{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2024,3,1]],"date-time":"2024-03-01T20:30:15Z","timestamp":1709325015374},"reference-count":30,"publisher":"Springer Science and Business Media LLC","issue":"12","license":[{"start":{"date-parts":[[2021,9,4]],"date-time":"2021-09-04T00:00:00Z","timestamp":1630713600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2021,9,4]],"date-time":"2021-09-04T00:00:00Z","timestamp":1630713600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"DFG-RTG","award":["2260 BIOQIC"]},{"name":"DFG Priority Programme DFG-SPP 1798","award":["1446\/21"]},{"name":"MATH+ -Berlin Mathematics Research Center","award":["Project EF1x1"]},{"DOI":"10.13039\/501100002347","name":"Bundesministerium f\u00fcr Bildung und Forschung","doi-asserted-by":"publisher","award":["BIFOLD - Berlin Institute for the Foundations of Learning and Data ref.01IS18025A","BIFOLD - Berlin Institute for the Foundations of Learning and Data ref 01IS18037I"],"id":[{"id":"10.13039\/501100002347","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Int J CARS"],"published-print":{"date-parts":[[2021,12]]},"abstract":"Abstract<\/jats:title>\n Purpose<\/jats:title>\n The quantitative detection of failure modes is important for making deep neural networks reliable and usable at scale. We consider three examples for common failure modes in image reconstruction and demonstrate the potential of uncertainty quantification as a fine-grained alarm system.<\/jats:p>\n <\/jats:sec>\n Methods<\/jats:title>\n We propose a deterministic, modular and lightweight approach called Interval Neural Network (INN<\/jats:sc>) that produces fast and easy to interpret uncertainty scores for deep neural networks. Importantly, INN<\/jats:sc>s can be constructed post hoc for already trained prediction networks. We compare it against state-of-the-art baseline methods (MCDrop<\/jats:sc>, ProbOut<\/jats:sc>).<\/jats:p>\n <\/jats:sec>\n Results<\/jats:title>\n We demonstrate on controlled, synthetic inverse problems the capacity of INN<\/jats:sc>s to capture uncertainty due to noise as well as directional error information. On a real-world inverse problem with human CT scans, we can show that INN<\/jats:sc>s produce uncertainty scores which improve the detection of all considered failure modes compared to the baseline methods.<\/jats:p>\n <\/jats:sec>\n Conclusion<\/jats:title>\n Interval Neural Networks offer a promising tool to expose weaknesses of deep image reconstruction models and ultimately make them more reliable. The fact that they can be applied post hoc to equip already trained deep neural network models with uncertainty scores makes them particularly interesting for deployment.<\/jats:p>\n <\/jats:sec>","DOI":"10.1007\/s11548-021-02482-2","type":"journal-article","created":{"date-parts":[[2021,9,4]],"date-time":"2021-09-04T11:02:26Z","timestamp":1630753346000},"page":"2089-2097","update-policy":"http:\/\/dx.doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":2,"title":["Detecting failure modes in image reconstructions with interval neural network uncertainty"],"prefix":"10.1007","volume":"16","author":[{"ORCID":"http:\/\/orcid.org\/0000-0002-1379-8627","authenticated-orcid":false,"given":"Luis","family":"Oala","sequence":"first","affiliation":[]},{"given":"Cosmas","family":"Hei\u00df","sequence":"additional","affiliation":[]},{"given":"Jan","family":"Macdonald","sequence":"additional","affiliation":[]},{"given":"Maximilian","family":"M\u00e4rz","sequence":"additional","affiliation":[]},{"given":"Gitta","family":"Kutyniok","sequence":"additional","affiliation":[]},{"ORCID":"http:\/\/orcid.org\/0000-0002-6283-3265","authenticated-orcid":false,"given":"Wojciech","family":"Samek","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2021,9,4]]},"reference":[{"issue":"48","key":"2482_CR1","doi-asserted-by":"publisher","first-page":"30088","DOI":"10.1073\/pnas.1907377117","volume":"117","author":"V