2020 Biomedical Informatics Seminars

December 11, 2020

Kai Zheng, PhD, FACMI, Professor, Department of Informatics; Professor, Department of Emergency Medicine; Director, Center for Biomedical Informatics, Institute for Clinical and Translational Science, University of California, Irvine.  "The Potential (or the Lack Thereof) of Voice-based Ambient Clinical Documentation Technologies."

Abstract:  Clinician burnout has become a prevalent phenomenon in US hospitals and clinics. A key contributing factor is the increased documentation burden attributable directly or indirectly to the widespread adoption of electronic health records. Many solutions have been proposed, from improving the usability of clinical information systems to deregulation efforts for reducing unnecessary documentation requirements. However, their effects on alleviating clinician burnout have only been marginal. An emerging family of solutions based on ambient automatic speech recognition (ASR) and national language understanding have received substantial attention in the past few years. These solutions, often referred to as "digital scribes," provide the capability for transcribing patient–clinician conversations in the exam room and generating clinical documentation automatically. Through a series of studies sponsored by SAP SE, we assessed the potential of such solutions. We first examined the work performed by human medical scribes, followed by an investigation into the performance of best-of-breed ASR technologies when applied to conversational dialogs taking place in the exam room. Based on the results, we concluded that: (1) medical scribes' work is incredibly versatile, far beyond simply transcribing patient–clinician conversations, and thus their role cannot be completely replaced by digital scribing agents; (2) ASR-produced transcripts of conversational clinical dialogs are of reasonable quality and are ready for trial use in real-world settings; and (3) additional research is needed on how to transform the resultant transcripts into structured data entries and free-text narratives that are accurate and are clinically meaningful.

Bio:  Kai Zheng, PhD, FACMI, is Professor of Informatics and Emergency Medicine at the University of California, Irvine. He is also Director of the Center for Biomedical Informatics at the UC Irvine Institute for Clinical and Translational Science. Prior to joining UC Irvine, he was Associate Professor of Health Management and Policy in the School of Public Health and Associate Professor of Information in the School of Information at the University of Michigan. He was Director of University of Michigan's Health Informatics Program preparing students for careers that will harness the power of information to enhance health and transform individual health and healthcare. Zheng's research draws upon techniques from the fields of information systems and human–computer interaction to study the use of information, communication, and decision technologies in patient care delivery and management. His recent work has focused on topics such as interaction design, workflow and sociotechnical integration, and diffusion and evaluation of health IT. His recent work also includes use of computational methods to augment the value of unstructured free-text data such as clinician notes and patient-generated content online. Zheng received his PhD degree in Information Systems from Carnegie Mellon University. He was the recipient of the 2011 American Medical Informatics Association (AMIA) New Investigator Award that recognizes early informatics contributions and significant scholarly achievements. He is an elected Fellow of the American College of Medical Informatics, and currently serves as Chair of AMIA's Clinical Information Systems Working Group.

December 4, 2020

Benjamin Smarr, PhD, Assistant Professor, Bioengineering, and the Halicioğlu Data Science Institute, UC San Diego.  "Wearables in public health monitoring - human and viral diversity in data."

Abstract:  Benjamin Smarr is the technical lead on TemPredict, a collaboration between UCSF, UCSD, MIT Lincoln Lab using wearable device data to build algorithms for early detection of COVID-19 illness onset. This effort has gathered >60K participants donating wearables data and surveys on conditions, symptoms, and test results. Despite initial skepticism that wearables could be useful in illness tracking, it quickly became clear that the data provides great promise. At the same time, these data are rich enough to capture the remarkable diversity of healthy baselines and illness manifestations. Such complexity helps us identify common trends, while highlighting the individual differences that make broad and uniform medical advances so elusive.

Bio:  Dr. Smarr got his PhD at the University of Washington studying the neural integration of daily and ovarian cycles. He was then an NIH fellow at Berkeley, where he demonstrated new approaches for extracting biologically important information from time series features in continuous physiological measurements. Dr. Smarr joined the Bioengineering faculty at UCSD in March 2020, with a joint appointment to the Halicioglu Data Science Institute. When he joined, COVID-19 was taking off in the US, so that his work has pivoted to using his time series approaches in service of the fight against the pandemic.

November 20, 2020

Vimal Mishra, MD, MMCi, Medical Director, Office of Telemedicine, Associate Professor of Medicine and Health Administration, Virginia Commonwealth University.  "AI in Healthcare: Fact, Fiction and Opportunity Examples from the Trenches."

