Project – AI Health Lab

Gender, Socioeconomic, and Racial Disparities in Medical Treatments for COVID-19 Patients

Jiarong Yang — Thu, 01 Jul 2021 19:58:35 +0000

A project aims to investigate socioeconomic and racial disparities in healthcare services during the global COVID-19 pandemic. Funded by Gates Foundation.

A variety of social inequities across domains have emerged apparently and rapidly during the global COVID-19 pandemic, which generate major public health threats to the most vulnerable communities and underrepresented groups. Thus, the objective of our study is to investigate socioeconomic and racial disparities in healthcare services during the global COVID-19 pandemic with the goal to provide evidence for public health policy makers to make informed decisions and to promote equity in healthcare. The expected outcomes will be statistically tested hypotheses backed up with extensive and systematic data analyses to identify patterns and trends of socioeconomic and racial disparities in medical treatments for COVID-19 patients.

PubMed Knowledge Graph

Jiarong Yang — Mon, 07 Jun 2021 21:09:34 +0000

We are building the high-quality PubMed Knowledge Graph by extracting biological entities using BioBERT and disambiguating millions of authors. It is published at Scientific Data.

PubMed^® is an essential resource for the medical domain, but useful concepts are either difficult to extract or are ambiguous, which has significantly hindered knowledge discovery.

To address this issue, we constructed a PubMed knowledge graph (PKG) by extracting bio-entities from 29 million PubMed abstracts, disambiguating author names, integrating funding data through the National Institutes of Health (NIH) ExPORTER, collecting affiliation history and educational background of authors from ORCID^®, and identifying fine-grained affiliation data from MapAffil. Through the integration of these credible multi-source data, we could create connections among the bio-entities, authors, articles, affiliations, and funding.

Data validation revealed that the BioBERT deep learning method of bio-entity extraction significantly outperformed the state-of-the-art models based on the F1 score (by 0.51%), with the author name disambiguation (AND) achieving an F1 score of 98.09%. PKG can trigger broader innovations, not only enabling us to measure scholarly impact, knowledge usage, and knowledge transfer, but also assisting us in profiling authors and organizations based on their connections with bio-entities.

Knowledge Graph can be downloaded at http://er.tacc.utexas.edu/datasets/ped.

COVID-19 Portal: Integrating Literature, Clinical Trials, and Knowledge Graphs

Jiarong Yang — Mon, 07 Jun 2021 19:55:42 +0000

COVID-19 portal is about researchers, universities, and bio entities related to COVID-19 which is based on CORD19 database. In this portal, you will know who is working on which biological entities relate to COVID-19, and the institutional profile on research related to COVID-19 (researchers, papers, and biological entities).

Based on research literature, clinical trials, PubMed knowledge graph, and the drug discovery knowledge graph, we created the COVID-19 portal to portray the research profiles of scientists, bio entities (e.g., gene, drug, disease), and institutions. It provides the following profiles related to COVID-19:

(1) the profile of a research scientist with his/her COVID-19 related publications and clinical trials, which can be ranked by year or by the number of tweets;

2) the profile of a bio entity which could be a gene, a drug, or a disease with articles and clinical trials mentioned this bio entity;

3) the profile of an institution with papers authored by researchers from this institution.

COVID-19 Portal Is Funded By NSF RAPID (2028717), Suit Endowment Fund And Mary R. Boyvey Dean’s Excellence Fund, And Is Hosted By The School Of Information At University Of Texas At Austin.

i-RadioDiagno: Human-centered AI Medical Imaging Diagnosis Tool

Ying Ding — Thu, 20 May 2021 00:58:41 +0000

The shortage of radiologists and the burnout of physicians create an urgent demand for immediate solutions. In this project, we will build the open-source human-centered medical imaging diagnosis tool (called i-RadioDiagno) which adds the prior knowledge of radiologists (e.g., radionics) into the deep learning algorithms to enable automatic or semi-automatic generation of diagnosis notes based on medical images. i-RadioDiagno will be integrated and tested in the clinical practice to enable radiologists and machines to work together by providing feedback loops to improve accuracy and adaptive learning. It will be built on Amazon SageMaker and Apache MXNet on AWS.

i-RadioDiagno is an open-source tool built on Amazon SageMaker Platform that integrates the existing radiomic feature extraction tools, and embeds it into the workflow of clinical decision support system (CDSS) to facilitate the diagnosis process of radiologists.

The user-friendly interface of i-RadioDiagno will allow radiologists to view the radiomic features and calculation in a visual manner, and their diagnosis procedure will become just several mouse clicks based on the pick lists generated by i-RadioDiagno and the diagnosis notes can be created semi-automatically.

By using i-RadioDiagno, all the information or footsteps during the diagnosis process of radiologists are captured, stored, and turned into high-quality labels for medical images which are critical for the AI-powered medical imaging diagnosis.

In this project, we will develop the open source tool called i-RadioDiagno which has the following components:

1) Radiomic feature extraction

2) Image to report model

3) feedback loop through clinical practice