ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
DateSpeakerAffiliationTitleDescriptionBio
2
9/6/16Carol FriedmanDepartment of Biomedical Informatics, Columbia UniversityOrientation for New Trainees
3
9/13/16Jonathan ChenDepartment of Medicine, Stanford UniversityWisdom of the Crowd or Tyranny of the Mob? Data-Mining Electronic Health Records for Clinical Decision SupportMedical decision making is fraught with uncertainty, with the majority of clinical decisions lacking adequate evidence to support or refute their efficacy. Medical practice is thus routinely driven by individual opinions and anecdotal experience. Knowledge dissemination is itself an open challenge with inconsistent implementation yielding wide practice variability, compromising both healthcare quality and efficiency.



The current state-of-the art in clinical decision support, such as order sets and alerts, reinforce best-practices but are limited in scalability by the manual production required of human experts and poor workflow integration. “Grand challenges” in clinical decision support thus include mining clinical data sources to automatically generate decision support content with seamless interface designs.



The meaningful use era of electronic health records creates the transformative opportunity of our time. Statistical approaches allow us to learn underlying patterns that reflect real-world standards of care vs. outliers. A particular focus for my current NIH Big Data 2 Knowledge K01 Career Development Award is adapting routine clinical data to empower individual clinicians with the collective experience of the many.