Abstract:  Artificial Intelligence (AI) is already shaping the health care landscape, and rapid advancements in AI-enabled technologies inspire further considerations vital to practicing evidence-based medicine and delivering high-quality clinical care. Join us for discussion on AI—from what it is and how it is practiced to how the health care industry can harness AI in ways that safely and effectively improve patient care.

Bio: Dr. Vimal Mishra is an associate professor of medicine and health administration at Virginia Commonwealth University Health System, specializing in internal medicine and health informatics. As a physician providing hospital-based medical care, his work lies at the intersection of delivering high-quality clinical care and leading innovative clinical and digital transformation. He serves as the medical director of telehealth and is responsible for providing the vision, strategy, thought, and clinical leadership to enterprise-wide virtual and digital health efforts. As a physician informaticist, he provides subject matter expertise in integrating medicine with computer/information science. He guides designing, planning, implementation, adoption, optimization of the electronic medical record and clinical decision tools to enable high quality, safe, effective, and efficient patient care.

Dr. Mishra serves as Director of Digital Health for the American Medical Association (AMA). In this role, he provides clinical and strategic leadership for AMA national programs focused on development, implementation, EHR optimization, and adoption of evidence-based digital health solutions. He serves as a clinical lead for programs like Physician Innovation Network (PIN), health tech accelerator relationships and provides research leadership to advance the AMA digital health research agenda.

His research interest focuses on complex health care processes and delivery and the application of multidisciplinary approaches that integrate systems science, artificial intelligence, informatics, operations research, and management methodologies to improve health care safety, quality, efficiency, and effectiveness. Dr Mishra has published significant numbers of peer-reviewed scholarly articles, has received several federal and state-funded grants, and holds invention disclosures. Dr. Mishra has been nominated as Richmond's top doctor in hospital medicine since 2017 and holds an award of excellence in service, clinical leadership, and research.

Dr. Mishra completed his internship and residency in Internal Medicine at Virginia Commonwealth University. He is board-certified in internal medicine and clinical informatics and holds a graduate degree from the Duke Fuqua School of Business.

November 13, 2020

Antoine Bosselut, PhD, Postdoctoral Researcher, Standford University.  "Neuro-symbolic Representations for Commonsense Knowledge and Reasoning."

Abstract:  Situations described using natural language are richer than what humans explicitly communicate. For example, the sentence "She pumped her fist" connotes many potential auspicious causes. For machines to understand natural language, they must be able to reason about the commonsense inferences that underlie explicitly stated information. In this talk, Dr. Bosselut will present work on combining traditional symbolic knowledge and reasoning techniques with modern neural representations to endow machines with these capacities.

First, Dr. Bosselut will describe COMET, an approach for learning commonsense knowledge about unlimited situations and concepts using transfer learning from language to knowledge. Second, he will demonstrate how these neural knowledge representations can dynamically construct symbolic graphs of contextual commonsense knowledge, and how these graphs can be used for interpretable, generalized reasoning. Finally, Dr. Bosselut will discuss current and future research directions on conceptualizing NLP as knowledge simulation, and the impact of this framing on challenging open-ended tasks.

Bio: Antoine Bosselut is a Postdoctoral Researcher at Stanford University and will join the Swiss Federal Institute of Technology in Lausanne (EPFL) as an assistant professor beginning in Fall 2021. He completed his PhD at the University of Washington and was a student researcher at Microsoft Research (2017-2018) and AI2 (2018-2020) during his studies. His research interests are in building systems that jointly represent knowledge and language to solve problems in natural language processing. He is also a Young Investigator at the Allen Institute for AI.

November 6, 2020

Nan Zhang, PhD, Professor of Information and Technology and Analytics, Kogod School of Business, American University.  "Implications of Data Anonymization on Disparity Detection."

Abstract: Research and practical development of data anonymization techniques has proliferated in recent years. Yet limited attention has been paid to examine the potentially disparate impact of privacy protection on underprivileged sub-populations. The objective of this talk is to examine the extent to which data anonymization could mask the gross statistical disparities between sub-populations in the data. We first describe two common mechanisms of data anonymization and two prevalent types of statistical evidence for disparity. Then, we develop conceptual foundation and mathematical formalism demonstrating that the two data anonymization mechanisms have distinctive impacts on the identifiability of disparity, which also varies based on its statistical operationalization. After validating our findings with empirical evidence, we discuss the practical implications, highlighting the need for data scientists and policy makers to balance between the protection of privacy and the recognition/rectification of disparate impact.

