Results for BIODS

Participants report on recent articles from the Biomedical Informatics literature or their research projects. Goals are to teach critical reading of scientific papers and presentation skills. Summer Quarter consists of critical review of relevant literature led by faculty associated with the Biomedical Informatics Program. May be repeated three times for credit.

This course introduces the data modalities and methods valuable to ask and answer probing and novel questions that advance biomedicine. You will get exposure to a variety of current data types from imaging and omics to patient-centric and digital health generated data types. You will also be exposed to the core methodological concepts useful to analyze these data in isolation or in combination. Specifically, in four separate modules taught by expert faculty in each area the basic principles of each module will be defined and explained. Module 1, Clinical Data and Systems, will explain the basics of Electronic Health Records, and how they operate in health care settings. Next, Module 2, Image Data Health Science, will focus on an introduction to the main imaging modalities in medicine and how methodological analysis using machine vision can be used on large studies. Module 3 will focus on fusing different data streams such as clinical, imaging, molecular and other data modalities. Finally, Module 4 will focus on reproducibility, evaluation and ethical issues when deploying models based on biomedical data, with emphasis on translation to practice. Emphasis will be placed questions, data and methods that advance health and medicine. Primary learning goals for this course include how to frame biomedical health questions, what data are needed to answer those questions, and what methodological constructs can be leveraged to probe and answer those questions. This course is a newly designed course for the PhD program of the Department of Biomedical Data Science but open to all. NOTE: For students in the Department of Biomedical Data Science Program, this core course MUST be taken as a letter grade only.

For graduate and medical students. High-throughput technologies and data science are essential tools in modern stem cell biology and cancer research. Students will gain practical exposure to bioinformatics concepts and techniques required to address biological questions within these research areas. The beginning of the quarter is focused on foundational principles underlying bioinformatics and genomics. Focus for the remainder of the quarter is on direct, hands-on experience with applications to common research problems. Topics include analysis of bulk and single-cell sequencing data, single gene to whole-genome analysis, machine learning, and data visualization. Intended for biology students without a background in computer science, or for students in a quantitative discipline interested in gaining exposure to key challenges in stem cell and cancer genomics. Basic programming experience is recommended but not required.

Introduction to the statistical analysis of several quantitative measurements on each observational unit. Emphasis is on concepts, computer-intensive methods. Examples from economics, education, geology, psychology. Topics: multiple regression, multivariate analysis of variance, principal components, factor analysis, canonical correlations, multidimensional scaling, clustering. Pre- or corequisite: 200.

The practice of data science is inherently a collaborative endeavor, and this course aims to equip data scientists and aspiring data scientists with the necessary skills for effective teamwork in clinical and translational research. The principles discussed in the course apply not only to the medical field but also to the behavioral and social sciences. Throughout the course, participants will explore various aspects of team engagement that are critical for the entire translational research process from study design to data management to data analysis to dissemination of findings. Key questions addressed include how to identify the required expertise for a team, how to integrate and onboard data scientists into teams, how to engage collaborators effectively by successfully leveraging multiple disciplines to jointly solve problems, how to educate the team on the role of the data scientist, as well as the data scientist's role in data collection, cleaning, and management. The course also delves into issues that impact rigor and reproducibility such as authorship, reasonable timelines, interpreting empirical findings, and the importance of statistical analysis plans and study registration. Material is taught through lectures, simulated role-playing exercises, and real-time demonstrations to enhance learning and practical application. Data scientists are working more and more as part of scientific teams. In this course, participants who are (or who are training to be) data scientists will learn optimal team science tools for engaging clinical and translational investigators in the collaborative research process. These principles apply across the medical, behavioral, and social sciences.Topic areas include: optimal team make up from a data science perspective; how to engage collaborators on study design; how to educate collaborators on engaging data scientists; how to educate collaborators on rigor and reproducibility principles such as creating a statistical analysis plan, pre-registering studies, and deciding on authorship; elements that comprise the ideal statistical analysis plan; how to play an integral role during data collection and data extraction phases of the study; and optimal approaches for dissemination of findings to the team and to the research community that adhere to rigor and reproducibility principles and that ensure integration of the data scientist?s voice. In addition to lectures, materials will be taught using simulated role playing and real-time demonstrations of collaborations.

