Rajaram Anantharaman is a Ph.D. student who was recruited through the Provost’s Strategic Funding Initiative to conduct research in the area of Big Data and Analytics. The selection process is very rigorous, and only the best applicants are considered. Raj started working with convolution neural networks and their application to oral diseases this spring. He envisions a future where detecting oral cancer may, “be as routine as drawing blood to detect other diseases.”
What do you hope to accomplish with your research?
The incidence of oral cancer worldwide is around 500,000 new cases every year, accounting for approximately 3% of all malignancies, thus creating a significant worldwide health problem (Johnson, 2011). Oral cancer has a tendency to be detected at a late stage which is detrimental to the patients because of its high mortality and morbidity rates. Early detection of oral cancer is therefore important to reduce the burden of this devastating disease (Messadi, 2013). My research deals with using computer vision and big data to detect oral precancerous conditions in its earliest stages. The hope is that my research will culminate in a solution that health workers can use to detect oral disease, get verified answers from the accompanying oral health knowledge base, and get patients referred to a dentist or oral surgeon for further diagnosis.
What got you interested in this topic?
I took a preliminary class in computer vision from Dr. Lee and it piqued my interest. We studied how we can train a convolution neural network to identify and classify images. Having worked at Cerner and other places in healthcare IT for over a decade, my natural inclination was to find relevant applications of this emerging technology to combat a health problem. Cancer is a terrible disease and there are millions of doctors, researchers, and other health workers already trying diligently to find a cure. My grandmother died of this terrible disease. While a great deal of biomedical research is happening in oral cancer, there is an opportunity to use computer vision, big data, and other computer technologies to help advance the cause.
Why did you choose UMKC?
My research requires help from different disciplines including computer science, medical informatics, and oral sciences. UMKC has all of the necessary resources for researchers like me who are trying to bring together different disciplines. UMKC’s programs are intrinsically designed to make this kind of collaboration happen.
What faculty are you working with?
I am primarily working with Dr. Yugyung Lee in the department of Computer Science. Apart from Dr. Lee, I am being advised by several other faculty members including Dr. Melanie Simmer-Beck from the department of Dental Public Health, Dr. Arif Ahmed from the department of Public Affairs and Dr. Mary Gerkovich from the department of Biomedical Informatics.
What do you hope to discover?
I hope to discover if it is really possible for an A.I. trained computer vision program to detect and classify oral diseases with any degree of confidence. We share a vision with IBM research (IBM Blog Research, 2016) who is doing similar research work in the area of melanoma. Our vision is that taking pictures to diagnose oral cancer and/or other soft tissue oral diseases might one day be as routine as drawing blood to detect other diseases.
1) Johnson NW, Warnakulasuriya S, Gupta PC, et al. Global oral health inequalities in incidence and outcomes for oral cancer: causes and solutions. Adv Dent Res. 2011;23 2:237–246.
2) Messadi, D. V. (2013). Diagnostic aids for detection of oral precancerous conditions. International Journal of Oral Science, 5(2), 59–65. http://doi.org/10.1038/ijos.2013.24
3) “Identifying skin cancer with computer vision.” IBM Blog Research, IBM Corporation, 14 Nov. 2016, www.ibm.com/blogs/research/2016/11/identifying-skin-cancer-computer-vision/. Accessed 25 Sept. 2017.