Bio: Dr. Nan Zhang is a Professor of Information Technology and Analytics in the Kogod School of Business at the American University. His research focuses on data analytics and information security/privacy. His work has received many awards, including the Communications of the ACM Research Highlight in 2020, the ACM SIGMOD Research Highlight Award in 2019, the National Science Foundation (NSF)'s CAREER award in 2008, and several best paper awards from various leading conferences. Before joining Kogod in 2018, he was a Professor of Information Sciences and Technology at the Pennsylvania State University, and a Professor of Computer Science at the George Washington University. Between 2016 to 2017, Dr. Zhang served as a Program Director at the NSF, with responsibilities in the Information Integration and Informatics (III), Secure and Trustworthy Cyberspace (SaTC), and BIGDATA programs.

October 30, 2020

Ming Tai-Seale, MD, Professor, Family Medicine and Public Heath, UC San Diego Health Sciences.  Facilitating Patient and Health Care Team Engagement during Covid-19 Pandemic with Multiple Functionalities of Electronic Health Records in Three Healthcare Delivery."

Abstract:  Meaningful use requirements and functionalities have laid the technological foundation for telehealth for many years. It was not until COVID-19 that the volume of telehealth video visits started to skyrocket. Delivering health services via telehealth presents an opportunity to support patients and physicians in shared decision making during this uncertain time when in-clinic encounters may be less preferred. Dr. Tai-Seale will discuss an ongoing multi-center cluster randomized trial funded by the Patient Centered Outcomes Research Institute (PCORI) that leverages the electronic health record system to design and implement user-centered EHR tools to facilitate patient-care team communication at three large health care systems before and during COVID-19 crisis.  

Bio:  Dr. Ming Tai-Seale is a Professor in UCSD's School of Medicine. Her research has been funded by the NIH, AHRQ, PCORI, and they investigate the practice of medicine, patient-physician communications, and mental health care. Her publications on patient-physician communication earned the Article-of-the-Year award from AcademyHealth. A co-authored publication on the practice of medicine in the age of electronic health records was the second most read paper in Health Affairs in 2017. She also serves as the Director for Outcomes Analysis and Scholarship at the UC San Diego Health Information Services, Director of Research in the Department of Family Medicine and Public Health, and Director of Research and Learning in the Population Health Services Organization at UC San Diego Health. 

October 23, 2020

Christopher Longhurst, MD, CIO and Associate Chief Medical Officer, Professor, UC San Diego Health.  "Clinical Informatics and COVID-19."

Abstract: The COVID-19 pandemic has impacted every aspect of our lives including healthcare delivery. The UC San Diego Health departments of Information Services and Quality & Patient Safety have supported these major transformations with new tools and processes from telehealth and analytics to exposure notification systems and daily tiered huddles. In this talk, Dr. Longhurst will explore some of these recent changes over the last 6 months, as well as resulting scholarship locally and nationally.

Bio:  Christopher Longhurst MD, MS, serves as the chief information officer and associate chief medical officer for quality and safety at UC San Diego Health. This dual, complementary role provides leadership in key functional areas to continuously improve efforts around operations, reputation and quality of care provided at UC San Diego Health.

He is the author and co-author of many publications on using technology and data to improve patient care and outcomes as well as quality improvement. Dr. Longhurst is an elected fellow of the prestigious American College of Medical Informatics.

October 16, 2020

Jejo Koola, MD, MS, Assistant Professor of Medicine, UCSD Health, Department of Biomedical Informatics.  "Modeling Learning Effects for Medical Device Safety Surveillance Using Parametric and Non-Parametric Methods."

Abstract: Hands-on experience with implantable medical devices significantly improves clinical outcomes through learning effects. The FDA has stressed the importance of increased post-market surveillance of drugs and medical devices. However, contemporary post-market surveillance strategies for medical devices currently do not separate risk attributable to the device itself versus risk attributable to the operator. Robust medical device safety surveillance that can provide information regarding the etiology of the safety signal requires accurate adjustments for the complex relationships among device characteristics, learning by physicians and institutions, and patient clinical heterogeneity.  

