Open Access
Add to BibSonomy
Abstract
This paper is an initial proof of concept for an optical speckled-based method for the evaluation of facial nerve paralysis. Differences between the affected and the healthy sides of the face were measured in patients with Bell's palsy—a peripheral facial nerve paralysis. The patients’ faces were illuminated with two symmetrical spots on their nasolabial folds and the reflected speckle patterns were analyzed. Muscle activity was evaluated by muscle tone contraction-release motion inducing associated skin tilting movements. The skin movements were imaged with a defocused lens, which enables extraction of the speckle pattern's time changing trajectory. We found an asymmetry ratio expression that may be the key for the estimation of the degeneration level in Bell's palsy.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Bell's Palsy is an idiopathic unilateral peripheral facial nerve palsy. The prognosis of Bell's palsy is tightly related to the maximal degeneration level, to degeneration rate and to time of initial function after onset of palsy [1–3]. Hence, a precise and sensitive evaluation is necessary for finding the appropriate treatment approach and the needed intervention. The current evaluation methods are mainly electromyography (EMG) and electroneurography (ENoG). Each method holds specific and unique limitations: The EMG may not reflect the nerve functioning in the acute phase of Bell's palsy while the ENoG, that uses electrical stimulus and recording of the compound action potential of the facial muscles, can provide useful information only in the early days of the insult due to the nerve degeneration characteristics [3,4,5]. In addition, both methods may cause some inconvenience to the patient.
The proposed optical set-up enables to detect minor movements of the muscle tone which cause skin perturbations. The muscle tone is the constant contraction-release movements that the muscles preform at rest, while no movement can be seen by a naked eye. In Bell's palsy the electrical signal that reaches the muscle is damaged due to the nerve degeneration and thus the muscle motion is affected. We assume that differences of skin movements between the paralyzed and the healthy sides of the face, measured by the speckle patterns analysis, may reflect the course and extent of underlying muscle injury.
2. Theoretical background
This paper presents an optical method that relies on speckle patterns examination for the analysis of facial muscle paralysis. Speckles are self-interference patterns occurring when coherent light hits a rough surface [6,7]. The roughness of the surface causes phase addition and constructive/destructive interference that appears as bright and dark spots. The muscle tone creates tilting movements in the tissues of the skin, and these movements are expressed as temporal changes in the speckle pattern. The correlation between muscle and skin movements is presented in Tenner's work [8,9]. According to Tenner et al., the muscle movement result in skin bulging which causes tilting movement of the skin. For focused imaging, the phase addition induced by the tilting movement will change the speckle pattern. Yet, when defocused imaging is used, an approximation of spatial Fourier transform is obtained, and the speckle pattern is preserved in its intensity for translational movement of the reflecting surface and it is shifted for angular movements (e.g. for the case of tilting). This is happening since linear phase addition in the spatial domain expresses as a transversal shift in the Fourier domain. This property is in the essence of our measurement methodology. We defocus our imaging lens and thus a different speckle like distribution is captured in our sensor. We perform spatial correlation between those intensity distributions obtained at each frame and by applying localization algorithms we find the location of the correlation peak. We plot the change with time of the position of the correlation peak and this signal is the signal from which we aim to extract our bio-medical estimation. This measurement methodology has been previously presented by former publications of the group [10]. To briefly support this concept, we assume a tilting movement of $({{\alpha_x},{\alpha_y}} )$ degrees which can be written as follows:
The main advantages of this optical approach are the option for non-contact evaluation and the nano-metric resolution achievable when using optical wavelengths.
3. Methodology and instrumentation
3.1. Analysis approach
The analysis procedure includes region of interest selection: 32 × 32 pixels from the speckle patterns center, frames correlation with the first frame and extraction of the speckle pattern's shift value with subpixel resolution by using cubic-Hermite interpolation [11]. Finally, concatenation of these shift values creates the speckle pattern tracking signal. The main assumption for the analysis approach is that muscles activity is pronounced in these tracking signals due to the muscles effect on the skin. By using the secondary speckle analysis and taking the advantage of its nano-scale sensitivity, we expected to recognize the muscle activity at rest, also known as the muscle tone. The energy of the speckle pattern tracking signal, noted as En, was extracted from the tracking signal x(n) according to the expression in Eq. (5), whereas x(n) is the trajectory of the speckle movement and N is the number of frames. The energy value was chosen thanks to its tight connection to the muscle movement: significant muscle movement causes major skin tilting, and this tilting is expressed as high shifts of the speckle pattern.
