Aaron Masino PhD

Aaron Masino PhD
Assistant Professor, Anesthesia and Critical Care Medicine
Perelman School of Medicine, University of Pennsylvania
Supervisor, Research Data Science
Children's Hospital of Philadelphia

Tuesday December 3, 2019
12:00 PM - 1:00 PM
Stokes Auditorium (Main Hospital)
Children's Hospital of Philadelphia

Learning Objectives

  • Understand risk associated with delayed time to antibiotic treatment for infants with sepsis.
  • Describe challenges associated with automated diagnostic tests in a continuous monitoring environment (e.g. ICU).
  • Differentiate global and prediction level machine learning model explanations.

Aaron Masino PhD
Aaron joined the Department of Biomedical and Health Informatics (DBHi) at The Children's Hospital of Philadelphia in 2011. His research interests include: precision medicine applications that incorporate patient specific, data driven mathematical models to provide clinical decision support; real-time population health applications that continuously integrate and analyze data to identify current population health trends; and fundamental data science methods in unsupervised learning over unstructured data, extensions of deep-learning methods, and development of uncertainty measures for non-linear models. His most recent research includes machine learning model development for sepsis prediction, natural language processing applications for information extraction from clinical text, and deep learning methods to detect adverse drug events described in social media. Previously, Aaron served as a senior scientist at MZA Associates Corporation where he developed advanced optics control algorithms. He received his PhD in applied mathematics from the University of Central Florida. He also holds a master's in aerospace engineering from the University of Colorado and a bachelor's in mathematics from Rutgers University.

Continuing Medical Education

ACCME Accreditation Statement
Children's Hospital of Philadelphia is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians.

AMA Credit Designation Statement
Children's Hospital of Philadelphia designates this live activity for a maximum 1.0 AMA PRA Category 1 Credit(s)™. Physicians should claim only the credit commensurate with their participation in the activity.

Disclosure Statement
Dr. Masino signed and completed Disclosure of Relevant Financial Relationships Forms and affirmed that neither he nor his spouse/partner/family members /have or had any relevant financial relationships or financial affiliations with commercial interest(s) within the past 12 months related directly or indirectly to this educational activity that may pose a conflict of interest within the context of this lecture.

Maintenance of Certification (MOC)
The American Board of Preventive Medicine (ABPM) has approved this activity for a maximum of one (1) LLSA credit towards ABPM MOC Part II requirements. Board certified physicians in Clinical Informatics who wish to receive MOC credit for this session must review the MOC questions below prior to the presentation. Answers will be posted on dbhi.chop.edu following the presentation.

1. (True or False) Increased time between clinician suspicion of sepsis and antibiotic administration in infants is a significant risk factor for increased mortality.

a. True
b. False

2. In an ICU setting, automated models (such as those based on machine learning methods) may process updates to patient data to make multiple longitudinal diagnostic predictions. Which of the following is a potential limitation as a clinical decision support tool that results from this?

a. False alarms
b. Vital sign data
c. None of the above

3. Machine learning model interpretation methods can provide either global or local explanations. Global explanations describe properties of the model that hold across the entire population. Local explanations provide information about an individual model output. Which is NOT a form of local explanation?

a. Prediction feature importance
b. Transparent model design
c. Prediction certainty
d. Similar instances (e.g. similar patients)

Nurse Attendance
Nurses who attend the live educational activity will need to sign in on the Nurse Sign-in Form for CME Activities and will be eligible to receive a certificate of completion, which may be used for re-licensure and recertification. For further information please contact Sharon Miller, MSN, RN, Nursing Professional Development Specialist.This email address is being protected from spambots. You need JavaScript enabled to view it..