Bio: Dr. Koola is a practicing physician with seven years practicing experience in multiple medical centers. During his fellowship in Biomedical Informatics at the Department of Veterans Affairs in conjunction with a master's degree from Vanderbilt University he gained formal skills in database management, data mining, information retrieval, software engineering, and natural language processing. He worked with patients who have advanced liver disease (cirrhosis) and constructed risk prediction models using machine learning techniques for mortality, readmission, and phenotyping (case identification). Additionally, he constructed machine learning models to phenotype Hepatorenal Syndrome (a complication of cirrhosis) using Natural Language Processing. He has been working on ways to use observational cohort data to predict decompensation in patients with significant medical comorbidities.

October 9, 2020

Lucila Ohno-Machado, MD, PhD, MBA, Professor and Chair, UCSD Health, Department of Biomedical Informaics.  "Reliable Response Data Discovery (R2D2) COVID19 Clinical Consultation using Patient Observations (C3PO)"

Abstract: There is an urgent need to answer questions related to COVID-19's clinical course and associations with underlying conditions and health outcomes. Multi-center data are necessary to generate reliable answers, but centralizing data in a single repository is not always possible. Using a privacy-protecting strategy, we launched a public Questions & Answers web portal (https://covid19questions.org) with analyses of comorbidities, medications and laboratory tests using data from 202 hospitals (59,074 COVID-19 patients) in the USA and Germany. We find, for example, that 8.6% of hospitalizations in which the patient was not admitted to the ICU resulted in the patient returning to the hospital within seven days from discharge and that, when adjusted for age, mortality for hospitalized patients was not significantly different by gender or ethnicity.

Bio: Lucila Ohno-Machado, MD, MBA, PhD received her medical degree from the University of São Paulo and her doctoral degree in medical information sciences and computer science from Stanford. She is Associate Dean for Informatics and Technology, and the founding chair of the UCSD Health Department of Biomedical Informatics at UCSD, where she leads a group of faculty with diverse backgrounds in medicine, nursing, informatics, and computer science. Also, she is the PI for the California Precision Medicine Consortium for the NIH All of Us Research Program. Prior to her current position, she was faculty at Brigham and Women’s Hospital, Harvard Medical School and affiliated with the MIT Division of Health Sciences and Technology. Dr. Ohno-Machado is an elected member of the American College of Medical Informatics, the American Institute for Medical and Biological Engineering, the American Society for Clinical Investigation and the National Academy of Medicine. She served as editor-in-chief for the Journal of the American Medical Informatics Association from 2011 to 2018. She directs the patient-centered Scalable National Network for Effectiveness Research, a large clinical data research network covering more than 30 million patients and 12 healthcare systems, and was one of the founders of UC-Research eXchange, a clinical data research network that connected the data warehouses of the five University of California medical centers. She was the director of the NIH-funded National Center for Biomedical Computing iDASH (integrating Data for Analysis, ‘anonymization,’ and Sharing) based at UCSD with collaborators in multiple institutions, as well as other NIH-funded consortia and research projects. Her research focuses on privacy-preserving distributed analytics for healthcare and biomedical sciences. She has received numerous awards for innovations in biomedical informatics.

June 5, 2020

Ming Tai-Seale, PhD, MPH, Professor of Family Medicine and Public Health, Director, Health IS Outcomes Analysis, University of California San Diego, "Enhancing Shared Decision Making in Primary Care Practice Using EHR Tools."

Bio: Ming Tai-Seale, PhD, MPH, Professor of Family Medicine and Public Health, UCSD School of Medicine, and Director of Outcomes Analysis at UCSD Health IS Department, is a health economist and health services researcher. Tai-Seale examines the impact of financial and organizational incentives on health services delivery and patient outcomes. She collaborates with physicians, patients, clinic leaders, and researchers from multiple disciplinary backgrounds to study patient-physician communication, primary care transformation, and how physicians allocate time and effort in office visits by taking a detailed look at patient-physician communication captured by video- or audio-recordings and the access log of EpicCare. A paper resulting from analyses of EpicCare Logs was the second most read paper in Health Affairs in 2017.

May 29, 2020

Larissa Neumann, MD, Visiting Scholar, UCSD Health Department of Biomedical Informatics, Physician, Department of Anesthesiology, Research Assistant, Institute for Medical Information Processing, Biometry, and Epidemiology (IBE), University Hospital of Ludwig Maximilian Universi, "The Evolution Of Tech Medicine In Germany: A Brief View Of My Journey From A Medical Doctor To A Clinician-Informaticist."

