Apple Watch is most precise wrist-worn heart rate tracker at 90% accuracy, study says

Posted:
in Apple Watch
It appears Apple's bet to pack a bespoke heart rate sensor into Apple Watch instead of off-the-shelf technology paid off, as Cleveland Clinic researchers recently found the wearable to be the most accurate wrist-worn fitness tracker on the market.









The study, which pitted Apple Watch against competitors Fitbit Charge HR, Mio Alpha and Basis Peak, found significant discrepancies between the variance between the consumer devices, reports Time.



Testing involved 50 subjects who were hooked up to an electrocardiogram while walking, running and at rest. Results were compared to to heart rate data from the consumer devices, with Apple Watch showing a 90 percent accuracy rate in most scenarios. The others dropped into the "low 80s," according to Dr. Gordon Blackburn, director of cardiac rehabilitation at Cleveland Clinic.



An unnamed consumer level chest strap monitor, presumably similar to those sold by Polar, was also tested and found to be 99 percent accurate.



"What we really noticed was all of the devices did not a bad job at rest for being accurate for their heart rate, but as the activity intensity went up, we saw more and more variability," Blackburn said. "At the higher levels of activity, some of the wrist technology was not accurate at all."



It is unclear whether Apple Watch accuracy diminished during rigorous exercise, and to what extent, though it is common for wrist-mounted monitors to suffer signal degradation during extreme movement. Since wearables rely on optical sensors to measure blood flow at a single point on the body, accurate tracking becomes a problem when the device shifts or lifts off from the skin.



Apple detailed how its in-house heart rate sensor prevents data anomalies caused by user motion in a recently published patent application. Apple's heart rate sensor is based on existing photoplethysmogram (PPG) technology, which employs a light emitter and sensor array to measure blood perfusion to the skin. With Apple Watch, signal data from at least two light guides are compared and contrasted using special software algorithms to correctly compensate for physiological changes (vasculature expansion and contraction) and device motion.



Even with a highly advanced sensor solution, Apple Watch is still positioned as a consumer device. Rumor has it that Apple is looking to push deeper into the health and medical industries with Apple Watch and its supporting ecosystem of apps. A first step might be providing healthcare professionals with rich data gleaned by HealthKit, a move hinted at by Apple's purchase of digital health records firm Gliimpse.



Most recently, a report in September claimed Apple is working on two new apps for Apple Watch, one for accurately tracking sleep patterns and another that measures heart rate recovery.
«1

