RO132296A0 - Method and device for studying dental restoration materials adherence to enamel and dentin - Google Patents

Abstract

The invention relates to a method and a device for studying the adherence of dental restoration materials to enamel and dentin. The claimed device comprises a laser excitation source (2) which emits a laser beam which is very precisely directed onto the sample to be investigated by means of an electrically-controlled mirror system (7), a selective voltmeter (3), which generates a reference signal for the intensity-modulated laser, an infrared chamber (4) which registers the thermal signal generated by the sample and has incorporated a synchronous detection module for receiving both the thermal signal and the reference signal transmitted by the selective voltmeter (3), which processes the received signals and generates some images of amplitude and phase corresponding to the excitation frequency, images which are afterwards displayed on a computer screen (5) wherefrom they can be exported for subsequent processing. The claimed method consists in using infrared synchronous detection thermography coupled with contour recognition image processing algorithms to emphasize the useful information in the defect area and remove noises, to finally obtain a binary image, namely a pixel matrix with values of 0 or 1 which delivers the particular signature of the surface defect.

Description

a) Specification of the technical field in which the invention can be used.

The invention relates to an automated experimental line of active thermoglia and the data analysis method for investigating enamel and dentin adhesion of dental restoration materials.

The concept is one found in principle in many other fields, which probably caused it to be taken over and used in dental medicine at one time, because of the undeniable benefits it can offer. Practically, in dentistry, the ultimate goal of adhesion is to join a solid substrate (hard / soft dental tissues - enamel, dentine, cement) with another substrate, represented by the restoration material or materials. Most of the times at the tooth / restoration interface there is a space accessible to the marginal infiltrations which in time favors the appearance of secondary cavities.

The lack of adhesion between the restorative materials and the tooth has been and remains a major problem in dentistry. The quality of adhesion and dental structures is a clinical and practical necessity. For this reason, research in the field of adhesion to dental structures using non-invasive techniques is a subject far from being completed.

Due to the high sensitivity of the proposed method, the invention can be used for the study of tooth-seal adhesion, while also allowing the testing of the sealing capacity of different types of dental adhesive systems, respectively of the commercial products existing within each class of dental adhesives. Therefore, the complete evaluation of a dental restoration by a non-invasive technique is an extremely useful tool for dentists.

b) Indication of the prior state of the art and indication of the documents underlying it.

Currently, in restorative dentistry there are two current investigative techniques used only to evaluate the marginal adaptation of composites to the tooth: (1) penetrating fluid method (2) scanning electron microscopy (SEM) [1] [2] [3] [4], Thermography is a widely used method in medicine [5], but too little in dentistry, despite its non-invasive character. In dentistry, IR thermography is mainly used to monitor the polymerization temperature of dental composites during the curing process [6] [7]. Therefore, there is not much data available in the literature on this topic. There are three recent publications (2013, 2015) in which the interface between the filling material and the tooth was evaluated by means of synchronous detection thermography (LIT). The amplitude and phase images were analyzed according to the excitation frequency and the diameter of the exfoliation zone for maximum contrast [8] [9]. Synchronous detection thermography has given promising results in detecting marginal and internal defects through irnagjxjjfor _x analysis.

Ζ · / '''foNCOf'XXj 1 (, -Ν'- ^P / I>' 7 -ri / s

to 2017 00298

5/18/2017 amplitude. The procedure allowed a diagnosis of very fine microfissures (up to 1pm). [10]

Mandelis et al. solved the thermal diffusion equation taking into account the phenomenon of multiple photon scattering in the enamel and analyzed the photothermal response of the tooth following the excitation process [11] [12] [13] [14]. In general, the greater the scattering of light in a certain volume of the tooth, the greater the probability of absorption of a photon in this region. Therefore, the thermal waves that are generated in the porous or defective regions of the tooth will have a greater amplitude than those generated in the intact enamel [15] [18]. Research has shown that PTR photothermal radiometry can successfully detect cracks in the tooth [15], and can also monitor early demineralization and remineralization of enamel [15] [16] [17],

At the national level, to our knowledge, there is no research group that applies synchronous detection thermography in medicine. There are only a few studies on the evaluation of dental polymers used in dentistry by passive thermography []. For these reasons, we believe that this procedure will bring many benefits at national level by developing new concepts and approaches regarding the non-destructive evaluation studies of dental biointerfaces.

c) Exposure of the invention in terms that will allow to understand the technical problem and the solution as claimed and the advantages of the invention in relation to the present state of the art.

