ABSTRACT
In order to continue providing users with the best in class activity monitoring experience, Basis, an Intel
Company, has progressively developed algorithms for heart rate and calorie detection. Hundreds of hours of in-
house testing and evaluation were invested in the development of these algorithms. To reinforce the reliability of the
algorithmic improvements, a validation study was performed at the Human Performance Research Lab at California
State Polytechnic University, Pomona (Cal-Poly Pomona), to assess the reliability of the Basis Peak.
The study aimed to validate the Basis Peak against a criterion measure for heart rate and calorie burn (also referred to
as energy expenditure). Additionally, the study intended to compare the reliability of the Basis Peak against a similar
activity tracker, the Fitbit Charge HR*. Both devices were subject to testing during which human subjects performed
an exercise protocol. Heart rate and expended energy values from both devices were respectively compared against
laboratory grade electrocardiograph (ECG) and indirect calorimetry.
Considerable agreement was observed between Basis Peak heart rate values and the ECG with an average difference
of 5.1%. Excellent correlation was also observed by the Basis Peak energy expenditure values and indirect calorimetry
with an average difference of 13.3%. The Fitbit Charge HR* had notably less comparable agreement with the ECG and
indirect calorimetry, with respective average differences of 8.7% and 27.3%.
INTRODUCTION
Commercially available activity trackers are in abundance, and their increasing prevalence indicates a growing
consumer market focus on personal health and well-being. Self-monitoring can provide invaluable information,
which in turn provides insights that can be applied towards weight regulation, athletic training, achievement of
healthier habits, or an overall general increase in awareness of one’s well being. There is a known association between
inactivity and an increase in risk for health problems such as obesity, cardiovascular disease, type 2 diabetes,
and mortality 1, 2, 3, 4
. Small increases in physical activity levels, when sustained, can have noticeable advantageous
impacts on individual health.
As activity trackers continue to become more prevalent, the need to validate their accuracy and reliability becomes
more important 6, 7, 8. In order to continue to enhance performance, Basis and Intel conduct continual assessment and
improvement of Peak’s sensors, algorithms, and associated technologies. In October 2014, Basis commissioned and
partnered with the Human Performance Center at the University of California, San Francisco Medical Center to validate
Peak and its algorithms. The results of the studies were conclusively positive and can be found here_______________.
In efforts to evaluate recent heart rate and energy expenditure algorithm improvements, Basis, an Intel Company,
commissioned a study at the Human Performance Research Lab at Cal-Poly Pomona. Unlike prior studies, resistance
exercises were included in this iteration’s protocol to increase the robustness of testing and better reflect a range of
real world usage. The primary aim of the study was to further validate the accuracy of Basis Peak’s heart rate and
calories monitoring with respect to benchmark measures acquired by electrocardiography and indirect calorimetry. A
secondary purpose was to compare the accuracies of the Basis Peak against that of the Fitbit Charge HR*.
METHODOLOGY
All experimental procedures were conducted on human volunteers in accordance with guidelines as
determined by the International Review Board of California State Polytechnic University, Pomona. The sample
pool consisted of 24 healthy subjects (12 males, and 12 females) between the ages of 20 and 31. Anthropometric
measurements were taken immediately prior to the start of data collection. Exclusion criteria consisted of any
significant medical diagnoses, including cardiovascular or pulmonary diseases, which may potentially increase the
risk of exercising or limit a subject’s ability to exercise.
All subjects underwent the same 77 minute protocol. Each participant completed an initial rest period (lying down)
of 15 minutes followed by 5 minute periods of each of the following activities: low cycling (60W resistance), intense
cycling (120W), walk (3.0-3.5mph speed), jog (4.0-5.0mph), run (5.5-7.0mph), arm raises with self-selected added
resistance (12 reps), lunges with self-selected added resistance (12 reps), and plank (60 second hold). In between
each exercise activity was a 5 minute rest period.
Each subject wore a Basis Peak on one wrist and a Fitbit Charge HR* on the opposite wrist. Wrist assignment (i.e. left
and right) alternated between subjects. Care was taken to follow proper wear guidelines as suggested for each device
9
. Criterion measurements of heart rate and energy expenditure were provided respectively by electrocardiography
(12-lead) and indirect calorimetry using a ParvoMedics TrueOne metabolic cart (met cart).
Heart Rate:
The Basis Peak demonstrated considerable
accuracy with an average difference of 5.1%
(see Table 1). Contrastingly, the Fitbit Charge
HR* demonstrated less favorable accuracy
with an average difference of 8.7%.
Exceptional correlation (R = 0.89) between
Basis Peak heart rate values and that of ECG
was seen (see Figure 2), while the observed
correlation for the Fitbit Charge HR* (R=0.80)
was notcieably lower (see Figure 3). Results
were statistically significant (p<0.001).
Table 1 | By activity, the average differences for heart rate from the Basis Peak and the
Fitbit Charge HR* as compared to ECG. The activity “All” refers to the entire trial period
of data collection.
Figure 3 | Comparison of all heart rate values from the Fitbit Charge HR* and the ECG
Figure 5 | Comparison of heart rate from the Fitbit Charge HR* and from the ECG for a female subject
Figure 7 | Comparison of heart rate values from the Fitbit Charge HR* and from the ECG for a male subject
Tests document performance of components on a particular test, in specific systems. Differences in hardware, software, or configuration will affect actual performance. Consult other
sources of information to evaluate performance as you consider your purchase. For more complete information about performance and benchmark results, visit http://www.intel.
com/performance
© 2015 Intel Corporation. All Rights Reserved. Intel and Basis Peak are trademarks of Intel Corporation in the United States and other countries.