A novel use of inertial sensors to measure the craniocervical flexion range of motion associated to the craniocervical flexion test: an observational study

This is a methodological study that looked at the reliability of the assessment of craniocervical flexion ROM associated with the CCFT using a wearable inertial sensor. The study follows a reproducibility protocol in three stages, including training, overall agreement, and study phase [33]. The training and overall agreement stages have already been completed and constituted the basis of the methods used on the present observational study. The present study addresses the study phase of the reproducibility protocol. In the training and overall agreement periods, the examiners discussed and agreed on the definition of the performance of the CCFT and the use of the inertial wearable sensors described in this section.

The study was approved by the Research Ethics Committee of Ceu San Pablo University (236/17/08). Participants provided informed written consent before being enrolled in the study and were able to withdraw their consent at any time during the study, in compliance with the WHO standards and the Declaration of Helsinki [34].

The participants to be investigated in this study were asymptomatic adults from a university community. Subjects were recruited through various channels and methods in a university campus, including advertisement, word of mouth and emailing.

Participants were eligible to be included in the study if they were between 18 and 30 years old with no known conditions affecting the craniocervical region in the past year and were able to successfully perform the five stages of the CCFT (the detailed procedure is described in the section on instrumentation and measures).

Subjects were excluded if they presented with any of the following criteria: (a) history of pain in the craniocervical region and/or shoulders at the time of the measurement or during the previous year, (b) history of surgery in the head or neck area, (c) previous diagnosis of TMDs, headaches or (d) any neurological deficits.

A sample of asymptomatic participants were chosen, since it was considered is the most appropriate sample to accomplish the objectives of the present study. The descriptive results of craniocervical flexion ROM associated with the CCFT could represent the normative values to provide a standardization that could posteriorly serve as a reference point for patients with pain conditions in future studies.

Once deemed eligible, subjects were invited to participate in the study and were requested to read and sign the informed consent prior to participation.

Prior to testing, the demographic characteristics of the subjects, such as age, gender, weight, height and hours of physical activity/week, were recorded.

The CCFT was performed as described in the most recent clinical protocol by Jull et al. [24]. First, subjects were asked to perform the CCFT to assess whether they were able to successfully perform the five stages of the CCFT. If they were able to do it, they were included; thus, this procedure also served as a warm-up and a training for the performance of the CCFT before the real testing. Participants were placed in a relaxed supine position with the knees flexed, forearms resting on the abdomen and the neck in a neutral position with the face and the line bisecting the neck longitudinally being horizontal to the plinth. The examiner placed an uninflated pressure sensor (Pressure Biofeedback Unit, Stabilizer™ Chattanooga Group, Hixon, TN, USA) behind the neck of the patient and inflated it to a pressure of 20 mmHg. It was connected to a high accuracy digital pressure gauge (1/4ʺ NPT M, 5 PSI; DPG210; Omega Engineering Limited, UK) through a flexible plastic tube (Fig. 1). The connection of the pressure biofeedback unit to digital pressure transducers has been reported in previous studies [14, 20, 30].

Then, the LCD screen of the digital pressure gauge was turned to the subject, who was instructed to gently and slowly perform a nodding action to elevate the target pressure from 20 to 22 mmHg and to hold for 3 s before relaxing and returning to the initial position. During the movement, the examiner requested that the subjects slowly feel the back of their heads to slide on the bed. This process was repeated through each stage of 2 mmHg increments of the test until 30 mmHg.

During the test, the examiner provided visual and verbal cues to guide the process with a correct technique and palpated the muscle activity of the superficial cervical flexors (i.e., sternocleidomastoid and scalene muscles) to avoid their use. Other signs of compensation or poor activation, including neck retraction or a lack of progressive increment of the craniocervical flexion ROM necessary for the CCFT, were monitored by an independent assessor. Neck retraction was prevented based on the procedure described by Falla et al. [30]. The head of the patient rested on a pressure sensor consisting of a thin pressure sensitive textile mat (Sensing Tex^®, Pressure mat dev kit, Barcelona, Spain), which displayed the pressure values in real-time on a computer screen. Any increase in pressure greater than 0.75 kg was considered neck retraction compensation and implied repetition of the test or exclusion of the subject due to unsuccessful performance of the five stages of the CCFT. This method for measuring neck retraction as a compensation during the CCFT (i.e., increments in weight of 0.75 kg) has been described in previous research [30].

During all of this procedure, a small (4 cm × 4 cm × 8 cm), light (< 200 g) wireless inertial sensor [35] (Werium Solutions©, Madrid, Spain) was adhered to the centre of the forehead of the subject, defined as the place where the lines that bisect the forehead longitudinally and horizontally cross. This sensor provided real-time monitoring of the progressive increment of the craniocervical flexion ROM necessary for the CCFT, which was displayed on a computer. This inertial sensor technology has previously shown good excellent intra-rater and inter-rater reliability in the measurement of cervical ROM [35, 36].