Comments

  • Reply 1 of 26
    I've noticed my Fitbit Charge HR doesn't seem very accurate on my longer runs (30KM's +). There are times when I know my heart rate is increasing as I'm running up hill and yet it shows no change or sometimes even a decline. But as the article says they are consumer grade devices and not expected to achieve 100% accuracy. Looking forward to having an AppleWatch one day so I can compare the results of it with my fitbit. Low 80's for accuracy compared to over 90% accuracy for the Apple Watch is a big difference.
    caliwatto_cobraGeorgeBMacalbegarcjony0jbdragon
  • Reply 2 of 26
    larryalarrya Posts: 587member
    Curious - how does this study declare "best on market" when only 3 other brands are compared?  No Garmin?  No  TomTom spark? No Suunto?  No modern Fitbit, say something that was released after 2014?  Can we have something better than hollow victories?
    freshmakersingularitygatorguywiggin
  • Reply 3 of 26
    nolamacguynolamacguy Posts: 4,758member
    it doesn't make much sense to use a british english word in the lead of your story for an american site. and yes while there are U.K. readers here, it still doesn't make style guide sense to do so. 
    pscooter63
  • Reply 4 of 26
    larryalarrya Posts: 587member
    I'm a dumb 'murcan and I know what "bespoke" means.  Maybe it was too many seasons of Top Gear?
    lollivercanadiandudesteveh
  • Reply 5 of 26
    it doesn't make much sense to use a british english word in the lead of your story for an american site. and yes while there are U.K. readers here, it still doesn't make style guide sense to do so. 
    I'm a British reader, and former publishing editor, residing in France. However, whilst few house style guides nowadays would recommend inserting periods ('full points') for the acronym UK, they would use capital letters for British, English and American... not to mention beginning sentences with It, and And. ;-)
    edited October 2016 robmfreshmakerjay-tlostkiwialexmacjony0lolliver
  • Reply 6 of 26
    larrya said:
    Curious - how does this study declare "best on market" when only 3 other brands are compared?  No Garmin?  No  TomTom spark? No Suunto?  No modern Fitbit, say something that was released after 2014?  Can we have something better than hollow victories?
    You realize that it takes a significant amount of time to perform a study, do the statistics, write the paper, then find a journal to get it published, then the journal has to do peer review on the article before it is published. It's not some dude in a garage running on his home treadmill and putting the numbers into his trial version of Microsoft Excel, then uploading it to his personal blog.
    edited October 2016 canadiandudejay-talbegarclolliverpscooter63
  • Reply 7 of 26
    What's always amazing me is how Apple rarely uses specs including comparisons to competition based on third party reviews in their ads. Others run with whatever spec they want to taut, , but Apple prefers the understatement here and focuses on user experience. Agreed, a lot of people would tend to believe Marketing independent on the presence or absence of facts behind their statements. OTOH spec-whoring is as useless on the other side of the spectrum. 
    What is amazing me is that to reach the goal of selling as many products and services possible they seems to still be convinced that a good product and hard work behind is the best guarantee for success when you go for margin and not just market share. You get a glimpse of the crazy hard work behind all this seemingly easy tech when the CEO laudes the team during a keynote. 
    edited October 2016 watto_cobrajbdragonstevehcalibadmonklolliver
  • Reply 9 of 26
    ppietrappietra Posts: 283member
    larrya said:
    Curious - how does this study declare "best on market" when only 3 other brands are compared?  No Garmin?  No  TomTom spark? No Suunto?  No modern Fitbit, say something that was released after 2014?  Can we have something better than hollow victories?
    The study doesn’t declare that. The study was only interested in seeing how reliable wrist-worn trackers can be, it doesn’t make any conclusion about which one is the best on the market.
    GeorgeBMaccali
  • Reply 10 of 26
    I have used the original Apple Watch for the past year and the accuracy of the heart rate tracker degrades significantly with intense exercises. If I work out with series of exercises, as the heart rate goes up rapidly the tracker keeps showing very low results. It still not accurate enough to be a fitness tracker.
    jdwcali
  • Reply 11 of 26
    it doesn't make much sense to use a british english word in the lead of your story for an american site. and yes while there are U.K. readers here, it still doesn't make style guide sense to do so. 
    It sounds to me like another red neck who doesn't know where is U.K. in a map...
    lolliver
  • Reply 12 of 26
    ppietra said:
    Interesting. Always pays to read the original. Based on that one page summary, I think this AI article might be wrong. The Mio Fuse had almost the same results as the Apple Watch so the sentence "The others dropped into the "low 80s," seems to only refer to the Fitbit and Basis
  • Reply 13 of 26
    lkrupplkrupp Posts: 9,991member
    Ever notice when any positive news is posted about Apple it’s always followed by people who come specifically to crap on it? It’s like they can’t stand to read anything nice about Apple.
    edited October 2016 GeorgeBMacalbegarclolliverpscooter63
  • Reply 14 of 26
    wigginwiggin Posts: 2,265member
    ppietra said:
    Interesting. Always pays to read the original. Based on that one page summary, I think this AI article might be wrong. The Mio Fuse had almost the same results as the Apple Watch so the sentence "The others dropped into the "low 80s," seems to only refer to the Fitbit and Basis
    I was wondering how the AI article managed to lump the Mio in with the other low-end watches as that is usually considered one of the better consumer optical HR monitors. Nice bit of selective journalism there, AI. LOL  

    It would be interesting to see a more comprehensive study with more devices, including more devices that are in the same higher-end category as the Apple watch (Garmin, Suunto, etc). But as the original research letter (ie, not a full-blown study) said, this was just a "convenience sample" (probably the devices the researches already personally owned LOL) of devices and "results should be confirmed with different types of exercises and with other devices."
    cali
  • Reply 15 of 26
    wigginwiggin Posts: 2,265member
    lkrupp said:
    Ever notice when any positive news is posted about Apple it’s always followed by people who come specifically to crap on it? It’s like they can’t stand to read anything nice about Apple.
    Without a doubt, the Apple watch is in the top tier of fitness watches. I think what some people "crap on" is the poor reporting by AI, which then leads to over-compensation in the follow-on comments due to the clearly demonstrated bias AI has for all things Apple. Many years ago AI was one of the best places to get all kinds of great info about Apple and the technologies in their products, including critical analysis of the competition. There is still an occasional gem of information to be found here (which makes me keep coming back), but it's mostly click-bait rubbish these days. Feeding meat to the base. This article is a prime example.
    singularitypscooter63
  • Reply 16 of 26
    GeorgeBMacGeorgeBMac Posts: 11,421member
    Interesting!   I may be one of the people who prompted that study!   For the past 3 1/2 years I've been reporting heart rate to Dr. Blackburn's staff and, for the past 18 months showing them wild swings of up to 50BPM during vigorous exercise (running) which perplexed them.   I was using a Polar H7 strap and and the ICardio app on my IPhone and they ran a bunch of tests on me but could neither find anything wrong nor duplicate the problem with their own equipment -- so they blamed it on the consumer grade equipment I was using.   (Actually, it turned out to be mostly caused by overtraining/overreach because, after a few days of rest it didn't happen -- which is why it didn't show up on their test because I always rested before their tests).