The device developed and the method of image analysis described in the present invention will allow the non-destructive detection of the microfissures located at the interface of the material / enamel or material / dentin with very high accuracy, while also allowing the testing of the sealing capacity of the different adhesive and / or dental composites.

dental restoration interfaces can have defects, known as microieakages (microfissures) that are not at all easy to diagnose. In this method, an intensity modulated excitation source (light) is sent to the sample to be investigated, generating a temperature field modulated at the absorption site. This modulated thermal field emitted by the sample (a complex signal having an amplitude and a phase) can be measured with a two-dimensional infrared detector (IR detector).

Difficult elements of the problem: The tooth has a multilayer structure (enamel, dentine, pulp and cement), each layer having its own inhomogeneities and thermal and optical properties. light scattering in the tooth is large, this being a very optically inhomogeneous environment. The depth of light penetration in the tooth depends on the energy of the incident photons: the large wavelengths (small energies of the incident photons) are less scattered compared to the short wavelengths (higher energies of the incident photons), and can therefore penetrate more. deep in the structure of the tooth. Therefore, the amount of light scattered is a key factor in determining the depth of light penetration into the tissue. A photothermal signature is only purified

W '4 IA

 to 2017 00298

05/18/2017 when heat energy is released following an absorption event of a photon in the analyzed tissue. Therefore, in order to obtain an optimum photothermal response (signal), the wavelength of the excitation source must be carefully chosen, taking into account both the scattering coefficients and the light absorption coefficients in the enamel and dentine. Due to the moderate absorption and strong scattering, the signal generation in such semi-transparent samples is weak, which is why they are difficult to investigate. To increase the signal-to-noise ratio, the energy of the incidence light source on the tooth may be increased, but this may cause the dental safety limit to be exceeded. For example, a laser power density of 2W / cm ² causes a temperature increase of approximately 5 ° C inside the pulp chamber, which can cause irreversible trauma to the tooth. One solution would be to reduce the optical power density in favor of longer signal acquisition times. This can be achieved by synchronous signal detection techniques.

The device and method described by the present invention are based on the use of IR synchronous detection thermography coupled with image processing algorithms for contour recognition, for the purpose of non-destructive characterization of dental interfaces. The advantage of the method consists in the possibility of extracting the useful signal from the thermal noise (signal / noise ratio> 100) and exploiting it by means of suitable image processing techniques. In addition, the automated and precise positioning of the Laser spot on the investigated surface, in the vicinity of the restored cavity, will allow to obtain thermal images having a very good contrast. This eliminates the need for optical alignment (very time consuming) necessary to investigate each cavity restored in part, a single calibration procedure is sufficient to obtain the information needed for a whole set of samples.

The novelty of the present invention consists of:

(1) the possibility of very precise control of the position of the laser spot on the investigated surface (with precision of the order of 1 pixel) by laser deflection on an electrically controlled mirror system.

2) implementation of image processing algorithms that improve the quality of the resulting images by accentuating the signature of the defective areas and reducing the blurring effect due to the effects of lateral heat dissipation.

d) Detailed description of the invention for which protection is sought.

The block diagram of the experimental line of synchronous detection thermography for the analysis of the dental interfaces is represented in figure 1.

