Team Members: Joshua Khorsandi, Kamie Mueller, Precious Oyinloye, Yue Yu, Sandhya Ganesh, Raajameenakshi Alahapphan, Divya Tinnium <\/p>\n\n\n\n
Department: Biomedical Engineering<\/p>\n\n\n\n
Divya Tinnium (Master\u2019s in BME \u201826), Precious Oyinloye (Master\u2019s in BME \u201826), Raajameenakshi Alahapphan (Master\u2019s in BME), Joshua Khorsandi (Bachelors in Biophysics \u201826), Yue Yu (Bachelors in Applied Math \u201826), Kamie Mueller (Bachelors in BME \u201826), and Sandhya Ganesh (Master\u2019s in BME \u201826) are a part of a Precision Medicine Team that is discovering phenotypes of early onset heart failure in women. Over 6 million people in the United States are diagnosed with some form of heart failure. Among women, it is one of the leading causes of mortality and morbidity. Historically, younger women have been excluded from heart failure-related studies, and their project takes a patient-enriched dataset of women and identifies subgroups (phenotypes) to see if they have different trends of hospitalization after their heart failure diagnosis. <\/p>\n\n\n\n
The team used the electronic health record (EHR) data from the Johns Hopkins Hospital and standardized the data using the Observational Medical Outcomes Partnership Common Data Model (OMOP CDM), which holds tables of standardized medical vocabulary across EHRs. The team gained familiarity with a platform to access, query and analyze EHR data – one they hadn\u2019t used before – and they even ended up writing original code to access and map data. By analyzing the data and implementing machine learning frameworks, the team was able to identify 5 distinct phenotypes of heart failure in younger women and analyzed the hospitalization trends of each. <\/p>\n\n\n\n
One of the largest problems they faced in the project was electronic health records and the dataset they used. There is a lengthy process of documentation review and deciphering to use data from different databases. Also, because they are real-world data sets, the team also had to address data quality limitations, such as the validity of diagnoses and the possibility of missing labs, so the team had to make decisions and justifications on how they used the health records. <\/p>\n\n\n\n
The team is collaborating with Dr. Anum Minhas, a cardiologist and clinician at the Johns Hopkins School of Medicine with an area in expertise in cardio-obstetrics, and a couple of people from her clinical team, like Dr. Robert B. Barrett, to gain clinical insights and interpretations from the models they create. Additionally, through weekly meetings with Dr. Greenstein, Dr. Taylor, and Dr. Ardekani, the professors teaching Precision Care Medicine, the team has gained advice and a perspective on the engineering side of their project. With an extremely diverse set of backgrounds (data science to applied math to biophysics to engineers to clinicians), the team is glad to have many different perspectives on the same problem and to collaborate to get results that can be explained to a broader audience. <\/p>\n\n\n\n
The project is being continued on into next year by Joshua (Bachelors in Biophysics \u201826), and the team hopes to shed light on and solve problems unique to women while creating more precise diagnoses about heart failure. They hope to be able to predict precisely individualized heart failure outcomes and risks of hospitalization, which may prove to be a lot more beneficial, accurate, and data-driven compared to the current model of heart failure diagnosis. <\/p>\n\n\n\n