mHealth Research Digest by Anupam Kumar


The angle of thorax rotation, measured in forward flexed position, is a common metric for assessing severity of scoliosis.

This rib hump angle is most often measured with a Scoliometer.

Prior reserach has confirmed that this device has good measurement reproducibility, with high inter-rater and intra-rater reliability.

However, these studies suffered from variability inherent in using human subjects–namely patient positioning, vertebral palpation, patient fatigue, and time passage between measurements were just some of the encountered sources of variability.

Recently produced smartphones, include the iPhone, include a micro-electric-mechanical-system (MEMS) accelerometer. The ability of this device to sense acceleration and inclination has been utilized by a number of native iOS applications produced by Apple itself and by developers for third-party apps.

The accelerometer has also been purposed towards clinical activities, especially those activities in need of precise angle measurements. In a recently published study, Izatt et al. compared the Scoliometer to the iPhone in order to compare relative effectiveness for measuring vertebral rotation in patients with scoliosis.


As noted above, previous investigations of Scoliometer efficacy have been saddled with variability introduced by using human subjects. The authors of this study sought to minimize these issues by employing professionally constructed plaster models of real scoliosis patients. Measurements with the iPhone were conducted using the Scoliguage application found in the Apple app store.

Additionally, the iPhone was placed in an acrylic sleeve that provided sufficient additional length to avoid underestimation of severe deformities.

Nine observers, at varying levels of clinical experience, measured the angle of rotation on eight different plaster torsos. The rib hump angles ranged between 6 and 30 degrees and aimed to be representative of severities encountered in clinical practice. At least one week passed between repeat measurements by observers, who were additionally blinded to their prior recordings. Intra- and inter-observer variability was subsequently assessed.


In total, 280 measurements were recorded by all observers. The overall rib hump angle measurement for the set of plastic moulds was 16 degrees +/- 5.8., similar to the range of severities seen in practice.

The mean absolute difference between pairs of measurements with the iPhone and scoliometer was 2.1 degrees with a 1 degree bias towards higher measurements with the smartphone. These figures are less than 5 degrees, the accepted standard for a clinical significant difference in rib hump angle.

Intra- and inter-observer variability was similar between the iPhone and Scoliometer. Consistent with previous research on the Scoliometer, the 95% confidence interval was higher for inter-observer than intra-observer variability. Compared to these earlier studies, the recorded confidence intervals and measurements errors were similar or improved.


Given the ubiquity of smartphones and tablet devices in the clinical setting and advancing accelerometer technology, their use in diagnosis of spinal deformities such as scoliosis is sure to increase exponentially. The iPhone in concert with the Scoliogauge application gives medical providers an easy, inexpensive and highly mobile avenue to conduct screenings in schools and in the community.  The authors conclude, based on the aforementioned data, that the iPhone is clinically equivalent to the standard Scoliometer in measuring rib hump angles.

Given the inadequate size of the iPhone alone for measuring severe spinal deformities without an the acrylic sleeve, it seems that longer devices like the Apple iPad or various other tablets may be better suited to fill this niche. However, a similar trial should be conducted to confirm the clinical utility of the iPad for this purpose before such a recommendation can be made.

The use of plaster torso models was a novel method for reduction of confounding variables encountered when using human subjects and allowed the authors to better isolate differences between the two measurement techniques.

This methodology should be considered in further exploration of truncal rotation measurement as newer methodologies are conceived over time. Extensive research is critical as new devices and software continue to be introduced into the medical setting at a blistering pace; for clinical decision-making to involve information provided by smartphone applications, such study is not only beneficial but also requisite in ensuring patient safety.

Authors: Maree T Izatt, Gary R Bateman, and Clayton J Adam
Institutions: Paediatric Spine Research Group, Queensland University of Technology and Mater Health Services,  Mater Children’s Hospital, Queensland, Australia
Original Abstract: PubMed