The selective lock-in voltmeter SR generates the reference signal for the intensity modulated laser (with the excitation frequency f ₀ ). By deflection on the mirror sjșifâ ^

£ 7 to 2017 00298

18/05/2017 electrically controlled, the laser beam is directed very precisely in the vicinity of the analyzed dental interface. The restored tooth slice is rigidly fixed to a sample positioning system. The thermal signal generated by the sample is recorded by the thermographic chamber. The thermal signal and the reference signal are sent to the synchronous detection module of the camera that processes the received information. Finally, the amplitude and phase images corresponding to the excitation frequency are obtained, practically filtering the signals with frequencies other than the excitation frequency fo. The resulting images are displayed on a computer, where they are exported for processing.

The efficient optimization of the process proposed by the present invention can be done only by using methods based on high precision experimental measurements, coupled with image processing algorithms. The amplitude and phase images obtained from the synchronous detection contain information about the presence of a crack or a demineralized area (compared to the intact enamel) which consists of a disturbance of the thermal wave above the fault zone. This variation may be only a few mK, impossible to detect by passive thermography (thermal fluctuations -0.5K). Image processing algorithms accentuate the useful information above the fault zone and eliminate noise, and finally a binary image (a pixel array with zero or 1 values) is obtained that provides the particular signature of the defect on the surface.

1. The automation of the experimental parameters to improve the contrast of the thermal images as the distance from the investigated interface at which the tissue is irradiated and / or the filling material must be controlled very precisely, the alignment process was automated by the deflection of the laser spot by a set of mirrors on two axes (XY), which are electrically controlled. The spot is deflected by the composite movement of the two mirrors. In this way the mechanical alignment of the sample / excitation system is avoided, a process that is very time consuming. The whole mirror control system must be capable of avoiding measurement errors [18] [19] [20]. These errors are:

(i) flatness errors in fixing on the optical mass of any element that makes up the optical system; (ii) parallax errors between the FLIR camera and the sample positioning system, errors due to the angular positioning of the laser after exiting the mirror system.

The mathematical model: the parallax correction

The flatness errors and the parallax errors are corrected using a type of transform called homographic transform or projection transform. This transform defines a geometrical relationship between all points in the input plane (A) and all points in the output plane (B). The use of the projection transform implies the existence of a minimum of 4 points in the plane A which have the known coordinates. These points must also correspond to a minimum number of 4 points in. tt Y '\\ // do »Q <> \\

·.! t a 2017 00298

18/05/2017 plan B. This type of geometric transform imposes at least the presence of a quadrilateral in the entry plane, thus resulting in a conversion system with 8 degrees of freedom.

The output image will have a modified geometric shape but having precisely known the coordinates of the points projected from the plane A to the plane B. Thus it is of interest to determine the unknown coordinates of a point from the plane B (u, v) by projecting an arbitrary point from the plane A having the known coordinates (x, y).

The generalized form of the projection transform is given by the system of equations (1):

V 0 ^x t 0 y ₀ 0 0 t 0 and 0 0 ^{x or} 0 y 0 y ₀ 0 y, 0 1 0 1 -/a ~ XM, ~ Xy. “>'O ^v o -y / i -yy of b c d e f = % ^v o u, ^v >, x "_. y "_. and 0 0 0 and, g RT-1 0 0 0 x '_I y ,, -! 1 -K-in-i .. _V '_ h

where the coefficients a, b, c, d, e, f, g, h are unknown, and the vector [w ₀ v ₀ - ··] represents the electric coordinates corresponding to plane B.

For each coordinate x, y in plane A, the new electrical coordinates (u, v) belonging to plane B are determined and expressed as follows:

__ ax + by + c ^u / ^h > + '(2, dx + ev + fv gx + hy +1

The geometrical significance of the calculated constants a, b, c, d, e, f, g, h is as follows: a: scaling coefficient on the x-axis having the value Y unchanged b: scaling coefficient on the x-axis proportional to the distance from Y to y origin c: translational coefficient of the origin on the x-axis d: scaling coefficient on the y-axis proportional to the distance of the x to the origin e: scalar coefficient on the y-axis having the value X unchanged f: translation coefficient of the origin on the y-axis g : proportional scaling coefficient for X and Y as a function of X h: proportional scaling coefficient for X and Y as a function of Y

The constants a, b, c, d, e, f, g, h are determined according to a preliminary calibration procedure: the initial geometric coordinates of the 4 calibration points to which the initial electric coordinates of the mirror correspond, after which the system is resolved of

to 2017 00298

18/05/2017 equations (1) and determine the new electrical coordinates belonging to plane B which are correlated with the geometric coordinates of plane A through the relation (2)

The projection transform is the most complex of the geometric transforms, because it includes in the set of mathematical operations an operation of; translation, rotation, shearing and scaling.

