This presentation will explore how Natural Language Processing can be used to extract information from reports in radiology, from breast density extraction in structured text to brain tumor characteristics in unstructured descriptions. The use of regex, cTAKES, and LSTM-based models will be highlighted.

Here are some optional pre-reading materials to brush up on topics related to this presentation:

• regex - Learn Regular Expressions
• cTAKES -  Mayo clinical Text Analysis and Knowledge Extraction System (cTAKES): architecture, component evaluation, and applications
• LSTM - A Gentle Introduction to Long Short Term Memory
• Embeddings - What are Word Embeddings for Text?

About the Speaker: 

Dr. Pablo Damasceno (pablo.damasceno@ucsf.edu) has over 15 years of experience in high-performance computing and has worked on a diverse range of problems, from biochemical engineering to mathematics to neuroscience. His current focus is on allying software (Deep Learning, Bayesian Inference, Network Theory) and hardware (high-performance computing, GPU acceleration, cloud computing) to enable scientific excellence in biomedical imaging. As a Senior Machine Learning specialist at UCSF’s Center for Intelligent Imaging, he helps develop, train, and deploy deep learning pipelines to complement clinicians’ examination workflows, improving speed, reproducibility, and reliability in diagnoses.

What is NLP@UCSF?

Are you curious about how natural language processing (NLP) can be applied in healthcare, clinical, and biomedical research? NLP@UCSF brings together learners, users, and doers to share and discuss NLP applications in medicine.

We meet each month via Zoom. Our community welcomes people of all backgrounds and skill levels, and all are welcome to attend and join the conversation! Check out our website to see upcoming meetups or access recordings from previous sessions. You can also join our Slack channel to stay connected with the community.