7:39 a.m. That’s the time that your smartphone’s sonar deems as optimal for you to wake up today. With its gentle vibration from your bedside table, you pick it up to turn off the smart alarm. As you do so, your phone asks for your permission to use the built-in sensors and camera to run your routine morning scan. It analyzes your voice; evaluates your stress level based on a facial scan; checks your vital signs; and notifies you to take a picture of that mole on your forearm in order to detect any anomalies.
Thereafter, it outputs a comprehensible report with recommendations which you can send over to your physician. This helps in early detection and prevention of ailments thanks to the minute changes indicative of a potential illness. All this is done in mere minutes, without even stepping out of your bed, every morning.
Such a (literally) healthy wake up isn’t too far away. In the near future your phone will pack sensors that will allow you to gauge your vitals and more in order to monitor ongoing treatments and prevent other diseases from progressing.
As a matter of fact, smartphones already have built-in sensors to analyse numerous parameters, without any add-ons. Cameras aren’t just for Instagram stories but can analyse skin lesions; voice recording isn’t only to ask Siri about the weather but can also detect Alzheimer’s; and you’re likely already familiar with the built-in fitness app in your phone. However, there are still some hurdles to overcome before we enjoy a morning screening as depicted.
Here we analyse the parameters that our smartphones can currently measure, the challenges preventing them from turning into an effective Swiss knife of digital health and give a glimpse of what awaits in the near future.
Every step counts
“Don’t forget to take your walk!” “It’s Day 3 of 5 of staying active!” These are notifications you’re probably familiar with thanks to built-in apps like Google Fit and Apple Health on your smartphone. By using your phone’s motion sensors (accelerometer, gyroscope, proximity sensor), they track your steps, estimate the calories you burnt and remind you to keep active regularly. The latter is particularly important as the WHO estimates that 1 in 4 adults is not active enough, which can increase the risk of cardiovascular complications and diabetes.
Additionally, these fitness apps can incorporate data about your sleep through smart alarm apps. Getting the right amount of sleep and waking up at the optimal time won’t only make you feel better but studies have also linked improper sleeping habits to suppressed immune systems and Alzheimer’s.
However, some experts cast a shadow of doubt over the reliability of energy expenditure and step count measurements. These show that our phones still have room for improvement on that front; but activity tracking is only the tip of the iceberg when it comes to what our phones are capable of measuring.
Eyeing hemoglobin count
In this near future, you won’t need to draw blood to evaluate your anemia risk. Rather, a simple picture of your inner eyelid, where microvasculature is visible, will do for your physician to evaluate remotely. This is exactly what a group of researchers have recently worked on.
By upscaling images of the inner eyelid taken by a smartphone’s camera, a set of algorithms can evaluate the blood haemoglobin content and help determine blood disorders. Early results show that this technique is comparable to traditional blood tests over a wide range of blood haemoglobin values.
“This new technology could be very useful for detecting anaemia, which is characterised by low levels of blood haemoglobin,” said the lead researcher. “This is a major public health problem in developing countries, but can also be caused by cancer and cancer treatments.”
Sensing oxygen saturation
You might, right now, be having a cough and sneeze every so often but not think much of it and go about with your routine. However, even without overt symptoms, there is the possibility of an underlying respiratory illness; even one that is related to COVID-19.
That’s why doctors are calling to have oxygen saturation (SpO2) measured for all at-risk patients. This parameter refers to the percentage of haemoglobin in your red blood cells bound to oxygen. A drop in the SpO2 value will indicate an underlying ailment even if no visible signs are present. How about monitoring Sp02 remotely, while in self-isolation, without risking cross-infection? This will soon be possible.
Some older Samsung smartphones featured pulse oximetry sensors that allowed SpO2 measurements and sharing of the results to a physician. Other apps claim to offer a similar function through a phone’s camera.
However, the Centre for Evidence-Based Medicine recently contested the accuracy of both Samsung’s sensors and apps. Researchers at the institution concluded that these methods raise “serious questions about the diagnostic accuracy”.
This is in part due to the inadequate wavelengths of light used in these sensors which can, however, be improved upon. Now that interest in this feature is gaining traction, we could see more accurate pulse oximetry sensors get integrated in our next smartphones.
Tap for temperature
With a simple tap to the forehead, your future phone will record your temperature. Traditionally, this apparently simple parameter was a challenge to measure with a phone. That’s in part due to the design aspect and also because the smartphone itself is a considerable source of heat (ever thought about frying an egg on the back of your phone after watching a couple of YouTube videos?). This makes recording ambient temperature quite the challenge, and the reason why Samsung dropped the idea.
