After browsing the internet looking at a forum for computer science, freshman Sasha Tripathi stumbled upon a Congressional App Challenge. Looking for something to do during COVID-19, Tripathi decided to build an app.
Tripathi spent the next six months learning Python and watching YouTube tutorials to learn code. Then, over nine months, he created an app called Skin Doctor. First, skin Doctor detects Melanoma, a type of skin cancer that affects melanocytes or pigment cells. Then, taking a picture of someone’s skin feeds the image into a neutral network that tells how likely that person has skin cancer.
“[My hope] is that my app will detect [the] cancer in the earliest stages and make sure people seek medical help if they suspect something, so they [won’t] have to [pay] for the entire cost. At lower-cost health care is more accessible, so the fact that my app works well is important so that the patients can check up on what could be super serious.” Tripathi said.
Tripathi’s app inspiration came from his previous science fair experiment with classifiers. This machine-learning algorithm assigns a class label to information provided to a computer or software program. He decided to extend the challenge into creating the Skin Doctor app. Tripathi spent his summer reading multiple books for more instruction on how to build his app, in addition to spending six months learning code.
“I brainstormed things that I thought were challenging enough, [so they] wouldn’t [think] ‘he did something that took two minutes to do.’ I wanted to let the judges know that I have worked on this for some time,” Tripathi said.
Tripathi used a program called Dart to create a skeleton of his app. Next, he researched various template apps and replicated their code to fit his vision for Skin Doctor, then fiddled around until it worked.
“I was fairly confident that I could get through the whole project, [but] when I hit a roadblock, it was frustrating. However, I kept my goal in mind of finishing the app, and it helped me get through, “Tripathi said.
During his process, Tripathi had some trouble building his camera. He had to scrap his framework and rebuild it from scratch completely. His neutral network was also not achieving the efficiency he hoped for, and it did not recognize variable melanomas.
“Before you start [working] and every time you hit a roadblock, talk to someone because often verbalizing your ideas can get your brain working. [Also] research helps speed up your process. Even if you get bored while you are researching, you will be done with the app in way less time,” Tripathi said.
Not only did research help with his rebuilding process, but Tripathi’s dad, who is a Computer Scientist, helped get him moving again.
“Whenever I would get stuck, I could ask him, and he would have some insight. His code affects millions of people, [which] is similar to what I want to do, [so] my dad was one of the largest help I had,” Tripathi said.
For years Tripathi has wanted to go into the computer science field and take after his dad, specifically game development. After fixing a few bugs on the Skin Doctor app, Tripathi hopes to publish it to the app store for consumer use.
“It’s a very open field because you can do whatever you want. You can control your world if you’re making a game, so you’re not constrained by the rules, [only] by your limits. Making games is an escape [for me] when having a stressful moment, and the ability to create something I would use means [more] to me.” Tripathi said.
Tripathi submitted his app on Nov. 1. 38 days later, he received an email from Congresswoman Ann Wagner, Chair of the House Suburban Caucus, stating that he had won the challenge.
“With the good feeling that comes with recognition, it compels me to do more artificial intelligence projects and to go to college to study Computer Science. It makes me proud since I put all my hard work into [my app]. It’s a little touching,” Tripathi said.