Abstract: The transition from a medical doctor using semi-analog devices to a clinician informaticist is uniquely exciting, with challenges and lessons. In this talk I will discuss what technologies were gradually introduced to German academic and professional life, how they were implemented, and how they impacted patient care my personal experience. From handwritten anesthesia care documentation to the use of electronic health record data and object-relational mapping, I will review the transformation of data acquisition and discuss the need for distributed analyses. I will also provide an overview of ongoing current SARS-CoV-2 projects at my home institution and in collaboration with USCD, including the implementation of an internal tracking system and a COVID-19 Dashboard.   

Bio: Dr. Neumann is currently a visiting scholar at the UC San Diego Health Dept. of Biomedical Informatics. She practices anesthesiology at the University Hospital of the Ludwig Maximilian University (LMU) of Munich, Germany, with additional specialization in critical care.  In January 2019, she joined the Anesthesia & Critical Care Informatics and Data Analysis Group (ACID) and is working as a research assistant for the Institute for Medical Information Processing, Biometry, and Epidemiology (IBE). Her research interests are in perioperative medicine and developing predictive models for critically ill patients. In addition to exploring machine learning methods to personalize blood transfusion triggers, she is excited about being a team member of building an Observational Medical Outcomes Partnership (OMOP) data warehouse to establish a cross-institutional distributed ledger. She is also passionate about participating in Datathons and learning more about varied aspects of Biomedical Informatics.

May 15, 2020

Erin Sundermann, PhD, Assistant Professor, Department of Psychiatry, University of California San Diego, "What a difference an X makes: The challenges of diagnosing and tracking Alzheimer's disease in women"

Abstract: The scientific community is increasingly recognizing sex differences in brain function and in the clinical manifestation of brain disorders, although biological sex is rarely considered when making diagnostic or treatment decisions. Women are predominantly affected by Alzheimer's disease (AD) and show a more aggressive clinical profile; however the reasons for these sex differences are not well understood. Throughout the lifespan, women tend to outperform men in measures of verbal memory. My work suggests that the female advantage in verbal memory acts as a domain-specific cognitive reserve enabling women to delay the clinical manifestation of AD until a more advanced disease stage compared to men but show more rapid decline thereafter. This line of research is clinically important because verbal memory tests are commonly used to diagnose AD and the norms for these tests are typically not sex-adjusted. I will discuss the implications of sex differences in cognitive/brain function and AD for both clinical and research practice and in the development of personalized disease interventions.  

Bio: I am cognitive neuroscientist and Assistant Professor in the Department of Psychiatry at UCSD. My broad research scope is in the investigation of sex differences in cognitive aging and Alzheimer's disease (AD) and the discovery of sex-specific biomarkers and risk factors for AD. I strive to address critical gaps in our understanding of sex differences in AD through a comprehensive research program involving neuropsychological, genetic, biomarker and neuroimaging data. More recently, my work has exposed the limitations of standard clinical tests of verbal memory in identifying women in the early stages of the AD trajectory due to a life-long female advantage in verbal memory that may serve as a form of cognitive reserve.

May 8, 2020

Charles Jaffe, MD, PhD, FACP, FACMI, CEO, Health Level 7 International, Visiting Scholar, UCSD Health Dept. of Biomedical Informatics, "FHIR Release 4: Changes to the future of Interoperability for patient care, research, public health, and the US payment system."

Bio: Charles Jaffe is the Chief Executive Officer of Health Level 7 International (HL7). He completed his medical training at Johns Hopkins and Duke Universities and post-doctoral training at the National Institutes of Health and the Lombardi Cancer Center. He has served in various academic positions in the Departments of Medicine and Pathology, as well as in the School of Engineering. Prior to joining HL7, he was the Senior Global Strategist at Intel. In addition, he led a national research consortium, found a consultancy for research informatics, served as the VP of Medical Informatics at AstraZeneca, and the VP of Life Sciences at SAIC. Dr. Jaffe has been the contributing editor for several journals and has published on clinical management, informatics deployment, and healthcare policy.

May 1, 2020

Eric Adler, MD, Professor of Medicine, Director of Cardiac Transplant and Mechanical Circulatory Support, University of California San Diego, "Development and Validation of MARKER HF: A Machine Learning Prognostic Tool for Patients with Heart Failure."

