This article originally appeared on VICE Mexico. Leer en Español.
When Julian Ríos was eight years old, a cancer misdiagnosis put his mother's life in danger. A few years later, she was misdiagnosed again. Now 18, Ríos runs a company that uses artificial intelligence to detect and prevent illnesses like cancer and diabetes.
"The inspiration came from a late diagnosis that they gave my mom," Ríos explains. "An error on the part of the doctors and the mammogram resulted in a late-stage cancer diagnosis that put her life at risk." The teenager dedicated himself to researching everything he could about cancer, including the problems that plague medical innovation and the challenges surrounding early breast cancer detection. "Then, I reached out to the two smartest guys I know and we got to work."
That's how Ríos founded Higia Technologies just over two years ago, along with Antonio Torres and José Ángel Lavariega. Now he has a team of 15 people in Mexico City and Monterrey in Mexico, Spain, Colombia, China, and Korea; specialists in biomedicine, artificial intelligence, computer science, and industrial design, along with a software team that develops mobile platforms.
His first product, the Eva Bra, is an insertable bra cup that uses thermal biosensors to analyze temperature changes and compares them against a database to detect possible health risks. The device is placed on the chest and is used once a week for 60 to 90 minutes. Ríos sold 75 percent of the initial 5,000 first-phase units a little more than a week after launch. He's also working closely with public and private institutions like the Instituto Mexicano de Seguro Social (Mexican Social Security Institute), TecSalud, and a number of anti-breast cancer associations, as well as Apple, Google, insurance companies, and more.
We spoke to Ríos to learn more about how the cup works and how he plans to use artificial intelligence to prevent disease.
Watch: Stella McCartney on Fake Fur, Rihanna, and Well-Fitted Bras
BROADLY: How did the idea of making Eva come about?
JULIÁN RÍOS: Initially, the idea was to quantify three things: The coloration, the texture, and the temperature of the breasts as a means of [cancer] detection. As we went along, we realized that color is really subjective, and texture wasn't a variable that aligned well with the technology. In other words, there weren't sensors that were really reliable to detect how elastic the fabric was. So we stuck with just the variable of temperature, and that's what Eva focuses on.
We also didn't know whether it was going to be a cup or a bra. The problem with a bra is the huge number of permutations that exist: You'd have to test with a number of cup and torso sizes. To avoid these variables, we decided that the product should be a cup or an insert, and not a bra.
How do these cups work?
The human body has this phenomenon called angiogenesis. When there's a tumor or an abnormal mass, that object is surrounded by blood vessels because the tumor needs to grow, and blood gives oxygen and nutrients that allow it to do just that. So, when you have breast cancer, there's more blood flow, and since blood is the main fluid circulating in the body, it increases one's temperature. Distinct tumors at distinct depths and of distinct sizes in the breast have thermal curves with very marked characteristics. In other words, a stage-three tumor looks thermally distinct compared to one in stage one or one in stage four.
Eva is a collection of thermal biosensors that are placed on the woman's breast and they record the temperature of the breast every second, assessing how the blood is flowing between the breasts. Once it has an idea about the blood circulation, the data is analyzed using artificial intelligence algorithms. These algorithms are essential; they go from being temperature readings to a diagnosis. This occurs in the following way: You have a patient's thermal conductivity curve and you analyze that curve, comparing it to an enormous database that includes thousands of confirmed cancer cases, or thousands of confirmed cases of mastitis or inflammation, etcetera.
Then the algorithm begins to say, "OK, this looks a lot like stage-two cancer, or a healthy breast, or a cyst." It's very easy to use. It's just taking the device and placing it on the breast; then, you take your phone, open the app, begin the reading, and all of the data is sent in real time via mobile app. The algorithms begin to work, and voila, produce a risk evaluation.
What are the data points that the device compare its information against?
Before thinking about commercialization or sales, clinical studies have to be conducted. These studies involve putting the device on thousands of women with breast cancer, thousands of women who are free of any type of inflammation, women who are ovulating [or] pregnant, women with reconstructive surgery, so that when the algorithm stumbles across an outlier—in other words, when it encounters the real world via the device—that data can be compared. This data collection has been done this whole year along with partners like Instituto Mexicano del Seguro Social and public and private hospitals.
Are you working on other products that use this technology?
Yes. One of them is called Aquiles ("Achilles" in Spanish), which is a rug that you use after taking a bath to measure how the temperature on the sole of your foot is changing. The blood is moving in your foot, and this can predict whether the patient is at risk of foot amputation because of diabetes. The other product is called Adán ("Adam"), which uses the same technology as Eva, but for testicular cancer.
The objective isn't just to create extraordinary products, but a strong company that becomes a means of attacking illnesses through artificial intelligence, current technologies, and science. Not just cancer, but in every aspect of daily life.