928-023-09886-4","author":"DS Br\u00e4m","year":"2023","unstructured":"Br\u00e4m DS, Nahum U, Schropp J, Pfister M, Koch G (2023) Low-dimensional neural ODEs and their application in pharmacokinetics. J Pharmacokinet Pharmacodyn. https:\/\/doi.org\/10.1007\/s10928-023-09886-4","journal-title":"J Pharmacokinet Pharmacodyn"},{"key":"9906_CR28","first-page":"26598","volume":"34","author":"I Kumar","year":"2021","unstructured":"Kumar I et al (2021) Shapley residuals: quantifying the limits of the Shapley value for explanations. Adv Neural Inf Process Syst 34:26598\u201326608","journal-title":"Adv Neural Inf Process Syst"},{"key":"9906_CR29","doi-asserted-by":"crossref","unstructured":"Rackauckas C, Ma Y, Martensen J, Warner C, Zubov K, Supekar R et al (2020) Universal differential equations for scientific machine learning. arXiv:2001.04385","DOI":"10.21203\/rs.3.rs-55125\/v1"},{"issue":"8","key":"9906_CR30","doi-asserted-by":"publisher","first-page":"1530","DOI":"10.3390\/pharmaceutics14081530","volume":"14","author":"Keutzer L, You H, Farnoud A, Nyberg J, Wicha SG, Maher-Edwards G, et al","year":"2022","unstructured":"Keutzer L, You H, Farnoud A, Nyberg J, Wicha SG, Maher-Edwards G, et al (2022) Machine learning and pharmacometrics for prediction of pharmacokinetic data: differences, similarities and challenges illustrated with rifampicin. Pharmaceutics 14(8):1530","journal-title":"Pharmaceutics"},{"issue":"4","key":"9906_CR31","doi-asserted-by":"publisher","first-page":"597","DOI":"10.1007\/s10928-021-09757-w","volume":"48","author":"E Sibieude","year":"2021","unstructured":"Sibieude E, Khandelwal A, Hesthaven JS, Girard P, Terranova N (2021) Fast screening of covariates in population models empowered by machine learning. J Pharmacokinet Pharmacodyn 48(4):597\u2013609","journal-title":"J Pharmacokinet Pharmacodyn"},{"key":"9906_CR32","doi-asserted-by":"publisher","DOI":"10.1007\/s10928-023-09887-3","author":"Y Wahlquist","year":"2023","unstructured":"Wahlquist Y, Sundell J, Soltesz K (2023) Learning pharmacometric covariate model structures with symbolic regression networks. J Pharmacokinet Pharmacodyn. https:\/\/doi.org\/10.1007\/s10928-023-09887-3","journal-title":"J Pharmacokinet Pharmacodyn"},{"key":"9906_CR33","doi-asserted-by":"crossref","unstructured":"Janssen A, Hoogendoorn M, Cnossen MH, Math\u00f4t RA, OPTI\u2010CLOT Study Group and SYMPHONY Consortium, Cnossen MH (2022) Application of SHAP values for inferring the optimal functional form of covariates in pharmacokinetic modeling. Pharmacomet Syst Pharmacol 11(8):1100\u20131110.","DOI":"10.1002\/psp4.12828"}],"container-title":["Journal of Pharmacokinetics and Pharmacodynamics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10928-024-09906-x.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10928-024-09906-x\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10928-024-09906-x.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,7,17]],"date-time":"2024-07-17T16:27:07Z","timestamp":1721233627000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10928-024-09906-x"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,3,26]]},"references-count":33,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2024,8]]}},"alternative-id":["9906"],"URL":"https:\/\/doi.org\/10.1007\/s10928-024-09906-x","relation":{},"ISSN":["1567-567X","1573-8744"],"issn-type":[{"value":"1567-567X","type":"print"},{"value":"1573-8744","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,3,26]]},"assertion":[{"value":"17 November 2023","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"9 February 2024","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"26 March 2024","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"M.H.C.'s institution has received investigator-initiated research and travel grants as well as speaker fees over the years from the Netherlands Organization for Scientific Research (NWO) and Netherlands National research Agenda (NWA), the Netherlands Organization for Health Research and Development (ZonMw), the Dutch Innovatiefonds Zorgverzekeraars, Baxter\/Baxalta\/Shire\/Takeda, Pfizer, Bayer Schering Pharma, CSL Behring, Sobi Biogen, Novo Nordisk, Novartis and Nordic Pharma and for serving as a steering board member for Roche, Bayer and Novartis for which fees go to the Erasmus MC as an institution.  R.A.A.M. has received grants from governmental and societal research institutes such as NWO, ZonMW, Dutch Kidney Foundation and Innovation Fund and unrestricted investigator research grants from Baxter\/Baxalta\/ Shire\/Takeda, Bayer, CSL Behring, Sobi and CelltrionHC. He has served as advisor for Bayer, CSL Behring, Merck Sharp & Dohme and Baxter\/Baxalta\/Shire\/Takeda. All grants and fees paid to the institution.  A.J., and F.C.B. have no conflicts of interest to declare.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}]}}