Abstract: Heart failure is a leading cause of morbidity and mortality worldwide. A central challenge of treating heart failure patients is the ability to accurately predict prognosis. Machine Learning (ML) has great promise for radically changing medical practice but has had a limited impact to date. We used ML to design a highly accurate risk score to predict mortality in HF patients. In this talk I will discuss the process we used to develop and validate this tool and discuss ongoing projects to evaluate MARKER prospectively.   

Bio: Eric Adler, MD, is a Professor of Medicine and Medical Director of the Heart Transplant Program at the University of California, San Diego. His laboratory's focus is how genetic mutations cause cardiomyopathy. To address this important topic he uses stem cells from patients, as well as mouse models of cardiac disease and human tissue. His recent work has been the study of Danon disease, a rare yet devastating inherited cause of heart failure. 

Dr. Adler is a principal investigator in clinical trials for all stages of heart failure and is specifically involved in clinical research examining the use of gene therapy and stem cells for treating heart disease. His work has been published in leading medical journals and has been supported by the National Institutes of Health (NIH) and the California Institute for Regenerative Medicine. He earned his medical degree from Boston University School of Medicine, completed his internship and residency at the University of Washington, and Cardiology Fellowship at Mount Sinai School of Medicine. He is board certified in cardiovascular disease and advanced heart failure.

April 24, 2020

Sally Baxter, MD, MSc, Postdoctoral Scholar, UCSD Health Department of Biomedical Informatics, UCSD Viterbi Family Department of Ophthalmology, Staff Physician, Veterans Affairs (VA) San Diego Healthcare System, "Predictive Modeling in Glaucoma using EHR Data."

Abstract: Glaucoma is the world’s leading cause of irreversible blindness. Although lowering eye pressure is the mainstay of therapy, many patients exhibit disease progression despite seemingly adequate control of eye pressure. As a result, understanding how other factors, such as systemic conditions and medications, contribute to worsening of glaucoma is critically important. We have conducted predictive modeling studies using electronic health record (EHR) data and found that systemic data have predictive value in identifying patients with primary open-angle glaucoma at risk of progression to surgical intervention. In this seminar, I will describe the process of conducting these studies, the results, and their implications. By describing work conducted using local data from UCSD as well as work using data from the nationwide NIH All of Us Research Program, I will share practical pearls and lessons learned about performing predictive modeling projects using different data sources.   

Bio: Sally L. Baxter is currently a National Library of Medicine postdoctoral fellow in biomedical informatics at UCSD. Her research interests fall within two broad areas: (1) Investigating how data from the EHR and other sources (e.g. sensors, patient portals) can be leveraged to better understand the connections between systemic diseases and medications with vision and other eye-related outcomes, and (2) Understanding how ophthalmologists at all career stages (including those in training) learn and interact with the EHR, and how these interactions might be improved in order to support high-quality care, enhanced patient experience, and ease of use for clinicians. She practices comprehensive ophthalmology with a specialization in complex cataract surgery at the San Diego VA. She was an Angier B. Duke Scholar at Duke University, received her Masters of Public Health from the London School of Hygiene & Tropical Medicine as a United States Marshall Scholar, and was a 21st Century Scholar at the Perelman School of Medicine at the University of Pennsylvania prior to coming to UCSD for her internal medicine and ophthalmology training.

April 17, 2020

Trevor Cohen, MBChB, PhD, FACMI, Professor, Biomedical Informatics and Medical Education Adjunct Professor, Department of Psychiatry University of Washington, "New Uses for Neural Embeddings: Biomedical Applications of Recent Advances in Neural Representations of Natural Language."

Abstract: Distributed vector representations (embeddings) of words, concepts and sentences have become increasingly prevalent as a fundamental unit of analysis in Natural Language Processing (NLP) on account of their ability to support generalization, and their alignment with the fundamental representational paradigm of neural network models. This talk will cover recent work in which such neural representations are applied to biomedical informatics problems, some of which fall outside their original conception as representations of natural language. The projects to be presented span application domains from post-marketing drug surveillance to digital phenotyping, but have a common thread between them: the desirable properties of neural representations for NLP, and how these can be leveraged to solve biomedical problems.   