Secondarily, the sensor system automatically recorded the cervical ROM achieved in each stage of the CCFT, represented as the degrees of motion of flexion from the starting neutral position of the neck, when the sensor was calibrated at 0.

The same independent assessor who monitored the pressure of the head against the pressure mat also monitored and recorded cervical ROM. Therefore, this assessor handled two computers connected by Bluetooth in real-time to the pressure and inertial sensors in order to detect whether any retraction compensation or insufficient craniocervical flexion occurred. Subjects who had any evidence of an incorrect execution of the CCFT during the initial screening procedure (i.e., warm up) were excluded from the study.

Figure 1 shows the placement of the pressure biofeedback unit, the inertial sensor and the pressure textile mat used in the measurement process.

The process of performing the CCFT described above was used at the screening stage to select the subjects for participation and also to evaluate the reliability of the craniocervical flexion ROM measurement associated with each stage of the CCFT. Examiners A or B randomly performed the CCFT at the initial selection stage. Then, the subjects included in the study attended the first session in which both examiners A and B performed the CCFT test once for each subject in a randomized order (Research Randomizer: https://​www.​randomizer.​org/​). Both examiners were mutually blinded to the results of the other examiner. This procedure was subsequently repeated by examiners A and B in a second session with an interval of 1 week, allowing for the calculation of the intra-rater and inter-rater reliability. Examiner C calibrated the inertial sensors and monitored and recorded the cervical ROM in all tests.

Immediately before the performance of the CCFT, examiners A and B also requested that subjects perform three full range active craniocervical flexion actions, which were also recorded by the inertial sensor. The measurement of the full ROM before the CCFT has been previously investigated using digital imaging techniques [30]. It allows for calculations of the percentages of the full ROM necessary to achieve each stage of the CCFT.

A flowchart of the whole testing protocol is shown in Fig. 2.

All data were analysed using the Statistical Package for Social Sciences software version 24.0 (SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606, USA). The normal distribution of quantitative data was assessed by the Kolmogorov–Smirnov test, and the mean and standard deviation (SD) for each normally distributed quantitative variable were analysed. The data corresponding to the ROM necessary to achieve each stage of the CCFT was expressed in degrees as means, SDs, 95% confidence intervals (CIs) and as a percentage of the full range of the craniocervical flexion ROM. The reliability of the craniocervical flexion ROM measurement associated with each of the five stages of the CCFT was calculated with intraclass correlation coefficients (ICCs) based on a two-way random effects model, a single rater type and an absolute agreement definition [37]. It included the analysis of “intra-rater” reliability (trial 1 versus trial 2) and “inter-rater” reliability (examiner A versus examiner B) of the measures. The standard error of measurement (SEM) was calculated considering the ICC, using the formula SEM = SD × \(\sqrt{1-\mathrm{ICC}}\) [38]. Differences between the stages of the CCFT in the amount of craniocervical flexion ROM were analysed by one way repeated measures analysis of variance (ANOVA).

International guidelines for reproducibility and validity research have reported that 40 subjects can be sufficient for reproducibility studies [33]. In our study, the sample size was calculated considering two raters, an expected ICC of 0.9 with 95% confidence interval (CI), and a confidence level of 0.05. Based on this calculation, 56 subjects were finally estimated to be included [39].

The use of technology is becoming an important method to evaluate cervical sensorimotor control [40]. This is the first study to investigate a system of inertial sensors to measure the ROM associated with the CCFT. However, the variability of the craniocervical flexion ROM was high and thus limits the possibility to accurately standardize a target ROM associated with each stage of the CCFT.

Our results showed that wearable inertial sensors were a reliable method to measure the craniocervical flexion ROM associated with each stage of the CCFT. One of the three components of the performance of the CCFT using the pressure biofeedback unit is to visually assess and confirm that the rotation of the head in the sagittal plane proportionally increase with the progressive stages of the test [15, 24]. However, the ability of the examiner to accurately observe the anterior rotation of the head is subjective, but the use of inertial sensors provides a precise electronic data record and feedback both for patients and examiners. Therefore, their use as described in this study could be implemented in future research and clinical practice as a user-friendly option to objectively monitor the ROM associated with each stage the CCFT, while computer feedback could provide performance-guidance for the examiner.

The execution of the CCFT requires that the clinician observes whether an altered movement strategy of a subtle retraction rather than a pure anterior sagittal rotation, occurs. This altered movement strategy has been associated with an average 5% reduction of the craniocervical flexion ROM at each stage of the CCFT [15]. The detection of this aberrant strategy is challenging for clinicians and it could also be facilitated by the use of inertial sensors, since they detect an absence of craniocervical flexion even though when there is an increase of the target pressure. This would indicate that a retraction movement is occurring without any craniocervical flexion.