Antun","year":"2020","unstructured":"Antun V, Renna F, Poon C, Adcock B, Hansen AC (2020) On instabilities of deep learning in image reconstruction and the potential costs of AI. Proc Natl Acad Sci 117(48):30088\u201330095. https:\/\/doi.org\/10.1073\/pnas.1907377117","journal-title":"Proc Natl Acad Sci"},{"key":"2482_CR2","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1017\/S0962492919000059","volume":"28","author":"S Arridge","year":"2019","unstructured":"Arridge S, Maass P, \u00d6ktem O, Sch\u00f6nlieb CB (2019) Solving inverse problems using data-driven models. Acta Numer 28:1\u2013174","journal-title":"Acta Numer"},{"key":"2482_CR3","unstructured":"Barber D, Bishop C (1998) Ensemble learning in Bayesian neural networks. In: Generalization in neural networks and machine learning. Springer, pp 215\u2013237"},{"issue":"5","key":"2482_CR4","doi-asserted-by":"publisher","first-page":"1190","DOI":"10.1137\/0916069","volume":"16","author":"RH Byrd","year":"1995","unstructured":"Byrd RH, Lu P, Nocedal J, Zhu C (1995) A limited memory algorithm for bound constrained optimization. SIAM J Sci Comput 16(5):1190\u20131208","journal-title":"SIAM J Sci Comput"},{"key":"2482_CR5","doi-asserted-by":"crossref","unstructured":"Carlini N, Wagner D (2017) Adversarial examples are not easily detected: bypassing ten detection methods. In: Proceedings of the 10th ACM workshop on artificial intelligence and security, pp 3\u201314","DOI":"10.1145\/3128572.3140444"},{"key":"2482_CR6","first-page":"877","volume":"1","author":"JS Denker","year":"1987","unstructured":"Denker JS, Schwartz DB, Wittner BS, Solla SA, Howard RE, Jackel LD, Hopfield JJ (1987) Large automatic learning, rule extraction, and generalization. Complex Syst 1:877\u2013922","journal-title":"Complex Syst"},{"issue":"1","key":"2482_CR7","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s11704-018-8900-4","volume":"13","author":"TG Dietterich","year":"2019","unstructured":"Dietterich TG (2019) Robust artificial intelligence and robust human organizations. Front Comput Sci 13(1):1\u20133","journal-title":"Front Comput Sci"},{"key":"2482_CR8","doi-asserted-by":"publisher","DOI":"10.1007\/978-0-8176-4948-7","volume-title":"A mathematical introduction to compressive sensing","author":"S Foucart","year":"2013","unstructured":"Foucart S, Rauhut H (2013) A mathematical introduction to compressive sensing. Applied and Numerical Harmonic Analysis, Birkh\u00e4user"},{"issue":"2","key":"2482_CR9","doi-asserted-by":"publisher","first-page":"455","DOI":"10.1093\/imaiai\/iaaa017","volume":"10","author":"R Foygel Barber","year":"2020","unstructured":"Foygel Barber R, Cand\u00e8s EJ, Ramdas A, Tibshirani RJ (2020) The limits of distribution-free conditional predictive inference. Inf Inference J IMA 10(2):455\u2013482. https:\/\/doi.org\/10.1093\/imaiai\/iaaa017","journal-title":"Inf Inference J IMA"},{"key":"2482_CR10","first-page":"1050","volume-title":"Proceedings of The 33rd international conference on machine learning, proceedings of machine learning research","author":"Y Gal","year":"2016","unstructured":"Gal Y, Ghahramani Z (2016) Dropout as a Bayesian approximation: representing model uncertainty in deep learning. In: Balcan MF, Weinberger KQ (eds) Proceedings of The 33rd international conference on machine learning, proceedings of machine learning research, vol 48. PMLR, New York, pp 1050\u20131059"},{"key":"2482_CR11","doi-asserted-by":"crossref","unstructured":"Garczarczyk Z (2000) Interval neural networks. In: 2000 IEEE international symposium on circuits and systems. Emerging technologies for the 21st Century. Proceedings (IEEE Cat No.00CH36353), vol\u00a03. Presses Polytech. Univ. Romandes, Geneva, pp 567\u2013570","DOI":"10.1109\/ISCAS.2000.856123"},{"key":"2482_CR12","doi-asserted-by":"crossref","unstructured":"Gast J, Roth S (2018) Lightweight probabilistic deep networks. In: 2018 IEEE\/CVF conference on computer vision and pattern recognition, pp 