EHRs records contain valuable information in unstructured format, and standardizing them has the potential to improve healthcare systems and led to save lives. 🏥
The methodology presented, called the Neuro-Symbolic System for Cancer (NSSC), is an end-to-end tool that enhances the named entity recognition and linking tasks. It is presented as an optimization problem to normalize free text as a low-cost solution. Additionally, it is designed as a disease-specific contextual adaptability framework, which allows it to be adapted to other diseases beyond cancer and to others vocabularies. The only ad-hoc step required is the generation of clinical entities, as an annotated corpus is needed to train this model. 📚
I hope you find this work useful and that it contributes to the emerging field of neuro-symbolic systems. By using this hybrid methodology, we can achieve interpretable solutions in a cost-effective manner. ⚖️
NSSC: A Neuro-Symbolic AI System for Enhancing the Accuracy of Named Entity Recognition and Linking from Oncologic Clinical Notes
Journal Paper published in Medical & Biological Engineering & Computing.
Like
Be the first to like this
Follow the Topic
Public Health
Life Sciences > Health Sciences > Public Health
Computer and Information Systems Applications
Mathematics and Computing > Computer Science > Computer and Information Systems Applications
Cancers
Life Sciences > Health Sciences > Clinical Medicine > Diseases > Cancers
-
Medical & Biological Engineering & Computing
This journal covers the entire spectrum of biomedical and clinical engineering. The journal aims to present exciting and vital experimental and theoretical developments in biomedical science and technology and to report on advances in computer-based methodologies in these multidisciplinary subjects.
Please sign in or register for FREE
If you are a registered user on Research Communities by Springer Nature, please sign in