In order to detect the asymmetry that clearly appears in Bell's palsy patients, separate energy values were calculated for each of the speckle pattern transversal shift direction, horizontal and vertical, corresponding to the x and y axes, for each side of the face. An example for this tracking signal is presented in Fig. 1.
3.2. Experiments and set-up
The remote evaluation of the facial nerve degeneration was performed by examination of the secondary speckle patterns. These patterns were created by illuminating the patient’s face with a laser beam in two key points which express mimic activity generated by the facial nerve [12]. The set up consist of 532 nm Thorlabs laser beam (CPS532, max power 5 mW) split by a 50-50 beam splitter into two equal beams directed by a mirror to both patient’s nasolabial folds, as shown in Fig. 2. The selection of the nasolabial folds enables consistency in the laser illumination location for different patients. The reflected secondary speckle patterns were acquired using a Basler camera (A312f, resolution 782 × 582 pixels, pixel size 8.3 × 8.3 µm) with 75 mm lens and frame rate of 200 fps. Since the speckle patterns are reflected from a biological living tissue, and the wanted information of the skin bulging may be disturbed by internal movements such as blood flow, a separation between light scattered from the surface from the one coming from deeper layers of the inspected tissue is necessary. This separation is achieved by two methods: The selection of an adequate wavelength and the selection of the speckle pattern’s center which is positioned around the illumination spot. First, the 532 nm wavelength was chosen for its low penetration to the deeper layers of the skin. The skin layer optical properties, in spite its heterogenous structure, are mostly determined by the melanin presence. Therefore, lower wavelength penetrates less into the depth of the skin tissue [13,14]. Second, speckles positioned away from the illumination spot, were generated from photons that traveled longer path into the tissue and passed longer scattering path that caused their formation more far away from the point in which they were injected into the tissue (the illumination spot). Hence, we first recognized the speckle pattern center and just then we defocused the lens and got speckles on size of 2 × 2 pixels.
Seven subjects in the ages 27-65 participated in these experiments: three healthy volunteers subjects, three subjects with recent severe left Bell's palsy and one subject with incomplete Bell's palsy recovery. The subjects sat about one meter from the optical set-up, and during the measurements they were asked to stay still and not perform any facial movement. Due to the short experiment duration no numerical correction for involuntary movements was needed, and the recorded speckles’ movements result only from the effect of the muscle tone. Each measurement was recorded for 15-20 seconds. The study was approved by the ethic committee of the SMC 4107-17.
4. Results and discussion
Each analysis was made using a 10-seconds window. The three healthy subjects (Subject 1-3) were recorded with consistent protocol of three consecutive 20-seconds periods per subject, hence six 10-seconds windows per subject were extracted. Due to the dynamic clinical changes of the patients’ state, the recording procedures among patients were inconsistent. The three subjects with Bell's palsy (subjects 4-6) were recorded as follows: Subject 4 –was recorded for 20 seconds, creating two 10-seconds windows, Subject 5 – was recorded for 15 seconds - creating two overlap 10-seconds windows (seconds 0-10 and seconds 5-15) and Subject 6 – was recorded twice for 20 seconds, hence creating four 10-seconds windows. The recovered patient, Subject 7, was recorded for 20 seconds, resulting in two 10-seconds windows.
The asymmetry expressed in Bell's palsy was examined by energies calculation. As mentioned, the energy of the speckles tracking signal in both axes, x and y, is related to the muscles’ activity in the illuminated location, due to the effect of the muscle's movement under the skin. At first, the median energy for one healthy subject and one Bell's palsy subject for x and y axes was plotted, as presented in Fig. 3. The energy calculation was conducted according to Eq. (5).
