Philip J. Freda

Biography

I am currently a project scientist at Cedars-Sinai Medical Center in Lost Angeles, CA. My research aims to identify risk factors of complex human diseases and poor clinical outcomes, with a primary focus on opioid use disorder (OUD) by identifying multilevel (individual, social, genetic, epigenetic, transcriptomic, clinical, and environmental) predictors using machine learning (ML) and artificial intelligence (AI). OUD and problematic opioid use have been shown to be associated with a host of other poor clinical outcomes. Thus, the identification of causative factors of OUD and problematic opioid use can assist in the development of risk assessment protocols and tools that can inform patients and clinicians of potential consequences of undergoing surgery or other medical procedures. To this end, I make extensive use of clinical notes, electronic health record (EHR) data, demographic information, societal data (including biometric and social media), and omics data to accurately predict poor clinical outcomes (including OUD) and identify predictors not usually captured using standard inference-based statistical approaches by employing various AI techniques including automated machine learning and natural language processing. I aim to create comprehensive patient profiles by combining diverse sources of available data to identify new dimensions of risk so accurate and broadly applicable risk assessment tools can be developed and applied to many diagnoses and medical procedures. My primary research goal is to reduce incidences of poor clinical outcomes to limit their adverse impacts on patients, families, and society.

Alongside my interests in predictive/precision medicine, I am currently engaged in the active development of automated machine learning techniques aimed at establishing associations between genotypes and phenotypes, as well as identifying sources of non-additive genetic variation, such as epistasis.