It can thus be stated that, by correctly manipulating the set of coordinates in plane A, respectively plane B, the singular use of the projection transform ensures the complete correction of the set of errors that appeared in the optical system. The construction of the system of linear equations involves a software interface for automatically taking the coordinates for the plane A represented by the FOV of the FLiR camera, respectively the plane β represented by the surface of the scanned sample, the software interface is presented in annex i.

2. Thermographic image processing The mathematical model used

2.1 Analysis of the amplitude images with respect to the amplitude image Ι _{Α (ΧιΥ} ), the proposed method for micro-crack detection consists of the following steps;

(i) introducing a threshold factor (a), which creates a mask for the amplitude image;

if A (x, y) j A _maK > a if A (x, y) / A _mm <a (3)

The mask, ie Φ _β (.χ, ν), is defined according to equation (3), where A _max is the maximum amplitude and A (x, y) is the pixel amplitude p _x , _y of coordinates (x, y). The higher the threshold factor a, the larger the masked area.

(ii) the application of the spatial derivative of order 2 (iaplacânan), in order to underline the local contrast, which contains the useful signal; (iii) the spatial filtering of the fluctuations of the excitation source and the binarization of the image (introduction of the second threshold factor b) ;

To obtain the image resulting from binarization, a second threshold factor b {b> 1) was introduced. The Laplacian operator is applied to the input image resulting in the image V ² 1 _A (p _xy ) = V ² A (x, y). The image resulting from applying the second threshold is given by equation (4).

to 2017 00298

05/18/2017

î if ^ A (x, y) ^ _a (x, y) <A _m Jb 0 //ν\4(χ,> -) χΦ _ο (χ, ^)> Λ, π / ^ά (4) where 4 _njn is the minimum value of the expression ν ² .4 (χ,>') ^χ Φ _ο (%,>').

The higher the threshold factor b, the higher the detection sensitivity but also the higher the noise. The larger the parameter c, the greater the number of interconnected pixels is considered significant to form a crack.

(iv) filtering the resulting image by introducing the third factor (c) which allows to keep a minimum number of adjacent pixels.

2.2 Analysis of phase images

The processing steps using the phase image are: (i) the preprocessing step, (ii) the edge detection stage based on the 2D Laplacian image and (iii) a refinement procedure based on the mathematical morphology to extract the surface defect.

Preprocessing involves a procedure for normalizing and scaling information, followed by a linear transformation based on the hyperbolic tangent function (tanh). The role of this nonlinear transformation is to improve the images, namely to decrease the background and to emphasize the useful information in relation to the threshold k. The result is an intensification of the contour information. The next step in the preprocessing stage is an anisotropic diffusion based on partial derivative equations (PDE) to eliminate noise and accentuate the contours (without destroying the contour information).

Considering the grayscale image obtained from pre-processing / 'as input image, after applying Laplacian we obtain another image denoted by L (i ^T ). The pixel values of the resulting image are described by eq. (5).

(5)

Scrolls in the case of the resulting image L (and ^T ) mark the contours.

