Objective analysis of the effectiveness of facial massage using breakthrough computed tomographic technology: A preliminary pilot study

Facial massage is empirically known to be associated with morphological changes, such as improvements in facial sagging. However, quantified objective evaluations of massage‐induced changes have not been performed to date. This preliminary pilot study aimed to verify the effectiveness of facial massages by using breakthrough computed tomographic technology.


INTRODUCTION
Interest in anti-aging therapies and beauty treatments has been increasing in aging societies. In addition to providing mental satisfaction, relaxation, and improvements in skin texture, 1-4 beauty treatments such as facial massages also result in morphological changes in the face, such as improvements in facial sagging and lifting effects on the cheeks. 5,6 The soft tissues, including the superficial musculoaponeurotic system (SMAS) and subcutaneous adipose tissue, 7,8 are distributed subcutaneously, and facial massage may change their morphology.
To date, the effectiveness of facial massage has been mainly evaluated by subjective assessments such as visual methods and photographic comparisons. However, techniques based on stereo-image correlation 5 and computed tomography (CT) 6 have been recently proposed for objective evaluation of the effectiveness of facial massage. Nevertheless, these evaluations were limited to changes in the surface morphology of the face, and they did not clarify the effects of facial massage on subcutaneous structures. Therefore, verifying the effect of facial massage, which has been discussed empirically, using an objective method was considered a highly noteworthy method.
Highly accurate and detailed three-dimensional (3D) CT images can be constructed using advanced spiral CT known as the multidetectorrow CT (MDCT) technique for capturing high-resolution images and a workstation for processing and analyzing a large volume of images. [7][8][9] Moreover, spiral CT examination is a highly objective examination. The spiral CT imaging data obtained using this approach contain a large amount of anatomical information, which can be used to construct 3D images of the surface and subcutaneous structures of the face. In addition, multiplanar reformatted images, such as axial, coronal, and sagittal images, can be useful for understanding the anatomical relationship between morphological changes of the face and subcutaneous structures.
Therefore, this preliminary pilot study aimed to analyze the changes in the cheek state and SMAS following facial massage and thereby objectively determine the effectiveness of facial massages by using the breakthrough CT technology. were enrolled in the present study. The mean (± standard deviation) age of these participants was 33.8 ± 3.56 (range, 29-37) years.

Participants
All five participants underwent CT examinations at the beginning of the study. Subsequently, they performed self-facial massage in accordance with a specified method for approximately 90 s ( Figure 1). The massages were performed using the same cosmetic emulsion daily in the morning and evening for 2 weeks. Self-massage is a way to alleviate muscle stiffness and to pull up the cheeks using the fingers, based on attention to the lymphatic flow. Two weeks after they started performing self-massages, all participants underwent a final massage by a professional technician using the same method as that used for the self-facial massage but for double the duration. Subsequently, all five participants underwent a second CT examination.

Image acquisition
All the CT examinations were conducted using a spiral CT having 320

Analysis of computed tomography images
Using the detailed CT imaging data of the C-Fs, facial 3DCT and recon-

Cheek thickness
Cheek thickness measurements were performed on both cheeks (total 10 cheeks) of five participants ( Figure 2). First, as shown in Figure 2A,

Shifts in the location of the malar top
Measurements of the massage-induced differences in the location of the malar top were performed on 10 cheeks of five participants ( Figure 3). As shown in Figure 3A, the malar tops of both cheeks were confirmed before and after the facial massage. As shown in Figure 3B, the facial massage-induced differences in the malar top location were determined in terms of the craniocaudal and horizontal distances.
The massage-induced shift in the malar top was calculated using the Pythagorean theorem with the craniocaudal and horizontal distances.

SMAS measurement
As shown in Figure

Statistical analysis
Participant age, cheek thickness, shifts of the malar top, SMAS-width, and SMAS-height are presented as mean ± standard deviation values.
The mean values for cheek thickness, SMAS-width, and SMAS-height before and after massage were compared using the mean values of a paired t-test.
A correlation analysis was performed to determine the relationship between the change rates of cheek thickness and SMAS-width due to facial massage. Similarly, the correlation between the change rates of cheek thickness and the SMAS-height due to facial massage was also determined.
Statistical analyses were performed using the StatMate V statistical software package (Nihon 3 B Scientific Inc., Niigata, Japan). Statistical significance was set at P < 0.05.

RESULTS
The overall quality of the 3DCT and reconstructed images of all participants was adequate and did not hinder measurements of the cheek thickness, location of the malar top, and the SMAS.