Bio: Trevor Cohen, MBChB, PhD is a Professor of Biomedical Informatics at the University of Washington in Seattle. His research focuses on the development and application of methods of distributional semantics – methods that learn to represent the meaning of terms and concepts from the ways in which they are distributed in large volumes of electronic text. The resulting distributed representations (concept or word embeddings) can be applied to a broad range of biomedical problems, such as: (1) using literature-derived models to find plausible drug/side-effect relationships; (2) finding new therapeutic applications for known medications (drug repurposing); (3) modeling the exchanges between users of health-related online social media platforms; and (4) identifying phrases within psychiatric narrative that are pertinent to particular diagnostic constructs (such as psychosis). An area of current interest involves applying literature-derived distributed representations in conjunction with observational data as a basis for machine learning.  More broadly, he is interested in clinical cognition – the thought processes through which physicians interpret clinical findings – and ways to facilitate these processes using automated methods. Before joining the University of Washington, he held faculty positions at Arizona State University, and at the University of Texas Health Science Center in Houston. Prior to this, and after training and practicing as a physician in South Africa, he completed his doctoral work at Columbia University in New York, with a research focus on the development of automated methods to enhance clinical comprehension in psychiatry.

April 10, 2020

Alysson Muotri, PhD, Professor of Pediatrics, Professor of Cellular & Molecular Medicine, UC San Diego, "Applications of Brain Model Technology."

Abstract: Structural and transcriptional changes during early brain maturation follow fixed developmental programs defined by genetics. However, whether this is true for functional network activity remains unknown, primarily due to experimental inaccessibility of the initial stages of the living human brain. We developed cortical organoids that spontaneously display periodic and regular oscillatory network events that are dependent on glutamatergic and GABAergic signaling. These nested oscillations exhibit cross-frequency coupling, proposed to coordinate neuronal computation and communication. As evidence of potential network maturation, oscillatory activity subsequently transitioned to more spatiotemporally irregular patterns, capturing features observed in preterm human electroencephalography (EEG). These results show that the development of structured network activity in the human neocortex may follow stable genetic programming, even in the absence of external or subcortical inputs. Our approach provides novel opportunities for investigating and manipulating the role of network activity in the developing human cortex. Applications for neurodevelopmental disorders and brain evolution will be discussed.  

Bio: Dr. Muotri earned a BSc in Biological Sciences from the State University of Campinas in 1995 and a Ph.D. in Genetics in 2001 from University of Sao Paulo, in Brazil. He moved to the Salk Institute as Pew Latin America Fellow in 2002 for a postdoctoral training in the fields of neuroscience and stem cell biology. He has been a Professor at the School of Medicine, University of California in San Diego since late 2008. His research focuses on modeling neurological diseases, such as Autism Spectrum Disorders, using human induced pluripotent stem cells and brain organoids. He has received several awards, including the prestigious NIH Director's New Innovator Award, NARSAD, , NIH EUREKA Award among others.

April 3, 2020

Terri Meier, CHFP, CSMC, CSBI, CRCR, System Director, Patient Revenue Cycle, UC San Diego Health "The Patient Financial Experience - Changes to Self-Pay Collection Strategies."

Abstract: The U.S. healthcare market is rapidly changing, particularly around the patient billing and payment experience. As patients become increasingly responsible for a larger portion of their medical bills, this has a direct impact on health system economics. Hospitals must learn to adapt in order to compensate for a fundamental change in revenue mix from payer to patient.  

Bio: Terri Meier is currently the System Director Patient Revenue Cycle at UC San Diego Health, providing leadership and strategic direction of UCSDH's Shared Business Office (SBO). SBO unifies the patient-facing element of the billing office for both hospital billing and professional billing with a single bill and single point of contact for customer service. Terri has worked exclusively in healthcare for 40 years serving in multiple leadership roles as well as healthcare consulting.

Immediately prior to UCSDH Terri served for four years as the Director, PFS and Director Professional Billing Organization for Stanford Health Care and the prior 12 years as the Director, Revenue Cycle Operations for Oregon Health & Science University.

Terri holds a healthcare financial management certification from the Oregon Graduate Institute as well as CHFP, CRCR, CSMC, CSBI from the Health Care Management Association (HFMA).  Terri is also the President for the 2020-2021 year for the San Diego Imperial Chapter of HFMA.

March 13, 2020

Anita Bandrowski, PhD., Founder & CEO, SciCrunch, Scientific Lead, Neuroscience Information Framework Project, UC San Diego, "Rigor and Transparency: Tools and Tricks."

