High school student Rian Tiwarihas developed a mobile app that uses artificial intelligence to help pregnant people spot nutrient deficiencies by scanning their fingernails. Tiwari’s app uses data from the scans to trigger diet and lifestyle recommendations, aiming to reduce the likelihood of a user developing anemia.
People with anemia have low levels of healthy red blood cells needed to carry oxygen to body tissues. More than 50 percent of pregnant people become anemic, according to the Cleveland Clinic, risking premature birth, low birth weight for the baby, and postpartum depression. Preventing anemia can be as easy as eating iron-rich foods such as beans, red meat, and dried fruit.
In 2020, as a sophomore at South Brunswick High School in New Jersey, Tiwari was learning remotely during the coronavirus pandemic, and he found himself bored. His father suggested he think about ikigai, a Japanese concept referring to something that gives a person a sense of purpose. Tiwari decided that his purpose was to help others through technology.
He began researching several chronic conditions and zeroed in on anemia. He found that apps already existed to monitor hemoglobin levels. But he also learned that a fingernail’s appearance can give clues about a person’s health. Discoloration, ripples, bumps, and other changes in a nail can be signs of nutrient deficiencies or disease. White spots, for example, might indicate a zinc deficiency. Brittle, cracking nails suggest low levels of folic acid.
Tiwari built an app that analyzes fingernail scans for signs of deficiencies in vitamin B12, calcium, zinc, and other nutrients. If these signs exist, the app recommends dietary and lifestyle changes, potentially preventing the development of anemia.
He presented his app at last year’s IEEE International Conference on Intelligent Reality during an event hosted in collaboration with Amazon Web Services. The event focused on how AWS is helping startups that are developing health care technology.
“Presenting my work in front of respected and accomplished engineers felt unreal,” Tiwari says. “I felt humbled by the opportunity to showcase my work and to learn from the other presenters.”
Turning an idea into a mobile app
In 2020 Tiwari and two of his classmates submitted a pitch and a business plan for their product to the Conrad Challenge, a competition for students developing technologies that address a global problem. Selected as finalists, they presented their idea virtually to the Conrad Challenge Innovation Summit at the Space Center Houston. Although they didn’t win, Tiwari moved ahead with developing the app.
He needed help, however. He found Viswanatha Allugunti, a solutions architect at Arohak—a software company in Monmouth Junction, N.J.—on LinkedIn and contacted him. Tiwari says he felt Allugunti’s experience in machine learning and business development would be beneficial.
“Having Dr. Allugunti as a mentor really helped me push myself and this project to the next level,” Tiwari says. “He taught me more about coding, how to file for a patent, and how to more effectively conduct research.”
Allugunti also helped him narrow down his target users from anyone with anemia to pregnant people, focusing on a group with a particular need for the technology, Tiwari says.
Tiwari filed for a U.S. patent for his invention in 2021. He was granted a patent in Germany last year.
How AI can read your nails
Tiwari created the algorithms his app uses by running data sets obtained from open-source website Kaggle through a machine-learning platform. The algorithm classifies images of nails based on their appearance. It looks for cracks, ridges, peeling, and discoloration. Lips and inner eyelids can show signs of nutrient deficiencies as well, but Tiwari chose to analyze nails, given that they are easier to photograph.
The app starts with a photo of a nail taken by a user. It then uses a device-based neural network to analyze the image and classify the nail as healthy or unhealthy, Tiwari says.
If, for example, the app detects the person has a folate deficiency, it recommends foods such as asparagus, spinach, and sunflower seeds, he says.
The app stores some medical information as well as records of the analytics and recommendations from past scans.
To test the app, Tiwari contacted maternal health organizations. He spoke with Reach, a nonprofit that provides mentorship and educational opportunities to people developing technology that enables individualized care for patients. The organization launched its own project—the Maternal Mortality Prevention Program—to ensure pregnant people’s access to health monitoring.
Tiwari says he worked closely with the organization’s president, Fran Ayalasomayajula, to learn what it takes to run a business. She helped him establish an advisory board and do patient outreach, he says.
He is working on getting the app to analyze images of lips and the inner eyelids as well, which can also show signs of nutrient deficiencies, he says. He also wants to add recommendations for medications and vitamin supplements.
Tiwari says he will be piloting the app this year. He plans to make it available for Android and iOS devices.
Inspired by IEEE
Tiwari says he was inspired to pursue a career in science, technology, engineering, and mathematics thanks to his father, Rajiv, who introduced him to IEEE’s products, services, conferences, and publications. The IEEE senior member serves as cochair of the IEEE Future Directions Committee.
Tiwari says IEEE Spectrum and many other award-winning IEEE publications his father subscribes to have made a significant impact on him. Reading articles published in Spectrum helped him discover that he wanted a career in technology, he says.
“So many of the cool technologies IEEE Spectrum writes about, such as quantum computing, sound like they are from a science fiction movie,” he says. “I want to work on these types of technological advancements.”
He says he plans a career in machine learning, specifically natural language processing.
“Helping patients monitor their health is only one of the problems I hope to solve using AI,” he says. “Developing the APT mobile app helped me discover that I want to expand my work into other fields such as language translation.”
The high school senior recently has been applying to colleges. His top choices are Cornell, Georgia Tech, the University of Michigan, and the University of Illinois because of the types of STEM programs they offer.
“The work students and faculty members in these universities are doing is extraordinary,” Tiwari says. “The programs offered at these schools can help me enhance existing technology or develop my own solution to language barriers that exist around the world.”
He says he’s looking forward to becoming an IEEE student member and is looking forward to conferences and networking opportunities.