Example of realization

The image obtained by the FLIR camera must be correlated with the correct coordinates in which the laser exposes the tested sample. For this purpose, before each measurement session, a calibration of the optical system and the control circuit of the mirrors is performed. The first sequence in the program generates a set of coordinates for such calibration; The USB-6211 Ni interface uses the two output channels AOO, AO1 to generate a control voltage for each x, y channel. Electrical control of the mirrors implies

to 2017 00298

5/18/2017 ordering them in 4 quadrants, so that the check of the control interface callback is limited to measuring the voltage of 0V on each axis. The values X = 0V and Y = GV correspond to the coordinate point 0.0 in the plane of the sample to be measured. To begin with, the coordinates 0,0 are generated for the correlation of the optical axis, with the electrical equivalent of the position of mechanical origin of the mirrors. Further, the mirrors are so controlled on the X-Y axes that the laser projects a rectangle bounded at the corners by a bright spot that delimits the optical field of the FLIR camera. The coordinates of these points can be controlled interactively from the control panel of the program in all four quadrants. In order to maintain good visibility of the calibration rectangle, the laser light intensity is initially not modulated and the signal sampling speed at the output of the control interface is 500 Hz.

The calibration points obtained on the sample surface and acquired using the FLIR camera are shown in figure 1. Figure 3 shows the sample surface investigated in the visible spectrum. For calibration, an opaque PVC sample was made that does not diffuse light. For the image obtained in IR, it is possible to observe the displacement of the optical center due to the accumulation of parallax errors. The reduction of the correlation errors between the image plane obtained by the FLIR camera, respectively the plane generated by laser scanning, is done by applying the mathematical model of parallax correction.

The numerical processing has the following performances: it corrects the parallax errors occurring at the level of the entire optical system and ensures the conversion between the reference system of the FLIR camera plane and the plane in which the working sample is found. Immediately after marking with a label assigned to the calibration coordinates, the laser is positioned at coordinates outside the FOV of the FLIR camera to avoid overheating of the sample.

The points of interest on the sample surface are marked to determine the trajectory on which the laser beam will explore the surface. An example of dynamic scanning of the surface of interest is shown in Figure 4. The constructive details of the sample support and the mirror support are shown in Figure 5.

The program that manages the FLIR camera allows to save the marked points in a txt file. This file is then read by the application written in LabView to automatically enter the corrections mentioned above. Performance of the block

The correction for a set of real coordinates is shown in Figure 6. The perfect correlation with respect to Figures 2 and 3. is shown. The point marked in red indicates the position of the optical center. The symmetrical displacement on the Y axis is due to the reversal of the image of the FLIR camera.

Image processing

Dental interfaces were analyzed using the alignment and image processing procedures described above. The thermal imaging camera used is 7200,

of 2017 00298 having a matrix of 256x320 quantum InSb detectors sensitive in the range 1.5pm5.1pm, a sensitivity in temperature of 25mK and focal length of the G1 type lens of 30mm. The acquisition frequency of the camera was 100 Hz. The excitation frequency of the laser radiation (laser Nd: YÂG, P = 50mW and Ă = 532nm) was f = 0.5Hz. The laser radiation was focused in the vicinity of the restored interface so that it is located in the heat diffusion area, in order to obtain information regarding the eventual discontinuities (microgaps) located at the interface.

The results of the image analysis are shown in Figures 7 and 8.

In addition to the discontinuity detected at the filler-dentin material interface starting from the amplitude image (Figure 7b), the presence of very fine localized discontinuities is observed in the filler-enamel material interface (Figure 8d). This means that the phase image is more sensitive than the amplitude image in microgapuriuri detection.

Bibliography [1] Meriwether, LA, Blen, BJ, Benson, JH, Hatch, RH, Tantbirojn, D, Versluis, A. Shrinkage stress compensation in composite-restored teeth: Relaxation or hygroscopic expansion, Dental Materials, vol. 29, Issue: 5, pp. 573-579, (2013).

[2] Tuncer D, Celik C, Cehreli SB et al. : Comparison of microleakage of a multi-mode adhesive system with contemporary adhesives in elass II resin restorations, Journal of Adhesion Science and Technology, vol. 28, pp. 1288-1297, (2014).

[3] Zhang Y, Yu Q, Wang Y., Non-thermal atmospheric plasmas in dental restoration: Improved resin adhesive penetration, Journal of Dentistry vol. 42, pp. 1033-1042, (2014).

