Validation of Amazon Halo Movement: a smartphone camera-based assessment of movement health

Movement health, encompassing the ability to perform daily living movements, is crucial for reducing injury risk, improving athletic performance, and enhancing overall well-being. While its benefits are well-established, quantitative assessment has been lacking. Current assessment tools, such as the Functional Movement Screen (FMS) and others, are often non-standardized, require specialized equipment, and clinician expertise, limiting accessibility. The proliferation of smartphone technology with advanced camera capabilities and AI algorithms creates an opportunity for innovative digital healthcare solutions. This study validates Halo Movement, a novel smartphone application leveraging computer vision and machine learning to assess movement health based on three key criteria: mobility, stability, and posture. Participants perform five pre-defined exercises, and the application analyzes the video recordings to generate a movement health score (0-100). This study aims to assess the validity of Halo Movement by comparing its scores to clinically validated functional movement tests and 3D motion capture data.

The existing literature highlights the significance of movement health and the lack of a universally accepted gold standard for its assessment. Several methods, including the FMS, LESS, and various range-of-motion tests, exist, but many are non-standardized, require specialized equipment and trained professionals, limiting widespread accessibility. Recent advances in computer vision and AI, combined with the ubiquitous nature of smartphones, present opportunities for developing user-friendly, accessible movement health assessment tools. Several studies have explored the use of markerless pose estimation algorithms from smartphone videos for fitness applications and clinical settings, demonstrating the potential of such technology for movement analysis.

This study employed a cross-sectional design with 150 participants aged 18-85, categorized into three groups based on self-reported activity levels and clinical diagnoses: athletes, healthy individuals, and movement-impaired individuals. Participants completed three trials of the Halo Movement assessment using their smartphones, followed by thirteen clinically validated functional movement tests, including range of motion measurements, balance tests (Sharpened Romberg, Star Excursion Balance Test, Clinical Test of Sensory Interaction on Balance), and strength/mobility tests (sit and reach, single leg hop). A full-body Xsens MVN Awinda motion capture system provided 3D joint angle and body segment position data during the Halo Movement assessment to extract metrics quantifying mobility, stability, and posture. Correlations between Halo Movement scores, functional movement test scores, and sensor metrics were analyzed using Pearson's or Spearman's correlation coefficients. Differences between the three participant groups were analyzed using two-sided t-tests. Intrasubject coefficient of variation was calculated to assess the repeatability of Halo Movement.

The study included 150 participants (Table 1), with a diverse representation of age, gender, race, and movement ability. Average Halo Movement scores were significantly different across the three groups: athletes (86.64 ± 5.69), healthy (81.28 ± 6.73), and movement-impaired (68.43 ± 6.74) (Fig. 3). Halo Movement showed moderate to strong statistically significant correlations (0.29 < *r* < 0.63; *p* < 0.05) with all thirteen functional movement tests (Fig. 4), except for one balance test (CTSIB (t)) likely due to a low ceiling effect. Statistically significant correlations (0.23 < *r* < 0.83; *p* < 0.05) were also found between Halo Movement scores and sensor-based metrics for mobility, stability, and posture (Fig. 5). The intrasubject coefficient of variation for Halo Movement was 1.64 ± 1.42%, indicating good test-retest reliability. The athlete and healthy groups did not differ significantly on most functional movement tests, suggesting Halo Movement might offer higher resolution for assessing functional movement in able-bodied individuals than existing tools.

This study provides evidence for the validity of Halo Movement as a smartphone-based assessment of movement health. The strong correlations with established functional movement tests and sensor-based metrics support its ability to accurately quantify movement health across a diverse population. The ability of Halo Movement to distinguish between athletes, healthy individuals, and movement-impaired individuals suggests its potential as a more accessible and comprehensive assessment tool. The higher resolution provided by Halo Movement compared to traditional functional movement tests, particularly in differentiating subtle movement discrepancies in healthy individuals, is a notable advantage.

Halo Movement offers a valid, cost-effective, and accessible alternative to existing functional movement screening tools. Its ability to quantify whole-body movement health using a smartphone makes it a valuable tool for individuals and healthcare providers. Future research should investigate the impact of targeted exercises on Halo Movement scores and their correlation with quality of life improvements. Further studies on broader populations, including children and individuals with severe movement impairments, are warranted.

This study's convenience sample may limit generalizability, particularly regarding the representation of individuals from diverse racial and ethnic backgrounds and those with severe movement impairments. While the chosen functional movement tests aimed for comprehensive body coverage, potential over- or under-representation of specific body parts or movement patterns could have influenced correlations. The study focused on validating the overall Halo Movement score, and further work is needed to validate the individual subscores for mobility, stability, and posture. The categorical grouping of participants (athlete, healthy, movement impaired) simplified a complex spectrum of movement abilities, requiring more nuanced analyses in future studies.