Examination of Fig. 3 rises the assumption that the symmetry expected for healthy subjects is pronounced in the energies, noted as En, of each element in the following ratio:
Another way to write the above expression is: To check this assumption, the two ratios in Eq. (7) were calculated for each ten seconds speckle patterns recording, and a scatter plot using these ratios as the axes is presented in Fig. 4.According to Fig. 4 one can notice that there is clear differentiation between the subjects’ state in this ratios’ space. Another way to present this ratios’ assumption is by calculating the quotient between the two ratios in Eq. (6). From Fig. 4 one can notice that the healthy markers are close to the ratio equality line, therefore the ratios quotient value is expected to be close to one for healthy patients and far from one for Bell's palsy patients:
Though these preliminary results are promising, more comprehensive research is needed to validate the presented results. Due to the low number of subjects, it was hard to determine whether this differentiation is due to the subjects’ clinical state or rather due to the variety among subjects. Also, subject recordings were made in different conditions. To confirm this evaluation method, larger population sample at same recording conditions is suggested.
5. Conclusion
In this paper we present a non-contact optical method for the evaluation of muscle activity in Bell's palsy. Secondary speckle patterns were recorded from the face, demonstrating differences between healthy and affected sides, at rest. We used a mathematical expression that may be the key for the evaluation of the facial asymmetry in Bell's palsy. Yet, this expression's reliability must be confirmed by extending the participants’ number.
This innovative non-invasive technique could be used for monitoring the course of Bell’s palsy. Future work should be focused on patients monitoring for long duration, proper calibration and during active muscle contractions.
This will enable non-contact patients monitoring during both, the acute and the recovery phases of Bell's palsy.
Disclosures
The authors declare no conflicts of interest.
References
1. E. Peitersen, “Natural history of Bell's palsy,” Acta Oto-Laryngol. 112(sup492), 122–124 (1992). [CrossRef]
2. U. Fisch, “Surgery for Bell’s palsy,” Arch. Otolaryngol., Head Neck Surg. 107(1), 1–11 (1981). [CrossRef]
3. D. H. Gilden, “Bell's palsy,” N. Engl. J. Med. 351(13), 1323–1331 (2004). [CrossRef]
4. D. Lee, “Clinical efficacy of electroneurography in acute facial paralysis,” J. Audiol. Otol. 20(1), 8–12 (2016). [CrossRef]
5. G. M. Weiner and N. J. Holland, “Recent developments in Bell's palsy,” BMJ 329(7474), 1104 (2004). [CrossRef]
6. J. C. Dainty, Laser speckle and related phenomena (Springer-Verlag, 1984).
7. N. Ozana, I. Margalith, Y. Beiderman, M. Kunin, G. A. Campino, R. Gerasi, and Z. Zalevsky, “Demonstration of a remote optical measurement configuration that correlates with breathing, heart rate, pulse pressure, blood coagulation, and blood oxygenation,” Proc. IEEE 103(2), 248–262 (2015). [CrossRef]
8. F. Tenner, M. Regensburger, A. Schramm, M. Sohle, K. Schwarzkopf, Z. Zalevsky, and M. Schmidt, “Evaluation of a laser-based sensor for the diagnosis of neurological disorders,” 2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 4231–4234 (2017).
9. F. Tenner, A. Schramm, M. sohle, M. Regensburger, E. Wirhmann, Z. Zalevsky, and M. Schmidt, “Towards a multi-sensor system for the diagnosis of neurological disorders,” 2016 IEEE International Conference on Advanced Intelligent Mechatronics (AIM), 495–500 (2016).
10. Z. Zalevsky, Y. Beiderman, I. Margalit, S. Gingold, M. Teicher, V. Mico, and J. Garcia, “Simultaneous remote extraction of multiple speech sources and heart beats from secondary speckles pattern,” Opt. Express 17(24), 21566–21580 (2009). [CrossRef]
11. J. D. Foley, A. van Dam, S. K. Feiner, and J. F. Hughes, Computer Graphics: Principles and Practice (Addison-Wesley, 1997).
12. S. H. Kim, E. W. Ryu, C. W. Yang, S. G. Yeo, M. S. Park, and J. Y. Byun, “The prognostic value of electroneurography of Bell's palsy at the orbicularis oculi versus nasolabial fold,” Laryngoscope 126(7), 1644–1648 (2016). [CrossRef]
13. G. V. G. Baranoski and A. Krishnaswamy, “An introduction to light interaction with human skin,” RITA 11(1), 33–62 (2004). [CrossRef]
14. Z. Gajinov, M. Matic, S. Prcic, and V. Duran, “Optical Properties of the human skin,” Serbian J. Dermatology Venereol. 2(4), 131–136 (2010). [CrossRef]