[4] Kim RJY, Choi NS, Ferracane J., Lee I. B.: Acoustic emission analysis of the effect of simulated pulp! pressure and cavity tvpe on the tooth-cornposite interfacial de-bonding, Dental Materials, you. 30, pp. 876-883, (2014).

[5] B.B. Lahiri, S. Bagavathiappan, T. Jayakumar, John Philip, Medical applications of infrared thermography: A review, Infrared Physics & Technology 55, pp. 221-235, (2012).

[6] Sertae Aksakalli, Abdullah Demir, Murat Selek, Sakir Tasdemir: Temperature increase during orthodontic bonding with different curing units using an infrared camera, Acta Odontologica Scandinavica (doi: 10.3109 / 00016357.2013.794954), (2013), [7] Gums M, Devito-Moraes A, Francei C, Moraes R, Pereira T, Froes-Salgado N, Yamazaki L, Silva L, Zezell D., Temperature Increase at the Light Guide Tip of 15 Contemporary LED Units and Thermal Variation at the Pulp! Floor of Cavities: An Infrared Thermographic Analysis. Dentistry vol.38.no, 3, pp. 324-33, (2013).

K '' NCi A '1 \ x iY Q U'. a

DM λ U? // <· '?

 to 2017 00298

5/18/2017 [8] Gu, Ja-Uk; Choi, Nak-Sam; : NDE of the internai hole defect of Dental eomposite restoration using infared loek-in thermography, Journal of the Korean Society for Nondestructive Testing, vol. 33, pp. 40-45, (2013), [9] Gu, Ja-Uk; Choi, Nak-Sam: Evaluation of Delamination of Dental Composite Restoration using lnfrared Lock-in Thermography, Composite Research, vol., 25, pp. 236 -240, (2012).

[10] M. Streza & ak: Lock-in thermography, penetrant inspection and scanning electron microscopy for quantitative evaluation of open micro-cracks at the tooth-restoration interface J.Phys. D: Appl. Phys, you. 48, pp. 105401 (2015).

[11] A. Mandelis & al. : Ninei dental depth profilometric imaging using simultaneous frequency-domain infrared photothermal radiometry and laser luminescence, Proceeding of SPIE, vol. 3916, pp. 130, (2000).

[12] A. Matvienko, R. Jeon, A. Mandelis, S. H. Abrams, Β. T. Amaechi: Photothermal detection of Incipient Dental Caries: Experimental and Modeling, Proc. of SPÎE you. 6759 67590J-1, (2007).

[13] RJ Jeon, A. Mandelis, V. Sanchez, and S. H, Abrams: Non-intrusive, non-contacting frequency-domain photothermal radiometry and luminescence depth profilometry of natural carious and artificial sub-surface lesions in human teeth, Journal of Biomedical Optics, vol. 9, no. 4, pp. 804-819, (2004).

[14] A. Matvienko, A, Mandelis, R. J. Jeon, and S. H. Abrams: Theoretical analysis of coupled diffuse-photon-density and thermal-wave field depth profiles photothermally generated in layered turbid dental structures, Journal of Applied Physics, vol. 105, no. 10, pp. 102022, (2009), [15] R. J. Jeon et al. : in vitro detection and quantification of enamel and root caries using infrared photothermal radiometry and modulated luminescence, Journal of Biomedical Optics, vol, 13, no. 3, pp. 034025, (2008).

[16] R. J. Jeon, C. Han, A. Mandelis, V. Sanchez, and S. Abrams: Diagnosis of Pst & Fissure Caries using Frequency-Domain Infrared Photothermal Radiometry and Modulated Laser Luminescence, Caries Research, vol. 38, no. 6, pp. 497-513, (2004).

[17] N. Tabatabaei, A. Mandelis, M. Dehghany, K. H. Michaelian, and Β. T. Amaechi, On the sensitivity of Thermophotonic Lock-In Imaging and Polarized Raman Spectroscopy to early dental caries diagnosis, Journal of Biomedical Optics 17, p. 02502, (2012).