Vast amounts of biomedical data are now routinely available for patients, raging from genomic data, to radiographic images and electronic health records. AI and machine learning are increasingly used to enable pattern discover to link such data for improvements in patient diagnosis, prognosis and tailoring treatment response. Yet, few studies focus on how to link different types of biomedical data in synergistic ways, and to develop data fusion approaches for improved biomedical decision support. This course will describe approaches for multi-omics, multi-modal and multi-scale data fusion of biomedical data in the context of biomedical decision support. Prerequisites: CS106A or equivalent, Stats 60 or equivalent.

Machine learning has driven remarkable advances in many fields and, recently, it has been pivotal in enhancing the diagnosis and treatment of complex brain disorders. Biomedical and neuroscience studies frequently rely on neuroimaging as it provides non-invasive quantitative measurement of the structure and function of the nervous system. Machine and deep learning methods can, for example, refine findings for specific diseases or cohorts enabling the detection of imaging markers at an individual level. This, in turn, paves the way for personalized treatment plans. In this course, we explore the methodological gaps in analyzing high-dimensional, longitudinal, and heterogeneous neuroimaging data and study novel, robust, scalable, and interpretable machine learning models for this purpose.Students have the option to enroll in the class for either 3 or 4 units. All students, regardless of their unit choice, are expected to attend every class session. The primary class content will cover the fundamentals of machine learning, offer some limited hands-on training, and explore the application of ML to neuroimaging. Those opting for 4 units will benefit from an extra hour of instruction weekly, diving deeper into core ML concepts and receiving extended hands-on training. The scheduling of this additional hour will be determined based on the availability of the students enrolled for 4 units to ensure a mutually convenient time slot. Undergraduate students and those who do not have ML backgrounds are advised to take the course for 4 units.

Data scientific analysis of genomics data has transformed biology, enabling myriad discoveries with enormous impacts on human and planetary health. Algorithms and statistics are central to knowledge of human and plant genomic variation, to microbiomes and carbon cycling in the ocean. This class will present the important open problems in the above application areas, pose them as statistical problems and explore core, unifying methods that are used to study them. We will cover diverse scientific application areas focusing on unifying ways they can be addressed statistics and informatics including (i) historical and computer-scientific approaches to addressing these problems where analysis begins with assembling and or aligning to a set of reference genomes (ii) 'statistics-first' approaches that operate on raw sequencing data to perform statistical inference for discovery. This class will present challenges and opportunities in using new methods that do not require a reference to illustrate how the planetary ecosystem can be investigated from a statistics-first perspective: from studies of microbial and plant life to humans. Motivation will be driven by current open and critical problems in planetary health, microbiome research and examples from human genomics. We will investigate statistical and informatic methods that can be used to address these problems including generalized linear models, Pearson's chi-square, permutation testing and present scientific examples/case studies where these tests fail to control the statistical level. Lectures will be pre-recorded with mandatory in-class discussions and problem sessions in class. Evaluation will be based on completion of ungraded problem sets with the major evaluation will be class projects.

The Data Studio is a collaboration between Spectrum (The Stanford Center for Clinical and Translational research and Education) and the Department of Biomedical Data Science (DBDS). The educational goal of this workshop is to provide data science consultation training for students. Data Studio is open to the Stanford community, and we expect it to have educational value for students and postdocs interested in biomedical data science. Most sessions are workshops that provide an extensive and in-depth consultation for a Medical School researcher based on research questions, data, statistical models, and other material prepared by the researcher with the aid of our facilitator. At the workshop, the researcher explains the project, goals, and needs. Experts in the area across campus will be invited and contribute to the brainstorming. After the workshop, the facilitator will follow up,helping with immediate action items and summary of the discussion. The last session of each month is devoted to drop-in consulting. DBDS faculty are available to provide assistance with your research questions. Skills required of practicing biomedical consultants, including exposed to biomedical and health science applications, identification of data science related questions, selection or development of appropriate statistical and analytic approaches to answer research needs. Students are required to attend the regular workshops and participate one to two consulting projects as team members under the supervision of faculty members or senior staff. Depending on the nature of the consulting service, the students may need to conduct numerical simulation, plan sample size, design study, and analyze client data. the formal written report needs to be completed at the end of consulting projects. May be repeated for credit. Prerequisites: course work in applied statistics, data analysis, and consent of instructor.