Abstract: Reproducibility is a broad topic difficult to understand and in general make much headway with. However transparency and rigor may be a tad more tenable things to understand, measure and address. SciScore, a tool funded under the SBIR mechanism (this is a commercial tool created by SciCrunch a UCSD start up company), that points out which rigor criteria are being addressed by which manuscripts. It processes the methods sections using text mining, classification and a set of rules, looking for things like "blinding", to indicate that investigators have addressed investigator bias. The tool scores each paper on a scale of 1-10 which roughly corresponds to how many of the criteria have been addressed by authors. In 2019, the tool was deployed on the OAI corpus of PubMed Central, which consists of about 2 million papers and it determined that while the overall score is increasing over time, the number of criteria met by authors remains below 50%, including fairly obvious criteria such as describing which sex of animal investigators are using.   

Bio: In the university, I lead the Resource Identification Initiative, an interdisciplinary group devoted to identification of scientific research resources. The initiative is designed to help researchers sufficiently cite the key biological resources used to produce the scientific findings reported in the biomedical literature. The group spans academia, publishers, funding bodies and commercial tool providers. It is the core principle of this group that reproducibility starts with identifiability and we work with many journals to improve the methods section in each and every paper published by helping authors disambiguate their resources with RRIDs. I also lead a company called SciCrunch that helps to create tools and interface with commercial organizations interested in improved rigor in research.

March 6, 2020

Lucila Ohno-Machado, MD, PhD, MBA, Professor of Medicine, Chair, UC San Diego Health Dept. of Biomedical Informatics, Associate Dean for Informatics and Technology, "Patient Preferences for Data Sharing."

Abstract: Dr. Ohno-Mached discussed how 1,200 patients from University of California San Diego and University of California Irvine indicated their preferences for sharing Electronic Health Records for research, which items they preferred to share, and what factors affected their choices. I will also provide an overview of technical and policy strategies to protect patient and institutional privacy when sharing data for research, including some differences between US and EU regulations. 

Bio: Lucila Ohno-Machado, MD, MBA, PhD received her medical degree from the University of São Paulo and her doctoral degree in medical information sciences and computer science from Stanford. She is Associate Dean for Informatics and Technology, and the founding chair of the UCSD Health Department of Biomedical Informatics at UCSD, where she leads a group of faculty with diverse backgrounds in medicine, nursing, informatics, and computer science. Also, she is the PI for the California Precision Medicine Consortium for the NIH All of Us Research Program. Prior to her current position, she was faculty at Brigham and Women’s Hospital, Harvard Medical School and affiliated with the MIT Division of Health Sciences and Technology. Dr. Ohno-Machado is an elected member of the American College of Medical Informatics, the American Institute for Medical and Biological Engineering, the American Society for Clinical Investigation and the National Academy of Medicine. She served as editor-in-chief for the Journal of the American Medical Informatics Association from 2011 to 2018. She directs the patient-centered Scalable National Network for Effectiveness Research, a large clinical data research network covering more than 30 million patients and 12 healthcare systems, and was one of the founders of UC-Research eXchange, a clinical data research network that connected the data warehouses of the five University of California medical centers. She was the director of the NIH-funded National Center for Biomedical Computing iDASH (integrating Data for Analysis, ‘anonymization,’ and Sharing) based at UCSD with collaborators in multiple institutions, as well as other NIH-funded consortia and research projects. Her research focuses on privacy-preserving distributed analytics for healthcare and biomedical sciences. She has received numerous awards for innovations in biomedical informatics.

February 28, 2020

Hooman H. Rashidi, MD, MS, FASCP

Bio: Dr. Rashidi combines his passion for patient care and education with his unique training in bioinformatics and computer programming to create innovative new tools and resources that improve clinical practice and health. Dr. Rashidi is the co-founder, developer, and senior editor of HematologyOutlines, an online atlas used internationally by medical schools and other training programs, and endorsed by the American Society of Clinical Pathology for clinical laboratory scientist and medical technologist training.

Dr. Rashidi also developed the educational app, HemeQuiz1, which includes quizzes in more than 15 categories, quick references, and a game center that allows users to compete with one another. HemeQuiz1 quickly became a top-selling medical app and available world-wide in 30 countries. In addition to the above, Dr. Rashidi has also recently published a print version of his renowned hematology atlas which has become a top selling atlas on Amazon.

Projects currently in progress by Dr. Rashidi and his team include development of artificial intelligence/machine learning (AI/ML) platforms applicable for diagnostic, educational and research use in multiple pathology subspecialty areas and other health science disciplines. Dr. Rashidi also is leading a project to improve clotting using poly-phosphates and silica nano-particles.