With the explosion of health information technology, there is potential to rapidly advance phenotyping in pain medicine to inform mechanistic gaps and identify novel treatments. The interface between traditional and novel (AI and machine-learning) techniques will be presented for pain research and precision pain care.
The Interface of Traditional and Machine-Learning Approaches to Assess Persistent Post-Surgical Pain
Spine Pain Mechanistic Phenotyping
Electronic Health Data Challenges and Applications in Pain Medicine Research and Treatment
- Define current approaches used to comprehensively phenotype patients with chronic spine pain, including deep characterization through patient-reported outcome measures, quantitative sensory testing, and neuroimaging techniques. Identify gaps in knowledge in spine pain phenotyping and provide a conceptual framework for how clinical phenotyping can interface with machine learning and big data.
- Define fundamental challenges to pain data organization and collection in electronic health data in the United States. Identify advanced analytic methods for working with electronic health data and current solutions to large-scale data modeling. Identify potential roles for informatics in pain research.
- Identify current limitations to perioperative risk stratification for persistent postsurgical pain and opioid use. Define how machine learning and data mining of the electronic health record can advance knowledge discovery to augment risk stratification, inform prevention efforts, and complement clinical treatment pathways.