Less user input and more silent monitoring will define the future of medical technology.
Ali has always smoked. Although he says that he started smoking one night in high school, the truth is that tobacco has always been a part of his life--his father was a smoker. Now at 25, Ali goes through half a pack of Newports every day.
“It’s down from last year, definitely.” Ali mutters. “I’ve been getting into Juuls. I hate the taste; but it’s helping me to wean off.”
Like many smokers, it is hard for Ali to ignore the effects of tobacco on the body. From the “Stop Smoking” signs printed in bold on the face on billboards to the foreboding television ads of smokers peeling off bits of their skin to figuratively ‘pay’ for a pack of cigarettes. Ali often considers his fate in these stark terms--for him, it’s not a question of if he gets cancer, it’s when.
Despite knowing the risks, Ali still finds it difficult to see a physician regularly. “I don’t have insurance right now. And that’s expensive. But when I did have insurance, I still wouldn’t go because it’s time that I didn’t have. I need to drive there, I need to make an appointment.”
Even when he does make it to a physician, Ali tends to avoid talking about his smoking. “It’s embarrassing man. It’s like, they know I’ve quit before, and they say, ‘so you’re smoking again,’ ‘oh you’ve increased the number of cigarettes!’ So it’s like, It’s hard to be a smoker in a world where everyone knows that smoking is bad.”
On the other end of the spectrum, Maurice is a 55-year-old laborer turned construction consultant. He oversees asbestos removal services for several large construction companies. “Everyone knows that they are at risk. Not even a question.”
While all contractors are legally required to medically screen their employees each year, Maurice notes that the quality of medical evaluation dramatically varies between organizations: “I’ve employed people working in asbestos abatement for decades and [not a single one] have a single issue with their lungs, no pain or anything. People who get COPD [or] tumors, I’ve seen, work at companies that cut corners on PFT (Pulmonary Fit Tests) or are lax on using safety equipment.”
Respiratory diseases such as lung cancer and Chronic Obstructive Pulmonary Disease (COPD) are often asymptomatic in their early stages. As a result, screening adherence rates for high-risk patients is a paltry 2%. While 88% of all patients diagnosed at stage I or II lung cancer survive for 10 years or more, only 16% of patients are identified at this stage. At later stages the prognosis is grim: only 5% of patients survive past 5 years.
Healthcare technology is rapidly transitioning from clinics and hospitals to the home. Heart rate monitors and sleep trackers are ubiquitous across fitness devices. Samsung phones can measure your blood oxygen saturation (SpO2). The Apple Watch can take an EKG reading from your wrist alone and is approved to detect irregular heartbeats (atrial fibrillation). Google and Microsoft are building healthcare platforms for monitoring consumer health from their smartphones.
The common denominator between all of these platforms is the need for active user engagement. Users must be resolute when tracking their health metrics across several platforms to get optimal results. Every calorie dutifully logged. Every step constantly tracked. Every waking hour sitting, standing, and sleeping carefully monitored. While retention rates for health apps are approximately 30%, the users that receive the most benefit from current health tech are a self-selecting population that are already conscientious of their health. For at-risk people like Ali and Maurice, this overly-complicated user interaction paradigm is a barrier that discourages rather than empowers.
The future of smart home devices dramatically changes this outlook.
The major theme at the 2019 Consumer Electronics Show was re-imagining the home of the future. Everything from toaster ovens to window shades to alarm clocks are connected and interconnected; powering these experiences are the next generation of digital assistants.
The data collected from these devices is an enormous opportunity for health monitoring. It’s not impossible to imagine a smart stove logging what foods you cook, and automatically notifying consumers of potential vitamin deficiencies. Smart medication trackers can notify patients of missed prescriptions and can work in tandem with a smart fridge to identify foods that might interfere with course of treatment. Smart air filters can integrate citywide pollen data with the current air quality in bedrooms and remind users to take an antihistamine before allergies are set off.
A personalized, data-driven health model will have an impact on users--from detection to treatment. What’s important, is that the beneficial insights are developed silently in the background, without explicit interaction, logging, or data entry from users. These advances can dramatically improve the screening, morbidity and mortality for chronic illnesses such as lung cancer, heart disease, and neurodegenerative disease. However, does the utility of smart health services outweigh losing aspects of our privacy at home? How much of our health data should we allow companies to collect, and what kinds of data?
Canairy is a smart device that monitors respiratory health silently while keeping privacy intact. Our approach is twofold, we first detect dramatic changes in voice due to respiratory disease; we then use our machine learning algorithms to classify detected coughs as one of several types of respiratory disease, including early stage lung cancer, COPD or asthma. When a malignant cough is detected, a notification with information on screening services is sent out, automatically, and without any need for user input. Importantly, all user data is collected locally on a Canairy device, and owners will have full control to manage any information collected.
For Ali, having agency in understanding his risk for respiratory disease significantly motivates him to see a physician. “I think if there was something automated, it would take away the stigma of getting yourself checked.”
At Canairy, our mission is to serve people at-risk for respiratory disease like Ali and Maurice; our goal is to build a smoke detector for respiratory health. Contact us to learn more: firstname.lastname@example.org
Canairy is a machine learning model deployed in smart home devices to detect the presence and progression of respiratory diseases such as asthma, COPD, and lung cancer.
From Universities: Icahn School of Medicine at Mount Sinai; Carnegie Mellon University; Cornell University; New York University; and The New School..
Team Members: Rayees Rahman, Chi-Chi Bello, Emily Wang, Amy Lei, Flora Wu, Jennifer Hsieh, Alexandra De Rosa.