Project Title: Computational Pathology Approach for Characterizing Kidney Biopsies and Predicting APOL1 Risk Variants
How would you sum up your research in one sentence?
Basically, I will try to teach computer to recognize important structures on digital kidney biopsy (glomeruli, tubules etc.), and then try to teach the computer to predict risk and outcome of chronic kidney diseases based on the presentation of these structures.
Provide a brief overview of the research you will conduct with help from the grant.
With the help of this grant, I will first work with renal pathologists to create several deep learning models for automated segmentation of kidney histologic structures on the digital pathology images. We will be using data from Nephrotic Syndrome Study Network (NEPTUNE) digital pathology repository to construct these models. Secondly, once I have the tools for histologic structure segmentation, I can derive sub-visual features from the kidney functional structures and couple with clinical data to develop computer aided diagnosis/prognosis tools. The outcome of the experiments (e.g. the most distinguishable feature to predict outcome) could potentially help uncover deeper biology behind the risk type of chronic kidney diseases.
What impact do you hope your research will have on patients?
APOL1 is a genetic variant in persons of African ancestry that accounts for high risk of chronic kidney diseases, some of the cases can even quickly progress into end stage kidney disease (also known as, kidney failure). It is estimated that the lifetime risk of kidney disease in APOL1 dual-risk allele individuals to be at least 15%. Currently there is no good way to determine the APOL1 risk type of the patient without a genetic test. I hope that I will be able to develop a computer-aided decision support system that accurately determines the APOL1 risk type from digital scans of patients’ biopsies. Thus better inform physicians for better projection disease outcome for direct patient care or kidney transplant recipients.
What inspired you to focus your research on the kidney?
First of all, I am a PhD candidate who works in the computational imaging field, and I noticed that image based artificial intelligence aided diagnosis or prognosis is largely unexplored in the kidney space. Second, as an amateur bodybuilder who consumes excessive amount of protein on a daily basis, I have always paid great attention to kidney health. Thus I decided I wanted to use my expertise to help push the computational imaging in kidney forward, and help facilitate precision medicine development in kidney space.
What are the major challenges to beginning a career in kidney research today?
It is particularly challenging in terms of research funding. The grants or fellowships in kidney research is not as common, which is why I was very much encouraged to apply for this opportunity.
In one sentence, please describe the importance of having grant funding available through the ASN Foundation.
Having grant funding available through the ASN foundation is crucial for me as an international student, because not only does it give me a kick-start as a new and relatively unestablished scholar to conduct the research I am interested in, but also provided a platform for me to showcase my research to a bigger audience.
Something you may not know about me is…
I started painting when I was 4, and I am pretty good at it.