3369\u20133378","DOI":"10.1109\/CVPR.2018.00355"},{"key":"2482_CR13","doi-asserted-by":"publisher","first-page":"145","DOI":"10.1007\/978-3-030-00928-1_17","volume-title":"Medical image computing and computer assisted intervention\u2014MICCAI 2018","author":"Y Huang","year":"2018","unstructured":"Huang Y, W\u00fcrfl T, Breininger K, Liu L, Lauritsch G, Maier A (2018) Some investigations on robustness of deep learning in limited angle tomography. In: Frangi AF, Schnabel JA, Davatzikos C, Alberola-L\u00f3pez C, Fichtinger G (eds) Medical image computing and computer assisted intervention\u2014MICCAI 2018. Springer, Cham, pp 145\u2013153"},{"key":"2482_CR14","doi-asserted-by":"publisher","first-page":"4509","DOI":"10.1109\/TIP.2017.2713099","volume":"26","author":"KH Jin","year":"2017","unstructured":"Jin KH, McCann MT, Froustey E, Unser M (2017) Deep convolutional neural network for inverse problems in imaging. IEEE Trans Image Process 26:4509\u20134522","journal-title":"IEEE Trans Image Process"},{"key":"2482_CR15","unstructured":"Kingma DP, Ba J (2015) Adam: a method for stochastic optimization. In: Bengio Y, LeCun Y (eds) 3rd international conference on learning representations, ICLR 2015, San Diego, CA, USA, May 7\u20139, 2015, conference track proceedings"},{"key":"2482_CR16","unstructured":"Kingma DP, Salimans T, Welling M (2015) Variational dropout and the local reparameterization trick. In: Proceedings of the 28th international conference on neural information processing systems\u2014Volume 2, NIPS\u201915. MIT Press, Cambridge, pp 2575\u20132583"},{"issue":"4","key":"2482_CR17","doi-asserted-by":"publisher","first-page":"817","DOI":"10.1007\/s00521-015-2109-3","volume":"28","author":"PA Kowalski","year":"2017","unstructured":"Kowalski PA, Kulczycki P (2017) Interval probabilistic neural network. Neural Comput Appl 28(4):817\u2013834","journal-title":"Neural Comput Appl"},{"key":"2482_CR18","first-page":"6402","volume-title":"Advances in neural information processing systems","author":"B Lakshminarayanan","year":"2017","unstructured":"Lakshminarayanan B, Pritzel A, Blundell C (2017) Simple and scalable predictive uncertainty estimation using deep ensembles. In: Guyon I, Luxburg UV, Bengio S, Wallach H, Fergus R, Vishwanathan S, Garnett R (eds) Advances in neural information processing systems, vol 30. Curran Associates Inc, New York, pp 6402\u20136413"},{"key":"2482_CR19","unstructured":"MacKay DJC (1992) Bayesian methods for adaptive models. PhD thesis, California Institute of Technology"},{"key":"2482_CR20","doi-asserted-by":"crossref","unstructured":"McCollough C (2016) Tu-fg-207a-04: overview of the low dose CT grand challenge. Med Phys 43(6 Part 35):3759\u20133760","DOI":"10.1118\/1.4957556"},{"key":"2482_CR21","doi-asserted-by":"publisher","DOI":"10.1137\/1.9780898719284","volume-title":"The mathematics of computerized tomography","author":"F Natterer","year":"2001","unstructured":"Natterer F (2001) The mathematics of computerized tomography. SIAM, Philadelphia"},{"key":"2482_CR22","doi-asserted-by":"publisher","unstructured":"Nix DA, Weigend AS (1994) Estimating the mean and variance of the target probability distribution. In: Proceedings of 1994 IEEE international conference on neural networks (ICNN\u201994), vol\u00a01, pp 55\u201360. https:\/\/doi.org\/10.1109\/ICNN.1994.374138","DOI":"10.1109\/ICNN.1994.374138"},{"key":"2482_CR23","unstructured":"Oala L, Fehr J, Gilli L, Balachandran P, Leite AW, Calderon-Ramirez S, Li DX, Nobis G, Alvarado EAM, Jaramillo-Gutierrez G, Matek C, Shroff A, Kherif F, Sanguinetti B, Wiegand T (2020) Ml4h auditing: from paper to practice. In: Proceedings of the machine learning for health NeurIPS workshop, proceedings of machine learning research, vol 136. PMLR, pp 280\u2013317. http:\/\/proceedings.mlr.press\/v136\/oala20a.html"},{"issue":"12","key":"2482_CR24","doi-asserted-by":"publisher","first-page":"5377","DOI":"10.1109\/TNNLS.2020.2966745","volume":"31","author":"F