News & Announcements

Discover the latest research developments and scientific achievements of the BMIR team

Dr. Tina Hernandez-Boussard Published with Healthcare Fraud Prevention Partnerships

Identifying Fraud, Waste, and Abuse in Genetic Testing

Dr. Tina Hernandez-Boussard and her team of researchers from the Stanford Center for Biomedical Informatics Research (BMIR) at Stanford identify and examine in a recently published white paper, the potential for fraud, waste, and abuse in genetic testing.

Genetic testing has rapidly grown over the past decade. Collectively, the healthcare industry faces billions in annual costs for these tests. This emerging area of medicine is susceptible to wide-scale fraud, waste, and abuse that increases healthcare spending and negatively affects the integrity and quality of the current healthcare system.

This white paper was developed in partnership with the Healthcare Fraud Prevention Partnership (HFPP), a public-private partnership of healthcare payers and allied organizations that seeks to identify waste, fraud, and abuse across the healthcare sector, and the Stanford University School of Medicine.

Tina Hernandez-Boussard,  Ph.D., is Associate Professor of Medicine (Biomedical Informatics) and of Biomedical Data Science within Stanford’s Center for Biomedical Informatics Research division.

Dr. Olivier Gevaert is an Author of an Article Published in npj Genomic Medicine

Dr. Olivier Gevaert led a study on “Whole-slide Images Reflect DNA Methylation Patterns of Human Tumors” published by npj Genomic Medicine showing an intriguing connection between epigenomic signatures and whole slide images of tumor tissues discovered using machine learning. Epigenomic patterns, in particular DNA methylation signatures, are changes on top of the DNA. DNA methylation changes do not change the DNA sequence, but instead, they affect how cells activate genes. In this study we show that these epigenomic patterns are reflected in whole slide images of tissues that are routinely captured by pathologists examining tumors. The authors concluded that their results provide new insights into the link between histopathological whole slide images and epigenomic signatures in human cancers. This work underlines the potential of associations between tumor tissue as visualized on whole slide images with the underlying DNA methylation states, providing new insights and understanding of how tumors develop at multiple scales. This may provide benefits in the future for pathologists to more routinely use available whole slide images as a proxy for determining the epigenomic make-up of human tumors.

DNA methylation is an important epigenetic mechanism regulating gene expression and its role in carcino-genesis has been extensively studied. High-throughput DNA methylation assays have been used broadly in cancer research. Histopathology images are commonly obtained in cancer treatment, given that tissue sampling remains the clinical gold-standard for diagnosis.
In this work, we investigate the interaction between cancer histopathology images and DNA methylation profiles to provide a better understanding of tumor pathobiology at the epigenetic level. We demonstrate that classical machine-learning algorithms can associate the DNA methylation profiles of cancer samples with morphometric features extracted from whole-slide images.
Furthermore, grouping the genes into methylation clusters greatly improves the performance of the models. The well-predicted genes are enriched in key pathways in carcinogenesis including hypoxia in glioma and angiogenesis in renal cell carcinoma. Our results provide new insights into the link between histopathological and molecular data.
In addition to Dr. Gevaert, the team further involved the first author, Dr. Hong Zheng, a postdoc in the Gevaert Lab; Mr. Alexandre Momeni and Pierre-Louis Cedoz, two students in the lab; and Dr. Hannes Vogel, a professor in the Department of Pathology at Stanford.

Dr. Olivier Gevaert Published in GigaScience

Dr. Olivier Gevaert is the lead author of the article “Benchmark of Long non-coding RNA Quantification for RNA Sequencing of Cancer Samples,” which was published in December 2019. The authors (Hong Zheng, Kevin Brennan, Mikel Hernaez and Olivier Gevaert) concluded that “considering the consistency with ground truth and computational resources, pseudoalignment methods Kallisto or Salmon in combination with full transcriptome annotation is our recommended strategy for RNA-Seq analysis for lncRNAs.

Olivier Gevaert, Ph.D., is Assistant Professor of Medicine (Biomedical informatics) and of Biomedical Data Science within Stanford’s Center for Biomedical Informatics Research division.

