TrainX – Named Entity Linking with Active Sampling and Bi-Encoders

Abstract: We demonstrate TrainX, a system for Named Entity Linking for medical experts. It combines state-of-the-art entity recognition and linking architectures, such as Flair and fine-tuned Bi-Encoders based on BERT, with an easy-to-use interface for healthcare professionals. We support medical experts in annotating training data by using active sampling strategies to forward informative samples to the annotator. We demonstrate that our model is capable of linking against large knowledge bases, such as UMLS (3.6 million entities), and supporting zero-shot cases, where the linker has never seen the entity before. Those zero-shot capabilities help to mitigate the problem of rare and expensive training data that is a common issue in the medical domain.

06/07/2020

Fenglin Liu, Chenyu You, Xian Wu and
Shen Ge, Sheng wang, Xu Sun

David A. Wood, Jeremy Lynch, Sina Kafiabadi and
Emily Guilhem, Aisha Al Busaidi, Antanas Montvila, Thomas Varsavsky, Juveria Siddiqui, Naveen Gadapa, Matthew Townend, Martin Kiik, Keena Patel, Gareth Barker, Sebastian Ourselin, James H. Cole, Thomas C. Booth

Chest X-ray analysis, disease classification, disease localization, triplet learning, region verification, multi-class activation map.

5:45

02/02/2021

GRASP: Generic Framework for Health Status Representation Learning Based on Incorporating Knowledge from Similar Patients

Ahmad B. Qasim, Ivan Ezhov, Suprosanna Shit and
Oliver Schoppe, Johannes C. Paetzold, Anjany Sekuboyina, Florian Kofler, Jana Lipkova, Hongwei Li, Bjoern Menze

compression, data-free, quantization, fine-tunning, knowledge-distillation, feature-visulization, computer-vision, statistics, similarity-measure, privacy

0:51

19/10/2020

continuous improvement methodology, test automation, knowledge base, benchmarking, differential diagnosis, clinical decision support, simulation

11:13

06/07/2020

Students Need More Attention: BERT-based Attention Model for Small Data with Application to Automatic Patient Message Triage

Active Latent Space Shape Model: A Bayesian Treatment of Shape Model Adaptation With an Application to Psoriatic Arthritis Radiographs

fast online adaptation, meta-learning, cardiac motion estimation, mismatched distribution, feature tracking, cardiac mri, optical flow

0:59

07/08/2020

Self-Supervised Pretraining with DICOM metadata in Ultrasound Imaging

Szu-Yen Hu, Shuhang Wang, Wei-Hung Weng and
JingChao Wang, XiaoHong Wang, Arinc Ozturk, Quan Li, Viksit Kumar, Anthony E. Samir

medical image segmentation, one-shot learning, atlas-based segmentation, correspondence, forward-backward consistency, cycle consistency, registration, anatomical priors

1:01

15/11/2020

Applied computing, Life and medical sciences, Health care information systems, Human-centered computing, Human computer interaction (HCI), Interactive systems and tools, Theory of computation, Theory and algorithms for application domains, Machine learning theory, Reinforcement learning, Sequential decision making

8:15

06/12/2020