Best practices for developing data science software for clinical and healthcare applications is a new seminar aimed to provide an overview of the strategies, processes, and regulatory hurdles to develop software implementing new algorithms or analytical approaches to be used in clinical diagnosis or medical practice. Upon completing this seminar, biomedical scientists implementing diagnostics, analytical, or AI-driven clinical decision support software should better understand how to protect, transfer, commercialize, and translate their inventions into the clinic. Topics include: Intellectual property strategies and technology licensing challenges; software development and quality best practices for the clinic; regulatory frameworks for clinical decision support and diagnostics informatics applications. It is open primarily to graduate students across Stanford and combines short lectures, guest industry speakers, and workshop sessions to allow participants to receive feedback on current related projects that are undertaking. Enrollment limited to 25 to allow participants present their current projects. Prerequisites: Basic experience in programing and algorithm or software tool development. Ideally, the participant is actively implementing a new method/process/application in software aimed to be used in the clinic.

Recent breakthroughs in high-throughput genomic and biomedical data are transforming biological sciences into "big data" disciplines. In parallel, progress in deep neural networks are revolutionizing fields such as image recognition, natural language processing and, more broadly, AI. This course explores the exciting intersection between these two advances. The course will start with an introduction to deep learning and overview the relevant background in genomics and high-throughput biotechnology, focusing on the available data and their relevance. It will then cover the ongoing developments in deep learning (supervised, unsupervised and generative models) with the focus on the applications of these methods to biomedical data, which are beginning to produced dramatic results. In addition to predictive modeling, the course emphasizes how to visualize and extract interpretable, biological insights from such models. Recent papers from the literature will be presented and discussed. Experts in the field will present guest lectures. Students will be introduced to and work with popular deep learning software frameworks. Students will work in groups on a final class project using real world datasets. Prerequisites: College calculus, linear algebra, basic probability and statistics such as CS 109, and basic machine learning such as CS 229. No prior knowledge of genomics is necessary.

This course offers an overview of statistical foundation for modern clinical trial design in precision medicine research. Starting from a quick review of traditional clinical development paradigm through Phase I to III clinical trials for medical product approval and Phase IV post-marketing studies for safety evaluation, and challenges in the time and society costs, we will introduce recently developed innovative designs and their statistical methodology across all phases of clinical trials. You expected to learn the statistical considerations for novel phase I-II trial designs, master protocols for umbrella, platform and basket trials, adaptive and enrichment designs including subgroup selections, estimand, surrogate and composite endpoints, integration of real-world evidence and patient-focused medical product development, and meta-analysis of clinical trial endpoints. Prerequisites: Working knowledge of statistics and R.

The importance of software to science has grown tremendously over the past 20 years. Proper use of standardized Software Engineering techniques, such as cloud computing, testing, virtualization, testing, and source control, is often necessary for high-quality, and replicable science. Software Engineering for Scientists is designed to help researchers, scientists, and non-domain-experts gain hands-on knowledge of the tools and practices that will make your day-to-day work more efficient and less error-prone, be it in academia or industry. Students will learn to adopt the most important and germane of the techniques used in the real world (from startups to large companies), and we will provide you with a good understanding of the tools, approaches, and tradeoffs inherent in writing any kind of program. The class is taught by an expert with 20 years of experience building software, managing engineering and product teams at companies including Google and Twitter who now works primarily in biotechnology and a Professor in the Biomedical Data Sciences with more than 30 years of experience teaching in bioinformatics at both Stanford and UCSF. Pre-recorded lectures will be provided and will cover topics from both a theoretical and practical perspective. In person lectures will be primarily interactive; we will spend time answering students' questions and talking about how these learnings could be useful to their research. There will be a number of assignments and a final project which can be based on students' existing research.

Applications of data science techniques to current problems in biology, medicine and healthcare. To receive credit for one or two units, a student must attend every workshop. To receive two units, in addition to attending every workshop, the student is required to write a two page critical summary of one of the workshops, with the choice made by the student.

