By Graham K. Rogers

Health monitoring using smart devices has grown much in the last few years. Diabetics use such devices, but the basic input comes using invasive methods of accessing blood. It is rumored that Apple is working on a non-invasive blood-glucose monitoring system that uses the Apple Watch; but the idea that this uses blue light has some confused.

Introduction

I am grateful to my colleagues at the Faculty of Engineering, Mahidol University, who have guided me through my examination of the questions here, particularly Dr Phornphop Naiyanetr (Bio-medical engineering), Dr. Aranee Panganard (Electrical Engineering) and Dr. Surachoke Thanapitak.

Blood-glucose monitoring

When the Apple Watch first came into my hands, I was fascinated with its ability to check my heart rate using green light. We were told by Apple that green was chosen because it contrasts with the red of blood and so makes detection easy. It turns out that not only is the heart rate detection fairly accurate, but that data from the watch can be used as a way to discover certain heart problems, such as arrhythmia. A recent survey conducted by the developer of the Cardiogram app using Apple's ResearchKit, found a high level of accuracy in making predictions.

A survey conducted by Stanford and published late last month found that the Apple Watch achieved the lowest overall error in both heart rate and energy expenditure, but was less accurate with exercise time and calorie expenditure (Journal of Personalized Medicine), so there is room for improvement in these areas.

As a device for measuring normal personal inputs it seems to be effective, along with FitBit and Microsoft Band, but its medical value with regard to the heart data is more valuable. Other types of input, such as weight measurement, blood pressure or temperature need third-party hardware, usually attached to a computer or smartphone. Currently most measurement of blood-glucose levels, critical for the health of the millions of people worldwide suffering from diabetes, not only use such external devices but almost all are invasive, and unpleasantly so.

As part of an earlier look at measurement devices I acquired a GMate blood monitor that works with the iPhone, but not directly with Apple's HealthKit. Every morning for a week, I pierced my finger with a lancet and placed a drop of blood on a test strip. This was inserted into a monitor which connected to the iPhone through the 3.5mm headphone jack (this works with the latest iPhones and Apple's Lightning to 3.5mm connector). An app would then calculate the levels of glucose in my blood.

Earlier at WWDC 2016, I had seen the Dexcom solution outlined during a workshop at the Moscone Center, and this appears similar to another invasive solution by Medtronic. By the continuous monitoring that these systems provide, a user can have a more complete idea of glucose levels and be provided with predictive analysis using software.

The system provides a certain amount of freedom to users which they would not otherwise have with frequent home monitoring - those with Type 1 diabetes need to monitor far more frequently than my once-a-day trial - but they are stuck with the inconvenient and unaesthetic solution of a probe inserted under the skin. As was pointed out to me when I asked about this, the alternatives are much worse.

There has been a certain amount of research into systems that are non-invasive, with the use of the invisible spectrum (infrared and radio), but these require a certain level of complexity (and expense) to provide accurate and user-friendly monitoring.

Non-invasive approaches

Both of these potential solutions use transmissions at (or past) the red end of the spectrum, while the blue light solution that Apple is rumoured to be developing is at the opposite end. Further along the spectrum from blue is ultra-violet which would be totally unsuitable for safety reasons. However, blue light has a wavelength slightly higher than green, which the Apple Watch already uses for measuring heart rate. Roughly, blue light is above 500nM and green light below this level.

In an article examining different laser light levels in glucose monitoring, H. Ali, F. Bensaali and F. Jaber write that "The highest absorbance is of NIR of wavelengths from 700 nm to 1000 nm making it the least suitable for BGM [blood glucose monitoring] applications", adding that 650 nm is an ideal choice for BGM: red light. Note that, by starting at 500nm, blue light was not considered at all.

The end purpose that Apple has under development is not known. It could be a monitor to detect changes in blood chemistry after exercise and eating, or have a semi-diagnostic intention that would help diabetics to monitor blood-glucose levels, in which case FDA approval would be needed. If the device were for the latter purpose, it would be in Apple's interests for it to be developed by a 3rd party as an accessory for the Watch (or, less likely, the iPhone).

A third-party accessory or other hardware solution would not need Apple to go through FDA processes which would delay any updates to the Watch. Apple has already inferred that it does not want its devices to be shackled by such processes. If the purpose were for performance monitoring, such FDA approval would not be needed. The device, whether a strap, an accessory, or incorporated into the LEDs of the Apple Watch, would be easier to make and use. It has been known for a while about the potential of OLED sensors and Patently Apple mentions this in relation to some of the experts that Apple has recruited.

Comments

Some of the answers depend on the final use or intention of such a monitoring device, and on its developer (whether Apple or 3rd party). With the World Wide Developers Conference beginning on Monday (5 June) in San Jose, answers may come then. Using certain colours of light has been shown to have potential for blood-glucose monitoring. Until the information is confirmed (or otherwise), on the question of blue light I am left scratching my head.

See also:

Anna Shcherbina et al. Accuracy in Wrist-Worn, Sensor-Based Measurements of Heart Rate and Energy Expenditure in a Diverse Cohort. Journal of Personalized Medicine.

Apple Drops Some Apple Watch Health Sensors ahead of Launch. Patently Apple.

H. Ali, F. Bensaali and F. Jaber. Novel Approach to Non-Invasive Blood Glucose Monitoring based on Transmittance and Refraction of Visible Laser Light. IEEE Transactions on Magnetics. [No URL at this time]

New Technology Surfaces that Could Assist Apple's Wearable Computer Project Patently Apple.

Hui-Chen Wanga, An-Rong Lee. Recent developments in blood glucose sensors. Journal of Food and Drug Analysis.

Masab Ahmad et al. Non-invasive blood glucose monitoring using near-infrared spectroscopy. EDN Network.

David C. Klonoff. Overview of Fluorescence Glucose Sensing: A Technology with a Bright Future. Journal of Diabetes Science and Technology

Graham K. Rogers teaches at the Faculty of Engineering, Mahidol University in Thailand. He wrote in the Bangkok Post, Database supplement on IT subjects. For the last seven years of Database he wrote a column on Apple and Macs. After 3 years writing a column in the Life supplement, he is now no longer associated with the Bangkok Post. He can be followed on Twitter (@extensions_th)