I will review my efforts developing a collaborative filtering machine-learning approach, analogous to Netflix or Amazon.com’s “Customer’s who bought A also bought B” algorithm, applied to clinical order entry. This automatically generated decision support content can reproduce, and even optimize, manual constructs like order sets while remaining largely concordant with guidelines and avoiding inappropriate recommendations. This has even more important implications for prevalent cases where well-defined guidelines and order sets do not exist. The same methodology is predictive of clinical outcomes comparable to state-of-the-art prognosis models (e.g., APACHE). With such infrastructure in place, embedded randomization of interventions and decision support could further allow us to explicitly build knowledge for the future, even as we enhance care today, in a closed-loop learning health system.
4
9/20/16Soumitra Sengupta, Kang ChenDepartment of Biomedical Informatics, Columbia UniversityDBMI New Students Information Security SummaryWe present background information about security requirements of Health Care information. Examples of information security threats, incidents and consequences are discussed in light of regulatory requirements. Only the first year students/postdocs are required to attend the presentation.
5
9/27/16Student Seminar: Tal and DrashkoDepartment of Biomedical Informatics, Columbia UniversityTBD
6
10/4/16Chad RossHanover ResearchHanover-Columbia OHDSI Seminar OverviewHanover Research, a grants consulting firm, will present on how their core services can support faculty and graduate students in broadening their research and education through grants and fellowship funding. Staff from Hanover’s Grant Development Center will provide a brief overview of their work with Columbia over the past year, as well as case studies on how to best leverage the firm’s services. Hanover’s Grant Development Center works with higher education institutions across the United States to locate private and federal funding opportunities, tailor projects to grantmakers’ priority areas, and increase the overall efficiency of clients’ grantseeking pursuits.
7
10/11/16Rong ChenClinical Genome Informatics, Mount Sinai School of MedicineUsing big data to interpret human genomes for diagnostics, therapeutics, and precision medicine
8
10/18/16Patrick RyanOHDSI, Department of Biomedical Informatics, Columbia UniversityMaking Evidence Great Again: Lessons from the Observational Health Data Sciences and Informatics (OHDSI) collaborativeThousands of researchers around the world are attempting to use observational data, such as electronic health records and administrative claims, to generate evidence about disease natural history, treatment utilization, and the effects of medical interventions. But are we actually doing more harm than good? The Observational Health Data Sciences and Informatics (OHDSI, http://ohdsi.org) collaborative was established as open science community to learn, develop, and apply scientific best practices to the appropriate use of observational data. OHDSI's mission is to "improve health, by empowering a community to collaboratively generate the evidence that promotes better health decisions and better care." In this talk, we will explore how the current practice in observational research is performing, and demonstrate how advances in empirical evaluations and large-scale analyses can improve the transparency, reproducibility, and reliability of population-level effect estimation. We will also highlight how joining and collaborating in the world's largest international network offers you opportunities for academic development (and high-impact publications) in methodological research, open-source development, and clinical evidence generation.
9
10/25/16Student Seminar: William Ogallo and Jing HeDepartment of Biomedical Informatics, Columbia UniversityTBD
10
11/1/16Student Town HallDepartment of Biomedical Informatics, Columbia UniversityStudent Town Hall
11
11/8/16Election Day
12
11/15/16AMIA
13
11/22/16Student Seminar: Silis Jiang and Alex HsiehDepartment of Biomedical Informatics, Columbia University
14
11/29/16Rajesh RanganathDepartment of Computer Science, Princeton UniversityDeep Generative Models and Survival AnalysisTo model data we desire to express assumptions, infer hidden structure, make predictions, and simulate new data. Probabilistic generative models provide a common toolkit to meet these challenges. In this talk, I will discuss deep exponential families (DEFs). DEFs are a class of model building blocks constructed from exponential families inspired by the hidden structure used in deep neural networks. I show how they can be used to model a large collection of documents. In the second half of the talk, I present a new model for survival analysis in the electronic health record (EHR) called deep survival analysis. Deep survival analysis departs from previous approaches in two main ways: (1) all observations, including covariates, are modeled jointly conditioned on a deep generative process; and (2) the observations are aligned by their failure time, rather than by an arbitrary time zero as in traditional survival analysis. We validate our approach on coronary heart disease (CHD) using 313,000 patients corresponding to 5.5 million months of observations.
15
12/6/16Alice KimCenter for Career Education, Columbia UniversityTBD
16
12/13/16Winter Break
17
12/20/16Winter Break
18
12/27/16Winter Break
19
1/3/17Winter Break
20
1/10/17Winter Break
21
1/17/17Kai WangDBMI/Institute for Genomic MedicineInterpretation of personal genomes in inherited diseases and cancerThe research in our lab focuses on developing informatics approaches to understand the association between genetic variants and diseases, facilitate genetic diagnosis of diseases, optimize the selection of treatment strategies, and improve patient prognosis. In this talk, I will first describe a suite of computational tools including Annovar, InterVar and Phenolyzer, which analyze both genome sequence data and phenotype terms to identify disease genes/variants in individual patients. I will next describe iCAGES, a statistical approach based on Annovar and Phenolyzer, for prioritizing cancer driver genes/variants for a patient using his/her genome sequencing data and gene expression data. Recent extension of iCAGES leverages a deep learning group lasso cox model that integrates genetic, molecular, proteomic, demographical and clinical profiles of cancer patients to optimize drug selection and predict patient survival.
22
1/24/17Student Seminar:Department of Biomedical Informatics, Columbia University
23
1/31/17David ShawD.E. Shaw ResearchCan Molecular Dynamics Simulations Cure What Ails Ya? [Special Seminar @ Vagelos Education Center, Room 201]Molecular dynamics simulations have in recent years been playing an increasingly important role in advancing our understanding of biological processes at an atomic level of detail. Such simulations are often capable of providing new insights into the interactions of biological macromolecules with each other and with endogenous and pharmaceutical ligands, suggesting that they may ultimately make important contributions to the process of drug discovery. This talk will examine several ways in which MD simulations might seem to be of potential use in designing new drugs. Examples will be given of both current capabilities and current limitations, with a focus on their potential implications for the use of MD as a tool for drug discovery.
24
2/7/17Michelle C. BensonOffice of Research Compliance & TrainingSafeguarding Your Research: Resources for Recognizing Research Misconduct and Avoiding Plagiarism
25
2/14/17Min SongYonsei UniversityAnalyzing Trend, Dynamics, and Diffusion of Knowledge in Cancer ResearchAbstract: Cancer is one of the leading causes of death worldwide, and its incidence is expected to exacerbate. To respond to the critical need from the society, there have been rigorous attempts for the cancer research community to develop treatment for cancer. Accordingly, we have observed a surge in the sheer volume of research products and outcomes in relation to neoplasms. In this talk, I will propose a new approach to understand the trends, dynamics, and diffusion of knowledge associated with cancer research. To this end, we collected over two million records related to cancer research from PubMed, the most popular search engine in the medical domain. Coupled with text mining techniques including named entity recognition and neural word embedding, the proposed approach will provide a new insight into understanding how the research of common cancers such as prostate, liver, and brain cancer have diffused and changed over time. At the end of the talk, the future applications and possible directions of the proposed approach will be discussed.
26
2/21/17Student Seminar:Department of Biomedical Informatics, Columbia University
27
2/28/17Lucila Ohno-MachadoDepartment of Biomedical Informatics, UCSDEcosystems for a Biomedical Research Commons: Finding and Using Data while Protecting PrivacyThe widespread adoption of electronic health record systems and data harmonization efforts across different health systems have made it possible to conduct analyses of large clinical data sets. In this presentation, I will present DataMed, a search engine for data sets, and I will describe our experience in developing a distributed clinical data research network for comparative effectiveness. This experience is being employed in the California Precision Medicine Consortium, which contributes to the All of Us program (formerly known as the Precision Medicine Initiative). I will discuss how the use of distributed analytics helped solve issues related to institutional policies while preserving patient privacy. I describe how vertically and horizontally partitioned data can be analyzed to create calibrated predictive models that have clinical significance.
28
3/7/17Nicholas GenesEmergency Medicine/Genomics, Mt. SinaiQuality report and Practice VariationMeasuring and reporting physician and hospital performance and quality has become a significant factor in reimbursement.

This presentation will explore:

1) the history of this practice, including notable successes and failures, 2) how quality is currently reported, and 3) how registries can leverage the capabilities of EHR to improve healthcare feedback and reduce practice variation.
29
3/14/17Spring Break
30
3/21/17Student Seminar: Albert and JonDepartment of Biomedical Informatics, Columbia University
31
3/28/17David GoldsteinInstitute for Genomic Medicine
32
4/4/17Chintan Patel & Sharib KhanApplied Informatics/TrialX
33
4/11/17Student Seminar: Andy and MollieDepartment of Biomedical Informatics, Columbia University
34
4/18/17
35
4/25/17Joel DudleyMount SinaiMoving from Precision Medicine to Next Generation Healthcare
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100