Applications of data science techniques to current problems in biology, medicine and healthcare. To receive credit for one or two units, a student must attend every workshop. To receive two units, in addition to attending every workshop, the student is required to write a two page critical summary of one of the workshops, with the choice made by the student

Applications of data science techniques to current problems in biology, medicine and healthcare. To receive credit for one or two units, a student must attend every workshop. To receive two units, in addition to attending every workshop, the student is required to write a two page critical summary of one of the workshops, with the choice made by the student

Generative AI and large-scale self-supervised foundation models are poised to have a profound impact on human decision making across occupations. Healthcare is one such area where such models have the capacity to impact patients, clinicians, and other care providers. In this course, we will explore the training, evaluation, and deployment of generative AI and foundation models, with a focus on addressing current and future medical needs. The course will cover models used in natural language processing, computer vision, and multi-modal applications. We will explore the intersection of models trained on non-healthcare domains and their adaptation to domain-specific problems, as well as healthcare-specific foundation models. Prerequisites: Familiarity with machine learning principles at the level of CS 229, 231N, or 224N

Each edition of the course focuses on one topic of research or translation. Students read, present and discuss papers from the literature.

This project-based course delves into the cutting-edge of Generative Artificial Intelligence (AI) and its transformative applications in the healthcare domain. As technology continues to evolve, so does the potential for AI to revolutionize healthcare practices, from diagnostics to personalized treatment plans. Participants will learn about the latest advances in Generative AI, exploring state-of-the-art models and techniques tailored for healthcare challenges. Students will be introduced to Human-Centered Design methodology -- involving empathy and needs finding, prototyping and iteration. Class projects will focus on deployment of Generative AI using datasets such as population biobanks, and training models from these population-scale datasets. Key topics covered include the utilization of Generative AI in medical image synthesis, enhancing diagnostic capabilities, and using genomes and protein language models for variant effect prediction. The course also navigates the ethical considerations surrounding the use of generative models in healthcare, addressing issues of privacy, bias, and interpretability. Through a combination of theoretical insights and hands-on practical sessions, participants will gain a deep understanding of how Generative AI is reshaping the healthcare landscape, and how they could have a positive impact. Guest speakers from venture capital and industry with real-world examples will illustrate successful applications of generative models in medical imaging, drug discovery, and patient care, and discuss the challenges they see in translation from research to implementation.Students will need to visit the link and fill out an application before they get a reg. code to register for course. https://dschool.stanford.edu/classes/generative-ai-for-healthcare

For students wishing to receive credit for directed reading or research time. Prerequisite: consent of instructor.

A seminar-style course with lectures on a range of computational topics important for modern data-intensive science, jointly supported by the Statistics department and Stanford Data Science, and suitable for advanced undergraduate/graduate students engaged in either research on data science techniques (statistical or computational, for example) or research in scientific fields relying on advanced data science to achieve its goals. Seminars will alternate a presentation of a topic, usually by an expert on that topic, typically leading to exercises applying the techniques, with a follow up lecture to further discuss the topic and the exercises. Prerequisites: Understanding of basic modern data science and competence in related programming, e.g., in R or Python. https://stats352.stanford.edu/

This course has the following broad goals: (1) To ensure that Stanford graduate students in data science are intentionally trained to effectively mentor people who may be different from them. (2) To sustainably develop pathways to increase access to higher education and to Stanford graduate programs in data science for individuals from backgrounds currently under-represented in those fields. During weekly class meetings, graduate student participants will learn strategies to create an inclusive environment, approaches to effective mentoring and coaching, and techniques to develop a personalized curriculum with the course staff and guest speakers. They will also be paired with current undergraduates from non-R1 schools with an interest in data science, recruited in partnership with faculty from those institutions. Participants will meet online weekly for one-on-one mentorship where you will expose your mentee to research in data science. During weekly online meetings, you will work with your mentee on a range of activities, planned with assistance from course staff, including planning their course of studies, navigating internship opportunities and preparing applications; tutoring in some aspects of data science; and guidance in engaging in mini-research projects, depending on their interests.

Students undertake investigations sponsored by individual faculty members. Prerequisite: consent of instructor.