Cheek thickness
The mean pre-massage thickness of the malar top was 22.0 ± 1.34 mm, and the cheek thickness decreased by 21.2 ± 1.08 mm due to facial massage. The mean change rate of cheek thickness was −0.8 ± 0.45%,

Shifts in the location of the malar top
As shown in Figure 6, the mean craniocaudal shift in the 10 malar tops of the five participants was 2.7 ± 1.44 mm, with nine malar tops shifting cranially and one remaining in the same location. The mean horizontal shift of the 10 malar tops was 2.7 ± 1.48 mm, with eight malar tops shifting laterally, one shifting medially, and one remaining in the same location. On the basis of the cranial and horizontal distances, the mean shifting distance of the malar top was calculated to be 3.9 ± 1.94 mm.

SMAS measurements and their relationships with cheek thickness
The mean pre-and post-massage SMAS-widths of the five participants' faces were 122.7 ± 4.35 mm and 122.4 ± 4.03 mm, respectively. The mean change rate of the SMAS-width was −0.20% ± 1.01%. The SMASwidth tended to decrease with massage; however, the difference was not statistically significant (P = 0.66).
The mean pre-and post-massage SMAS-heights of the 10 cheeks of the five participants were 37.8 ± 13.57 mm and 38.7 ± 13.75 mm, respectively. The mean change rate of the SMAS-height was 2.6% ± 2.6%, and the mean SMAS-height significantly increased after facial massage (P < 0.05). In general, soft tissues, including the SMAS and subcutaneous adipose tissue, were shown to be susceptible to the effects of gravity. [10][11][12] The gravity vector applied to the face is different in the upright and supine positions, and the morphology of the soft tissues is also different. Therefore, the facial appearance in a supine position is younger than those in standing or sitting positions. 10,11,13,14 Thus, evaluations The workstation is installed with an application that uses various preset reconstruction algorithms, including dedicated face scan algorithms, to reconstruct 3DCT images of the face. 15,16 Volumerendering, based on an edge-detection image processing system, was used in these 3D reconstructions. Spiral CT imaging data contain 3D information and include a coordinate axis for the area imaged using the spiral CT system. One of the advantages of the workstation is that the same point can be precisely indicated on 3DCT and all multiplanar reformatted axial and sagittal images by utilizing spiral CT coordinates.
Moreover, in addition to the facial 3D state, the anatomical relationship between the facial appearance and the subcutaneous structures can be accurately displayed with this approach. 7,8,11 A further innovation of our study was to use stationary facial bones, which were not changed by massage, as the landmarks for creating images for comparison before and after facial massage. This allowed us to quantify minute soft-tissue changes caused by facial massage.
The malar tops of nine cheeks in five participants were shifted cra- This study was a pilot study conducted prior to a large-scale study that would enroll several participants. The possibility of quantifying the fine changes caused by facial massage was attempted. The

F I G U R E 6
Shift in the location of the malar top due to facial massage. Individual differences in the massage effect were recognized, and there was a difference in the massage effect between the left and right sides even in one participant. They are as follows. Blue: participant 1, red: participant 2, black: participant 3, orange: participant 4, green: participant 5. On the right cheeks of the five participants, the malar tops shifted cranially toward. In assessments of horizontal shift, the malar tops shifted laterally and medially in four and one participant, respectively. On the other hand, the left cheeks of four participants showed cranial shifting of the malar top. In assessments of horizontal shift, the malar tops of four participants shifted laterally. The mean craniocaudal shift of the 10 malar tops was 2.7 ± 1.44 mm, with 9 malar tops shifting cranially and remaining at the same location. The mean horizontal shift of the 10 malar tops was 2.7 ± 1.48 mm, with eight malar tops shifting laterally, one shifting medially, and one remaining in the same location F I G U R E 7 Correlation of the change rates of superficial musculoaponeurotic system (SMAS)-height and cheek thickness. In the 10 cheeks of the five participants, a significant negative correlation was detected between the change rates of cheek thickness and SMAS-height (r = −0.63; P < 0.05) limitations of this study were the small-sample size, limited age range (29-36 years), and the factors that it does not take gender differences in the skin texture as well as the environment surrounding participants into consideration. However, despite the small number of participants, we were able to quantify the effectiveness of facial massage and conduct an objective analysis because the analysis using breakthrough CT technology is highly objective. We believe that the confirmation of the effects of massage, which had been discussed empirically, in an objective method is a highly significant finding. Future studies should aim to examine how factors such as age, gender differences, facial structures, and environment are involved in the effectiveness of facial massage in a larger number of subjects.

CONCLUSIONS
Using breakthrough CT technology, we conducted a detailed analysis of the effects of massage on the facial surface in this preliminary pilot study. Facial massage appeared to show lifting and tightening effects.
It caused the cheeks to shift cranially, and thick cheeks became thinner. Moreover, the SMAS-height increased. Our results provide useful information for beauty treatments and could contribute to the objective scientific literature for facial massages. We expect that these diagnostic imaging data for facial massage will facilitate the development of new techniques.