Learning More for Free - A Multi Task Learning Approach for Improved Pathology Classification in Capsule Endoscopy

The progress in Computer Aided Diagnosis (CADx) of Wireless Capsule Endoscopy (WCE) is thwarted by the lack of data. The inadequacy in richly representative healthy and abnormal conditions results in isolated analyses of pathologies, that can not handle realistic multipathology scenarios. In this work, we explore how to learn more for free, from limited data through solving a WCE multicentric, multi-pathology classification problem. Learning more implies to learning more than full supervision would allow with the same data. This is done by combining self supervision with full supervision, under multi task learning. Additionally, we draw inspiration from the Human Visual System (HVS) in designing self supervision tasks and investigate if seemingly ineffectual signals within the data itself can be exploited to gain performance, if so, which signals would be better than others. Further, we present our analysis of the high level features as a stepping stone towards more robust multi-pathology CADx in WCE. Code accompanying this work will be made available on github.

[1]  Rong Lin,et al.  Gastroenterologist-level Identification of Small Bowel Diseases and Normal Variants by Capsule Endoscopy Using a Deep-learning Model. , 2019, Gastroenterology.

[2]  Anastasios Koulaouzidis,et al.  Capsule endoscopy of the small bowel. , 2016, Annals of translational medicine.

[3]  Leland McInnes,et al.  UMAP: Uniform Manifold Approximation and Projection , 2018, J. Open Source Softw..

[4]  Eric Goldberg,et al.  Detection of Lesions During Capsule Endoscopy: Physician Performance Is Disappointing , 2012, The American Journal of Gastroenterology.

[5]  A. Atsawarungruangkit,et al.  Understanding deep learning in capsule endoscopy: Can artificial intelligence enhance clinical practice? , 2020 .

[6]  J. Saurin,et al.  CAD-CAP: a 25,000-image database serving the development of artificial intelligence for capsule endoscopy , 2020, Endoscopy International Open.

[7]  J. Kim,et al.  Current and Future Use of Esophageal Capsule Endoscopy , 2018, Clinical endoscopy.

[8]  Hélder P. Oliveira,et al.  Lesions Multiclass Classification in Endoscopic Capsule Frames , 2019, CENTERIS/ProjMAN/HCist.

[9]  Roberto Cipolla,et al.  Multi-task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[10]  Youngbae Hwang,et al.  Application of Artificial Intelligence in Capsule Endoscopy: Where Are We Now? , 2018, Clinical endoscopy.

[11]  Iasonas Kokkinos,et al.  UberNet: Training a Universal Convolutional Neural Network for Low-, Mid-, and High-Level Vision Using Diverse Datasets and Limited Memory , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[12]  Martial Hebert,et al.  Cross-Stitch Networks for Multi-task Learning , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[13]  Sule Yildirim Yayilgan,et al.  Stochastic Capsule Endoscopy Image Enhancement , 2018, J. Imaging.

[14]  Bonnie Webber,et al.  Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 2, Short Papers , 2017 .

[15]  Jordi Vitrià,et al.  Generic Feature Learning for Wireless Capsule Endoscopy Analysis , 2016, Comput. Biol. Medicine.

[16]  Javier Del Ser,et al.  Vision-based personalized Wireless Capsule Endoscopy for smart healthcare: Taxonomy, literature review, opportunities and challenges , 2020, Future Gener. Comput. Syst..

[17]  Rich Caruana,et al.  Multitask Learning: A Knowledge-Based Source of Inductive Bias , 1993, ICML.

[18]  Eyal Klang,et al.  Deep learning for wireless capsule endoscopy: a systematic review and meta-analysis. , 2020, Gastrointestinal endoscopy.

[19]  S. Syed,et al.  Potential for Standardization and Automation for Pathology and Endoscopy in Inflammatory Bowel Disease. , 2020, Inflammatory bowel diseases.

[20]  Joachim Bingel,et al.  Identifying beneficial task relations for multi-task learning in deep neural networks , 2017, EACL.

[21]  Sharib Ali,et al.  Endoscopy artifact detection (EAD 2019) challenge dataset , 2019, ArXiv.

[22]  Jonathan Baxter,et al.  A Bayesian/Information Theoretic Model of Learning to Learn via Multiple Task Sampling , 1997, Machine Learning.

[23]  Rich Caruana,et al.  Multitask Learning , 1997, Machine Learning.

[24]  Y. Yang.  The Future of Capsule Endoscopy: The Role of Artificial Intelligence and Other Technical Advancements , 2020, Clinical endoscopy.

[25]  Xiaodong Liu,et al.  Representation Learning Using Multi-Task Deep Neural Networks for Semantic Classification and Information Retrieval , 2015, NAACL.

[26]  Rich Caruana,et al.  Multitask Learning , 1998, Encyclopedia of Machine Learning and Data Mining.

[27]  Xiaoou Tang,et al.  Facial Landmark Detection by Deep Multi-task Learning , 2014, ECCV.

[28]  Dirk Hovy,et al.  Multitask Learning for Mental Health Conditions with Limited Social Media Data , 2017, EACL.

[29]  Jordi Vitrià,et al.  Using the Triplet Loss for Domain Adaptation in WCE , 2019, 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW).

[30]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[31]  Leland McInnes,et al.  UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction , 2018, ArXiv.