[puszczyk]

The in-ear HR is interesting. I wonder if Apple can make it accurate, e.g., Sennheiser Momentum are not very accurate[0]:

> Look, I’m not gonna waste your time – this thing is dumpster-fire inaccurate level in almost every realm of heart rate accuracy except for indoor cycling.

[0]: https://www.dcrainmaker.com/2024/05/sennheiser-momentum-temp...

[intrasight]

The hardware tech in place for HR is even more interesting.

"AirPods Pro 3 introduce a custom photoplethysmography (PPG) sensor that shines invisible infrared light pulsed at 256 times per second to measure light absorption in blood flow. Combined with sensor fusion from the AirPods Pro accelerometers, gyroscope, GPS, and a new on-device AI model on iPhone ..."

This same hardware, or its next iteration, will probably allow for silent speech recognition with a software update. This will be a total game-changer in terms of human-computer interactions.

Sadly, it'll be up to Apple to make this a reality since they don't expose the raw sensor data.

[coldtea]

Apple wouldn't put in a medical sensor if it wasn't accurate, that's not their style and they know they'll get the most scrutiny (whereas nobody specifically has it in for Sennheiser, they're just one more brand among 100s).

If it wasn't yet up to par it would make it just give "warnings" instead of numbers (like the apnea and blood pressure measurements).

[adastra22]

This is anicdata, but my watch is consistently 5-10bp higher than my clinically measured heart rate. Always and consistently. It's an actual issue for my health care that my doctor and I took some time to track down, and turns out Apple Watch is wrong.

[Foivos]

They still include the blood oxygen sensor in their watches, which is impossible to be accurate because it is doing measurements at the wrist.

[coldtea]

And yet:

The researchers had 24 healthy participants wear an Apple Watch Series 6—which is outfitted with the same blood oxygen reader as the newer Apple Watches—on their left wrist while placing the medical-grade reader on their left middle finger. The participants wore breathers that slightly reduced the oxygen they took in over a few different phases, and the researchers recorded the blood oxygen levels recorded by each device at 30-second intervals. The Apple Watch is very close to achieving accuracy levels of a pulse oximeter. The blood oxygen level (which is called SpO2 in the study) bias across all data points was 0 percent. Here’s the rest: “The bias for SpO2 less than 90% was 1.2%. The differences in individual measurements between the smartwatch and oximeter within 6% SpO2 can be expected for SpO2 readings 90%-100% and up to 8% for SpO2 readings less than 90%.”