The Shah Lab Green Button Team Awarded The 2019 ISP Star Award

The Shah Lab Green Button team (center, from left to right: Ken Jung, Vladimir Polony, Alison Callahan, Saurabh Gombar) received a 2019 ISP Star Award, which recognizes individuals from the Stanford Medicine community who, through their extraordinary efforts, embody the strategic priorities of our Integrated Strategic Plan (ISP): Value Focused, Digitally Driven, and Uniquely Stanford. By developing a technology that amplifies the clinical impact of our world-class faculty and creates outcome-driven recommendations at the point care, the Green Button exemplifies the transformative power of collaboration at Stanford Medicine. They were presented the award by David Entwistle (President and CEO of Stanford Health Care; left), Lloyd Minor (Dean of the School of Medicine; second from right) and Paul A. King (President and CEO of Stanford Children's Health; right).



Blood test could transform tuberculosis diagnosis, help monitor treatment effectiveness

A simple blood test that can accurately diagnose active tuberculosis could make it easier and cheaper to control a disease that kills 1.5 million people every year.

Four image feature model predicts the presence of EGFR mutations in non small cell lung cancer

The Gevaert lab together with the department of radiology developed a model based on only four image features that predicts the presence of EGFR mutations in non small cell lung cancer. This work shows that macroscopic phenotypes such as tumor shapes, textures and tumor environment reflect their molecular properties. 

Stanford HAI

 The gap in our ability to systematically evaluate how clinical decision support affects how clinicians think and behave hinders the design and implementation of safe and effective AI for patient care. BMIR researchers aim to create a proof of concept of a method for generating a comprehensive, low-bias feature pool of downstream clinical decisions made in the real world setting that can be potentially generalized to become a new way to study the effects of AI on how clinicians make decisions.



Gevaert and Gentles Featured in New PanCan Papers

BMIR'S Andrew Gentles and Olivier Gevaert, and Aaron Newman from the Department of Biomedical Data Science, contributed greatly to the completion of The Cancer Genome Atlas (TCGA) PanCan initiative, a decade-long process that culiminated in the release of a series of 27 papers.

Researchers Help Develop Technique for Assessing, Reducing Risk of Future Stroke

Using health records, Stanford researchers developed an algorithm for scoring the risk of a stroke patient experiencing a heart condition known as atrial fibrillation, a major risk factor for a second stroke.

Gevaert Lab featured in Science Translational Medicine editorial on meta-analysis method

Katie Planey developed a method called CoINcIDE, a meta-analysis method to look for clusters across several data sets.

Gevaert and Cheerla Create Web-Based Tool for Cancer Diagnosis and Treatment Options

What do a Stanford Assistant Professor and Monte Vista High School student have in common? Disrupting the current state of the art in cancer diagnosis and treatment!

Hernandez-Boussard and Colleagues Find that SSRIs Reduce the Effectiveness of Hydrocodone and Codeine

Research led by BMIR's Tina Hernandez-Boussard, colleague Ian Carroll and grad student Arjun Parthipan found that patients on SSRIs such as Zoloft, Paxil and Prozac who were prescribed prodrug opoids such as hydrocodone or codeine experienced more post-surgery pain than patients not on SSRIs. Hernandez-Boussard's team built a machine-learning algorithm that predicts how a patient will respond to different types of opoids. 

Read the NPR article

Read the Scope article


Wanted: More Data, the Dirtier the Better

The computational immunologist Purvesh Khatri embraces messy data as a way to capture the messiness of disease. As a result, he’s making elusive genomic discoveries

To distill a clear message from growing piles of unruly genomics data, researchers often turn to meta analysis a tried and true statistical procedure for combining data from multiple studies.

Adventures of a Proud Data Parasite

Analyzing other people's data has been linkened to research parasitism. Purvesh Khatri is "repurposing data to ask and answer questions that are not addressable using traditional approaches". He is "a proud data parasite."

“I think I’ve finally found my calling in immunology.”

Having come to the United States from India in 1998, Purvesh Khatri, originally planned a career as an electronics and communications engineer.

Stem Cell Discoveries

The Gevaert lab discovered two types of intestinal stem cells using bioinformatics analysis of transcriptome data of several populations of intestinal stem cells. This work was a highly collaborative effort in the context of a national network of intestinal stem cell researchers.

Khatri and Team Discover Biomarker for Flu Susceptibility

How should an algorithm generate recommendations for patient care?

Dr. Jonathan Chen and his colleagues are working to answer the question, "How should an algorithm generate recommendations for patient care?" This question was tackled as the latest step in Dr. Chen’s quest to build OrderRex, a tool that will mine data from electronic health records to inform medical decisions.

Mining information contained in clinical notes could yield early signs of harmful drug reactions, study shows

Nigam Shah and his colleagues have developed a method to extract data from records of interactions between doctors and patients to yield information about possible drug side effects.