Dasariraju ’23 Published in Bioengineering Journal
This summer, Satvik Dasariraju ’23 published his science research project in the latest issue of the journal Bioengineering.
This summer, Satvik Dasariraju ’23 published his science research project in the latest issue of the journal Bioengineering.Through the iResearch Institute, a summer research program geared towards high school students pursuing scientific research projects, Dasariraju received mentorship from a Stanford student and the head director of iResearch to create a method of diagnosing leukemia using machine learning algorithms.
On his interest in science, Dasariraju said, “I’ve always been really interested in science. Ever since I was four I have been reading a lot of books about science. This summer, when I did the program at iResearch Institute, I really enjoyed learning about biomedical engineering and its importance while being able to implement computational skills and engineering techniques to solve medical problems.”
Writing for the journal also gave Dasariraju experience and insight into scientific writing and the process of scientific inquiry. “Writing a scientific research paper is very different from writing papers in English...but the biggest thing I took away is that if you’re passionate, focused, and really driven, then there’s nothing that will stop you,” Dasariraju said.
For his research process, Dasariraju’s first step involved processing thousands of images of white blood cells from Internet databases to identify which important parts to isolate. Secondly, he extracted 16 different leukocyte features, creating multiple algorithms to aid the process. Through his research, he identified two novel color features of white blood cells. After the features were identified and extracted, Dasariraju trained a machine-learning algorithm in Python to classify and detect immature blood cells that are typically found in leukemia. This machine learning classifier achieved 92.99 percent accuracy in detecting immature leukocytes, and 93.45 percent accuracy is classifying them, higher accuracies than current detection methods.
The implications of his projects are far-reaching. Explaining his program, Dasariraju said, “[The algorithm] can be used as a support tool for doctors to efficiently, accurately, and quickly diagnose leukemia. The current method for diagnosing leukemia is very time consuming and prone to a lot of error because the clinician needs to look at each blood cell and classify it…An automatic approach to classifying blood cells is much more efficient.” The current length of leukemia diagnosis is around two to three weeks, which requires a series of imaging and blood tests that are processed by a physician. When considering the severity of leukemia, saving time in a diagnosis could become the difference between life and death.
Looking forward, Dasariraju hopes to publish more papers in high school and college while continuing his involvement in the bioengineering and computer science fields. On his accomplishment, he said, “Don’t let your age limit you. Just because you don’t have a Ph.D. doesn’t mean you can’t do research. Be driven and be motivated in your goals. No matter what age you are, you can always contribute to science and gain a better understanding of the world to help other people.”