[18] Harvey Rhody, Chester F. Carlson, Geometric Image Transformation, Rochester Institute of Technology, 2005 [19] National Instruments, DAQ M Series, NI USB-621x User Manual, 2009 [20] THORLABS, User guide for GVSxOl and GVSxO2 , Scanning Galvo Systems, 2015

Claims (7)

1) Automated experiment device and image analysis procedure to identify the micro-fissures that appear at the interface between a filling material and the dental tissue, characterized in that it is based on the use of synchronous detection thermography to obtain amplitude and phase images of the investigated areas. which are subjected to a harmonic excitation regime, coupled with the analysis procedure of the resulting images in order to obtain the areas with defects located at the tooth / restoration interface and with the optimization of the laser scanning process on the surface. 2) Automated experimental device for determining the areas of defects located at the tooth / restoration interface according to claim 1), characterized in that it consists of the thermographic chamber with the built-in synchronous detection mode (FLIR 7200 model), dual-axis controllable mirror system electric motor with a galvanized motor (THORLABS GVSM-G02M), Laser excitation source (Nd: YAG) with power P-5GmW and Ă = 532nm, interface for positioning and control of mirrors, National Instruments NI USB-8211, selective voltmeter lock-in SR, Sample and mirror positioning system specially designed for the presented application, Mirror control program, developed on the LabView 2015 platform. 3) An automated mirror control method according to claims 1) and 2), characterized in that the precise control of the laser spot on the surface (pixel by pixel) is realized by implementing the mathematical model of parallax correction. 4) The positioning bracket of the restored dental slice according to claims 1) and 2) placed in the piano of the thermographic chamber and which consists of a fixed body 1 rigidly mounted on the opto-mechanical support by means of two screws 2 and containing a cut-out provided with grooves 3 that serve to attach the movable body 4. 5) Electrically controlled mirror holder according to claims 1) and 2) which consists of an aluminum plate 10mm thick provided with 2 holes with mirror fixing screws (5) and an opening (6) with locking screw (7) ) allowing vertical adjustment. The mirrors are fixed to the vertical rod using the adapter (8). 6) Process for analyzing the amplitude images according to claim 1), characterized in that a threshold factor (a) is introduced which masks the laser spot, the spatial derivative of order 2 (laplacian) is applied in order to accentuate the contrast in the areas otherwise, the resulting image is binarized by introducing a second threshold factor (b) and the noise is filtered by introducing the filter coefficient (c). 7) Process for analyzing the phase images according to claim 1), characterized in that the image is rescaled, the anisotropic diffusion is applied, the spatial derivative of order 2 (laplacian) and the noise filtered using mathematical morphology. to 2017 00298 05/18/2017 Brief presentation of explanatory drawings Figure 1. This figure shows the block diagram of the experimental device described by the present invention. The notations in this figure refer to: 1- sample support 2- laser excitation source 3- signal generator (lock-in voltmeter) 4- infrared camera with synchronous detection mode Built-in 5- computer for data acquisition 6- Interface for positioning and controlling mirrors, National Instruments NI USB6211 7- electrically controlled mirror system (THORLABS GVSM-002M) Figure 2 - Thermal image of the calibration points Figure 3 - Surface of the sample investigated in the visible spectrum Figure 4 - Dynamic scanning of the surface Figure 5a - This figure shows the constructive details of the test stand, in front view (from the thermal imaging chamber) and top view. The notations in this figure refer to: 1- mounting bracket 2- two fixed screws 3- grooves 4- mobile body Figure 5b - This figure shows the constructive details of the mirror support (front and top view). The notations in the figure refer to: 5- 10mm thick aluminum plate 6- screw mounting slot 7- locking screw hole 8- locking screw that allows vertical adjustment 9- adapter fixing the mirrors on the vertical rod Figure 6 - Correction block performance for a set of real coordinates Figure 7a - Second spatial derivative of the amplitude image Figure 7b - Binary image of the dentin / filler interface defect Figure 8a - Phase image Figure 8b - The corrected image after applying the hyperbolic tangent function Figure 8c - The second spatial derivative of the corrected phase image