Similar to previous research using digital imaging [14, 15, 30, 31], the present study has reliably observed that the amount of craniocervical flexion progressively increases during the five successive stages of the CCFT. Full craniocervical flexion ROM values observed in our study (14.6° ± 4.2°) are in line with those observed by Uthaikhup and Jull in a young healthy population (14.3º ± 5.2º) [31]. However, they seem significantly higher than those previously reported by Falla et al. [30] in young healthy adults (8.5º). This difference could be explained by the instructions given to the patients during the performance of the full range of active craniocervical flexion. With regard to the ROM values reached in each stage of the CCFT, the percentage of full craniocervical flexion observed in our study (22 mmHg: 27%; 24 mmHg: 37.5%; 26 mmHg: 47.4%; 28 mmHg: 54.9%; and 30 mmHg: 62.7%) also seem quite similar compared with those previously observed by Uthaikhup and Jull [31] in young healthy individuals. However, they seem slightly lower when compared to Falla et al. results [30], especially for the last two stages (22 mmHg: 24.9%; 24 mmHg: 41.9%; 26 mmHg: 55.8%; 28 mmHg: 66.5%; and 30 mmHg: 76.6%). These differences may be explained by the fact that participants in our study had increased values of the full range of the craniocervical flexion ROM, and therefore, the percentage of the full ROM necessary to reach the target pressure level was reduced.

To the authors’ knowledge, no previous research has investigated inertial sensors to measure the ROM associated with the CCFT or as an option to test and retrain the DCF muscles. The CCFT is limited for the patient in a supine position, since the air-filled pressure sensor needs to be compressed between the cervical lordosis and the bed. Although training approaches of the DCF muscles based on the CCFT have been demonstrated to improve the muscle function, previous research has observed that craniocervical flexor training in the supine position using feedback from an air-filled pressure sensor did not produce changes in muscle activation during a standing functional upper limb task. Thus, it was suggested that rehabilitation of the cervical muscles should be extended to include training in functional postures and tasks [32].

The results of the current study have shown that the variability of craniocervical ROM associated with each stage of the CCFT observed between subjects limits the possibility to standardize a set of targets of craniocervical flexion ROM equivalent to each of the pressure targets of the pressure biofeedback unit. This variability might be explained by various factors, such as anthropometric characteristics, cervical laxity and anatomical characteristics of the cervical lordosis or the occiput prominence, considering that the stages of the CCFT specifically monitor the slight flattening of the cervical lordosis that compresses the pressure biofeedback unit sensor.

The inclusion of young healthy participants in the present study may have reduced this variability, so larger variability ranges would be expected in patients who cannot successfully complete all stages of the CCFT or with pain conditions associated with poor sensorimotor control, losses of segmental mobility or alterations in central processing. Larger variability ranges of craniocervical flexion ROM during the CCFT have also been reported in the elderly and considered indicative of impaired fine motor and cognitive skills [31]. Moreover, differences between each of the stages of the CCFT \ in terms of the mean data of craniocervical flexion ROM and percentage of full ROM are always small (< 1.5º or 11% of the difference between two consecutive stages). Therefore, this factor also limits the possibility to accurately standardize a set of targets of ROM equivalent to each of the pressure targets of the pressure biofeedback unit, since the SD of the measures and the SEM of each stage tend to overlap.

Further research is needed to analyse the relationship between different targets of craniocervical ROM and the amplitude of DCF muscle activity, allowing for the standardization of a set of targets of craniocervical ROM associated with specific amplitudes of DCF muscle activity. Therefore, future research could investigate whether inertial sensors associated with computer feedback could provide an alternative option to test and retrain the DCF muscles in multiple positions or functional postures and tasks, accomplishing the needs of clinicians who treat patients with multiple craniocervical pain conditions.

This is the first study to investigate a novel inertial technology that could assist clinicians when performing the CCFT in patients with neck pain and associated conditions (i.e., headache, TMDs and craniofacial pain). However, the present study has various limitations. First, the results of this study are limited to the specific characteristics of the study population, which included asymptomatic, young Caucasian participants from a university community. Although this type of sample may be appropriate to describe the normative values of craniocervical flexion ROM that should be achieved in normal conditions during the CCFT, future research could investigate whether its performance is influenced by age, anthropometric features or anatomical characteristics of the cervical spine and the cranium. Previous research has observed that healthy elders showed larger variability in craniocervical flexion ROM for the five stages of the test as compared to young subjects [31]. Second, the muscle activation of superficial muscles while performing the CCFT was not monitored by electromyography. This could have allowed for an objective measure of subjects who could have been excluded from the study due to superficial muscle activity. Third, some factors could have provoked a different performance of the CCFT in each trial separated by 1 week, such as the previous training on the test or the physical and psychological status of the patient on the day and time they were measured.

Inertial sensors are a user-friendly option to objectively monitor the ROM associated with each of the pressure targets of the pressure biofeedback unit during the CCFT, providing a real time computer feedback. Therefore, their use as described in this study could be implemented in clinical practice and research to confirm that the craniocervical flexion ROM proportionally increases with the progressive stages of the CCFT and to detect any aberrant strategy.

Further research is needed to evaluate the specific muscle activity of the DCF muscles associated with various targets of craniocervical ROM to accurately determine which craniocervical flexion ROM best targets the DCF muscles.