Dr. Vanessa Julia Carpenter (right) spoke with IDA journalist Anne-Cathrine Jensen on a webinar for the Danish Engineer Association (IDA) about bridging gender bias in health tech and building better health tech.
“By combining design and technology, we can create innovative solutions that matter to the people and the planet,” said Dr. Carpenter in her opening.
However, so far, not all technological inventions made much sense.
Not all things “smart” are useful
“Technology goes in wonderful and weird directions. Back when I started, when the Internet of Things was brand new and hot, there was a heavy put-a-chip-in-it-mentality – let’s make it smart,” recalled Dr. Carpenter.
Interdisciplinary collaboration can help close the knowledge gap in women’s health while incorporating meaningfulness and attention to data safety.
For example, she mentioned an egg tray that tells your phone when you run out of eggs. Another example was the smart wine bottle with a screen that tells you all you want to know about the region, etc.
“Unfortunately, such a bottle produces a lot of electronic waste. There were even smart condoms letting you know how much you’re thrusting and being able to compare with your friends,” said Dr. Carpenter.
A critical point in tech and design
“Technology is fast, cheap, and easy to create. AI can help us write, draw, create, and code. But convenience and surveillance go hand in hand,” Dr. Carpenter pointed out.
While apps on our devices make our lives easier, they also know a lot about us, gathering data on us.
“In an age of increasing AI and interconnection with the digital, and where we don’t know what information we can trust, it is especially relevant now to ask why,” underscored Dr. Carpenter. “This is why I work with designing for meaningfulness.”
Meaningful design incorporates personal development, identity, values, priorities, purpose, and value over function. For example, Dr. Carpenter mentions a step counter—a smart wearable that encourages walking at least 10,000 steps daily.
“It could tell me how many steps I’m walking. But it doesn’t actually matter. It matters whether or not I’ve reached my goals of becoming a healthier version of myself,” Dr. Carpenter explained.
As humans are biased, AI is biased
Dr. Carpenter pointed out that AI is based on data fed by humans. And as humans become biased throughout our lived experience, AI will inadvertently inherit the same bias. Bias in AI arises in different ways depending on how we train it.
“We have three main types of data bias. One is called standard or training bias, based on what the AI has seen. Another is algorithmic bias, based on what the developer has seen. The third is cognitive bias – how our own experience and preference leak into everything we do, ” explained Dr. Carpenter.
Inventions based on data from men
In her book Invisible Women—Exposing Data Bias in a World Designed for Men, Caroline Criado Perez names several examples of data bias that result in things being designed for men and thus not being fit—or even safe—for women.
As AI voice assistants are trained on male-biased data, female voices are less likely to be accurately processed, according to Dr. Carpenter. “Fx voice recognition often doesn’t recognize female voices because they are on a different frequency than male voices, which the AI has been trained on,” said Dr. Carpenter.
On average, men’s voices operate at 100-120 Hertz, whereas women’s are at 200-220 Hertz.
Medicine made for men
This bias also appears in health research. The description inside a medicine package usually distinguishes between grown-ups and children. However, as children metabolize medicine differently, women don’t react the same way to medicine as men.
“Women metabolize some medicines differently from men,” said Dr. Carpenter.
As an example, she mentioned that the recommended dose for women of the sleep medicine Zolpidem, also known as Ambien, is too high.
“That can lead to dangerous side effects like impaired driving and memory issues. This error shows that there’s a critical need for sex- and gender-specific medical research. Drugs must be tested and dosed appropriately for women to prevent harm and improve healthcare outcomes,” underscored Dr. Carpenter.
Unknown female heart attack symptoms
The knowledge gap in men’s and women’s health caused by bias can have dire consequences when it comes to recognizing health issue symptoms in women. There is much less research on what symptoms women experience compared to those of men, for example, symptoms of a heart attack.
Unlike men, it turns out that women do not experience the same symptoms as men when they experience a heart attack. Rather than getting chest pains and discomfort in the left arm, symptoms in women range from upper back, neck, or jaw pain, indigestion, heartburn, nausea, or vomiting to extreme fatigue, dizziness, and shortness of breath.
“The neck or jaw pain is really the big signifier in women,” said Dr. Carpenter.
Most people don’t know about this difference because most research has been based on men. According to Dr. Carpenter, 20 to 25 percent of participants in clinical trials related to heart attacks today are women.
Interdisciplinary collaboration needed
“Addressing bias in tech can benefit everyone, and innovation through interdisciplinary teamwork can resonate far beyond health,” said Dr. Carpenter.
Dr. Carpenter explained how interdisciplinary collaboration can help bridge gender bias in health tech and narrow the knowledge gap in women’s health caused by bias in research, development, and treatments. One example is using the health data collected by smart wearables.
“Smart wearables are revolutionizing how we monitor health and fitness, presenting exciting opportunities in women’s health—an area historically overlooked in medical studies,” said Dr. Carpenter.
By creating health devices designed using the meaningful mechanics method, women can learn more about their health and share the knowledge with their doctors. For example, Dr. Carpenter mentioned a course she teaches at Copenhagen Business Academy (KEA). In the course, students combine technical engineering with jewellery design to prototype beautiful yet functional wearables for some of the 34 known menopause symptoms. The purpose of the designs is to make a tangible impact on women going through menopause.
“Addressing bias in tech can benefit everyone, and innovation through interdisciplinary teamwork can resonate far beyond health,” said Dr. Carpenter.