Rodrigues","year":"2020","unstructured":"Rodrigues F, Pereira FC (2020) Beyond expectation: deep joint mean and quantile regression for spatiotemporal problems. IEEE Trans Neural Netw Learn Syst 31(12):5377\u20135389","journal-title":"IEEE Trans Neural Netw Learn Syst"},{"key":"2482_CR25","first-page":"234","volume-title":"Medical image computing and computer-assisted intervention\u2013MICCAI 2015, Lecture Notes in Computer Science","author":"O Ronneberger","year":"2015","unstructured":"Ronneberger O, Fischer P, Brox T (2015) U-Net: convolutional networks for biomedical image segmentation. In: Navab N, Hornegger J, Wells WM, Frangi AF (eds) Medical image computing and computer-assisted intervention\u2013MICCAI 2015, Lecture Notes in Computer Science. Springer, Berlin, pp 234\u2013241"},{"issue":"388","key":"2482_CR26","doi-asserted-by":"publisher","first-page":"871","DOI":"10.1080\/01621459.1984.10477105","volume":"79","author":"PJ Rousseeuw","year":"1984","unstructured":"Rousseeuw PJ (1984) Least median of squares regression. J Am Stat Assoc 79(388):871\u2013880","journal-title":"J Am Stat Assoc"},{"issue":"5","key":"2482_CR27","doi-asserted-by":"publisher","first-page":"756","DOI":"10.1109\/JPROC.2021.3052449","volume":"109","author":"L Ruff","year":"2021","unstructured":"Ruff L, Kauffmann JR, Vandermeulen RA, Montavon G, Samek W, Kloft M, Dietterich TG, M\u00fcller KR (2021) A unifying review of deep and shallow anomaly detection. Proc IEEE 109(5):756\u2013795. https:\/\/doi.org\/10.1109\/JPROC.2021.3052449","journal-title":"Proc IEEE"},{"issue":"1","key":"2482_CR28","doi-asserted-by":"publisher","first-page":"45","DOI":"10.1023\/B:MACH.0000008084.60811.49","volume":"54","author":"DMJ Tax","year":"2004","unstructured":"Tax DMJ, Duin RPW (2004) Support vector data description. Mach Learn 54(1):45\u201366","journal-title":"Mach Learn"},{"key":"2482_CR29","unstructured":"Wang H, Xingjian S, Yeung DY (2016) Natural-parameter networks: a class of probabilistic neural networks. In: Advances in neural information processing systems, pp 118\u2013126"},{"key":"2482_CR30","doi-asserted-by":"publisher","unstructured":"Yang D, Wu W (2012) A smoothing interval neural network. Discrete Dyn Nat Soc. https:\/\/doi.org\/10.1155\/2012\/456919","DOI":"10.1155\/2012\/456919"}],"container-title":["International Journal of Computer Assisted Radiology and Surgery"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11548-021-02482-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s11548-021-02482-2\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11548-021-02482-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2021,11,25]],"date-time":"2021-11-25T11:30:56Z","timestamp":1637839856000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s11548-021-02482-2"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,9,4]]},"references-count":30,"journal-issue":{"issue":"12","published-print":{"date-parts":[[2021,12]]}},"alternative-id":["2482"],"URL":"https:\/\/doi.org\/10.1007\/s11548-021-02482-2","relation":{},"ISSN":["1861-6410","1861-6429"],"issn-type":[{"value":"1861-6410","type":"print"},{"value":"1861-6429","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,9,4]]},"assertion":[{"value":"8 April 2021","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"10 August 2021","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"4 September 2021","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"L.O. co-chairs the DAISAM working group at the ITU\/WHO Focus Group AI4H.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}},{"value":"This article does not contain any studies with human participants or animals performed by any of the authors.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval"}},{"value":"No new patient data were acquired as part of this work; public data were used from ExternalRef removed.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Informed consent"}}]}}