By Kevin Uchimura
Kevin Uchimura has been working with Project HoneyBee to help develop a database of biosensors developed for the consumer-market. He completed his Masters in ASU’s Biomedical Engineering program in May 2013. He helps to manage a laboratory group developing the next generation of biosensors and his interests include new gadgets and creatively applying them to medicine.
Kevin has provided Project HoneyBee with invaluable technical expertise and we’re thrilled to present his key findings and musings in monthly installments.
The first two articles are devoted to exploring the challenges behind benchmarking different devices and why it’s so important to the data validation process. Since validating wearable technology in clinical settings to help pinpoint the transition from disease to health is central to Project HoneyBee’s efforts, Kevin’s work in benchmarking consumer biosensors is crucial to our success.
Why are we Benchmarking?
The field of wearable biosensors is a rapidly evolving market; manufacturers are scrambling to leverage new technologies in communication, connectivity, and miniaturized electronics to apply to traditional medical devices. The dynamics of this emerging field of consumer-marketed devices are complex: manufacturers are trying to offer more complex health metrics to satisfy the demands of the “wellness” movement, while simultaneously avoiding FDA purview.
The wellness movement is characterized by an interest in one’s health and fitness, especially through the use of devices which had traditionally been confined to the clinic or the research laboratory. Manufacturers are following a general pattern of introducing medical technology to a population increasingly concerned for their own health. The humble pedometer has evolved into the FitBit, a high-tech activity tracker capable of detecting stair-climbing, sleep quality, and other information. The Scanadu Scout, the 1.6 million dollar darling of crowdfunders, is a melting pot of health metrics, collecting the electrocardiogram, body temperature, blood pressure, and the photoplethysmogram.
For an engineer, a device’s “robustness” refers to the veracity of the produced data under all operating conditions. Biosensors vary tremendously in this respect. Deciphering adequate working conditions for each device is critical to testing whether those biosensors are collecting data accurately.
Benchmarking these new devices is a challenging prospect, primarily because of the lack of any standard method of evaluation.
Consumer devices exist on a sort of spectrum: on one side, those devices which aspire to diagnostic use, and on the other, those devices which are primarily wellness devices. The AliveCor portable single-channel ECG and the Scanadu Scout are consumer devices actively seeking FDA approval. These manufacturers recognize the value of marketing a product as a bona fide medical device. However, other manufacturers that are unwilling to undergo onerous and expensive FDA regulations simply specify that their product is not intended for diagnostic purposes and may still apply nebulous promises of improved health.
While FDA approval is not a guarantee of device quality, it does provide the discerning physician or biomedical engineer some assurance that a product is not likely to burst into flames or suffer some other, less dramatic, design defect. Furthermore, companies that pursue validation are far more likely to publish detailed design specifications or results from a clinical trial.
Devices that do not aspire to FDA regulation are under no obligation to follow established technical standards that set requirements on how accurate a device must be within certain conditions. Interestingly, the FDA does not require that device manufacturers adhere to relevant standards, but such renegades must generally offer a very compelling reason for noncompliance. Therefore, understanding whether a certain data set is accurate for its working condition is made that much more difficult. In the next post, I will continue the discussion around the complexities inherent in benchmarking, and explore the shortcomings involved in not ensuring interoperability between devices.
The post Benchmarking Biosensors: The Necessity and the Challenges appeared first on Sustainable Health.