    But interestingly, recently I switched from the Polar H7 being monitored by the ICardio app on my IPhone (aka "DigiFit") to an Apple Watch and the watch always shows a steady, consistent heart rate even if I have linked it to the Polar H7 heart rate strap.

    The difference, I believe, lies in sampling rate:  while the ICardio app on the IPhone would sample many times a second and then display a graph of my heart rate where the swings showed up, the exercise app on the Apple Watch samples far less frequently (in order to conserve battery) and then only displays average heart rate.   Basically, the Apple Watch tends to miss the spikes -- which is (I believe) proven because when I use both the watch (using its wrist sensors) and the strap connected to ICardio on my IPhone simultaneously, the average heart rate on the IPhone is often 5-10 beats higher than the what the watch shows (it has never been lower).

    This may also be why Apple has admitted that the heart rate measured by the watch is not overly accurate when it is monitoring exercise such as weight lifting or boxing -- it misses the spikes.

    From my experience, if you want an accurate heart rate, you still need to use a chest strap and IPhone.   And the study confirmed that saying that the chest strap is 99% accurate.   Part of that is due to methodology:   the chest strap measures the same electrical signals as an EKG which is the gold standard; while the wrist monitors measure blood flow through the wrist by shining green light through the skin.  It sounds to me like the wrist based light sensors are simply inherently less accurate.   But, what is also interesting is that it is theoretically possible to measure those same electrical signals that the EKG measures at the wrist rather than the chest.   I wonder if Apple looked into that or will look into that?

    In the meantime, I know that if I want a highly accurate heart rate I need to use my Polar H7 strap and my IPhone.  But, for most runs, just the Apple Watch will give me 'close enough'.   
    brucemccali
  • Reply 17 of 26
    ppietrappietra Posts: 283member
    wiggin said:
    ppietra said:
    Interesting. Always pays to read the original. Based on that one page summary, I think this AI article might be wrong. The Mio Fuse had almost the same results as the Apple Watch so the sentence "The others dropped into the "low 80s," seems to only refer to the Fitbit and Basis
    I was wondering how the AI article managed to lump the Mio in with the other low-end watches as that is usually considered one of the better consumer optical HR monitors. Nice bit of selective journalism there, AI. LOL  

    It would be interesting to see a more comprehensive study with more devices, including more devices that are in the same higher-end category as the Apple watch (Garmin, Suunto, etc). But as the original research letter (ie, not a full-blown study) said, this was just a "convenience sample" (probably the devices the researches already personally owned LOL) of devices and "results should be confirmed with different types of exercises and with other devices."
    I think you should make that comment to Time magazine, which made that allegation, not Appleinsider.
    lolliver
  • Reply 18 of 26
    Attention! The market space also include Garmin, and I am not seeing any data with them, pitted against the Apple Watch.  Other are also excluded, like those from Samsung, and Jawbone. So, not a complete picture of the situation.

    Just so, you know, I do not own an Apple Watch. For the kind of activities I do, mainly the gym, and hiking, I chose Garmin products, the Forerunner 15, and a good GPS, the eTrex Touch 35.  I need to have a GPS track for at least 8-10 hours.  The Apple Watch is not capable of doing that on a single charge
  • Reply 19 of 26
    nolamacguynolamacguy Posts: 4,758member
    it doesn't make much sense to use a british english word in the lead of your story for an american site. and yes while there are U.K. readers here, it still doesn't make style guide sense to do so. 
    I'm a British reader, and former publishing editor, residing in France. However, whilst few house style guides nowadays would recommend inserting periods ('full points') for the acronym UK, they would use capital letters for British, English and American... not to mention beginning sentences with It, and And. ;-)
    blame the iOS keyboard, it swapped in the periods.

    as for my case usage, it's irrelevant -- I'm just an anonymous commenter on a rumors site, not a publication that presumably has a style guide. guides have value for a reason. thats why the "But you do it too!" argument is a fallacy, an ad homenim, because it's attacking the speaker and not the point raised.


    edited October 2016
  • Reply 20 of 26
    nolamacguynolamacguy Posts: 4,758member
    alexmac said:
    it doesn't make much sense to use a british english word in the lead of your story for an american site. and yes while there are U.K. readers here, it still doesn't make style guide sense to do so. 
    It sounds to me like another red neck who doesn't know where is U.K. in a map...
    which has nothing to do with my point. I'm willing to bet most american readers have no idea what "bespoke" means.
    edited October 2016
Sign In or Register to comment.