However, with the prevalence of IR sensors in cameras, like the one used in the iPhone’s FaceID, temperature measurement is set to be easier and more commonplace with phones. For instance, dubbed as “the most 2020 phone of 2020”, Huawei’s latest phone, the Honor Play 4 Pro, integrates an IR temperature sensor at the back. With a tap to the forehead, the phone can establish the user’s temperature.
With phones making temperature measurements as easy as a tap, patients and physicians will monitor this parameter remotely, which, if elevated, can be indicative of the need to self-isolate.
Making diagnosis of rare genetic conditions more common
Our upcoming high-tech Swiss army knife of digital health will not only be a boon for the general public but also for physicians. Yearly, some half a million children are born with a rare hereditary disease worldwide. Their identification poses a challenge since doctors might not have encountered such rare cases in their career. However, early detection is crucial to provide adequate treatment and support.
With this in mind, scientists at the Kaunas University of Technology (KTU) in Lithuania developed an app to identify Huntington’s disease, a genetic condition leading to physical and cognitive impairments. It can manifest in both adults and children, with the latter being rarer and harder to diagnose. KTU’s app aims to diagnose the condition in at-risk patients when there are no visual symptoms.
It achieves this with a set of tests that evaluates the patients’ physical and cognitive skills and early signs of decline. “Although there is no known treatment for Huntington’s disease, it is estimated that a patient can gain 3-16 years of healthy life if the disease is diagnosed early,” said KTU Professor Rytis Maskeliunas.
Other researchers in Germany trained an A.I. to identify rare conditions like Mabry syndrome and Kabuki syndrome, which have specific facial characteristics, from portrait pictures and genetic data. Their method improved the accuracy of diagnosing these rare conditions.
Pocket aspiring dermatologist
In our introductory near-future scenario, we mentioned the ability to check one’s skin lesions through a phone. This is already possible today thanks to dermatology apps. With a simple snapshot of a mole, apps like SkinVision can determine the risk of your skin lesion. Upon the evaluation of a suspicious mole by the algorithm, the picture is further analysed by actual dermatologists, who give further recommendations. Moreover, the app sends reminders to regularly re-evaluate a mole after a period of time.
As innocuous as they might seem, skin lesions warrant careful attention as they can evolve into skin cancers. The WHO estimates that yearly, there are 2-3 million non-melanoma skin cancers and around 132,000 melanoma skin cancer cases. The latter is particularly serious due to its ability to metastasise to other organs. In the U.S., nearly 20 Americans die from melanoma every day, according to the American Academy of Dermatology. With the ease of early detection through a phone, there can be a drastic drop in such cases.
Decoding speech for neurodegenerative cues
In line with more specific diagnoses, doctors will be able to detect complex neurodegenerative conditions with a smartphone. Canadian firm WinterLight Labs developed an A.I. that analyses subtle cues from a patient’s voice. Such an analysis allows for the detection of conditions like Alzheimer’s, with an 82% accuracy.
“Our platform can analyse natural speech to detect and monitor dementia, aphasia, and various cognitive conditions,” the company said. “Using a short one-minute sample of speech, WinterLight can characterise the speaker’s cognitive, acoustic and linguistic state, including lexical diversity, syntactic complexity, semantic content, and articulation”. While it is implemented in a robot companion, the technology can easily be adapted on a smartphone.
For instance, KTU’s app mentioned in the previous section can also evaluate other neurodegenerative diseases such as Parkinson’s, Alzheimer’s or dementia. With such developments, it’s not difficult to imagine a doctor sliding out their phones during an appointment to enhance their diagnosis.
A regulated digital health Swiss knife
More than easing the lives of citizens of developed countries, low- and middle-income ones will largely benefit from a Swiss knife of digital health. In these resource-limited settings, having access to accurate and vital parameters measurements will be a real game-changer. Without the need for pricey equipment or any attachments, caregivers will perform quality, low-cost diagnoses with a single device.
Indeed, a current smartphone’s sensors can measure a plethora of parameters already. A review by scientists from McMaster University in Canada attested to their “incredible role” as low-cost solutions for early diagnosis and remote monitoring.
However, the authors also highlight the need for “rigorous clinical trials” to determine their safety and efficacy. They call for more defined guidelines by regulatory authorities before labelling an app as a “medical device”. Of course, there is also the ever-present matter of stricter controls over data privacy as these are often not addressed in studies.
We also saw how experts contest the reliability of smartphone’s pulse oximeters, calorie expenditure estimates and even step counts. Thankfully, these are features that can be remedied. With increased demand, companies will enhance their accuracy and reliability. With proper regulations and secure implementations happening in parallel, we will benefit from a real Swiss knife of digital health; and wake up to a similar routine as pictured in this article’s opening.