Illness-detecting sensors: Detecting diseases before it’s too late

IMAGE CREDIT:
Image credit
iStock

Illness-detecting sensors: Detecting diseases before it’s too late

Thrive from future trends

Subscribe today to equip your team with the leading trend and foresight platform used by multidisciplinary, future-focused teams working across departments in Strategy, Innovation, Product Development, Investor Research, and Consumer Insights. Convert industry trends into practical insights for your business.

Starting at $15/month

Illness-detecting sensors: Detecting diseases before it’s too late

Subheading text
Researchers are developing devices that can detect human illnesses to increase the likelihood of patient survival.
    • Author:
    • Author name
      Quantumrun Foresight
    • October 3, 2022

    Post text

    Illness-detecting sensors can help monitor a virus’ spread and identify potentially developing cancers. During the COVID-19 pandemic, the use cases for illness-detecting sensors became more apparent. 

    Illness-detecting sensors context

    Early detection and diagnosis can save lives, particularly for infectious diseases or illnesses that may take months or years for symptoms to show. For example, Parkinson’s disease (PD) causes motor deterioration (e.g., tremors, rigidity, and mobility issues) over time. For many people, the damages are irreversible when they discover their illness. To address this issue, scientists are researching different sensors and machines that can detect illnesses, from those that use dogs’ noses to those that employ machine learning (ML). 

    In 2021, a coalition of researchers, including the Massachusetts Institute of Technology (MIT), Harvard University, Johns Hopkins University in Maryland, and Medical Detection Dogs in Milton Keynes, found that they can train artificial intelligence (AI) to mimic the way dogs smell out disease. The study found that the ML program matched the success rates of dogs in detecting certain illnesses, including prostate cancer. 

    The research project collected urine samples from both diseased and healthy individuals; these samples were then analyzed for molecules that could indicate the presence of disease. The research team trained a group of dogs to recognize the smell of diseased molecules, and researchers then compared their success rates in identifying illness to those of ML. In testing the same samples, both methods scored more than 70 percent accuracy. Researchers hope to test a more extensive data set to pinpoint the significant indicators of various diseases in greater detail. Another example of an illness-detecting sensor is the one developed by MIT and Johns Hopkins University. This sensor uses dogs’ noses to detect bladder cancer. However, while the sensor has been successfully tested on dogs, there is still some work to be done to make it suitable for clinical use.

    Disruptive impact

    In 2022, researchers developed an e-nose, or an AI olfactory system, that can potentially diagnose PD through odor compounds on the skin. To build this technology, scientists from China combined gas chromatography (GC)-mass spectrometry with a surface acoustic wave sensor and ML algorithms. The GC could analyze odor compounds from sebum (an oily substance produced by the human skin). Scientists then used the information to build an algorithm to accurately predict the presence of PD, with an accuracy of 70 percent. When scientists applied ML to analyze the entire odor samples, the accuracy jumped to 79 percent. However, scientists acknowledge that more studies with an extensive and varied sample size need to be conducted.

    Meanwhile, during the height of the COVID-19 pandemic, research on data collected by wearables, such as Fitbit, Apple Watch, and Samsung Galaxy smartwatch, showed that these devices could potentially detect viral infection. Since these devices can collect heart and oxygen data, sleep patterns, and activity levels, they could warn users of potential diseases. 

    In particular, Mount Sinai Hospital analyzed the Apple Watch data from 500 patients and discovered that those infected by the COVID-19 pandemic displayed changes in their heart variability rate. Researchers are hoping that this discovery can lead to the use of wearables to create an early detection system for other viruses like influenza and the flu. A warning system can also be designed to detect infection hotspots for future viruses, where health departments can intervene before these diseases develop into full-blown pandemics.

    Implications of illness-detecting sensors

    Wider implications of illness-detecting sensors may include: 

    • Insurance providers promoting illness-detecting sensors for patient healthcare information tracking. 
    • Consumers investing in AI-assisted sensors and devices that detect rare diseases and potential heart attacks and seizures.
    • Increasing business opportunities for wearable manufacturers to develop devices for real-time patient tracking.
    • Physicians focusing on consultancy efforts rather than diagnostics. For example, by increasing the use of illness-detecting sensors to assist in diagnosis, physicians can spend more time developing personalized treatment plans.
    • Research organizations, universities, and federal agencies collaborating to create devices and software to enhance diagnostics, patient care, and population-scale pandemic detection.

    Questions to comment on

    • If you own a wearable, how do you use it to track your health stats?
    • How else may illness-detecting sensors change the healthcare sector?

    Community forecast feedback

    View the community's ratings after you leave your own below.

    Average year

    All readers

    Average year

    Qr readers

    --

    Average vote

    All readers

    Average vote

    Qr readers

    --

    Average vote

    All readers

    Average vote

    Qr readers

    --

    Average vote

    All readers

    Average vote

    Quantumrun readers

    --
    --
    --

    Average vote

    Company readers

    Insight references

    The following popular and institutional links were referenced for this insight: