Real-time imaging method
Patent Information
- Application Number
- EP2023836838
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2022-12-20
- Filing Date
- 2023-12-20
- Publication Date
- 2025-10-29
AI Technical Summary
Current Mueller polarimetric imaging techniques face challenges in achieving real-time imaging and efficient post-processing due to slow acquisition and processing times, especially when dealing with multispectral approaches and the need for simultaneous measurement of multiple wavelengths, which limits their application in biomedical fields like colposcopy and endoscopy.
A multispectral and color polarimetric imaging system utilizing a multi-sensor camera with dichroic prisms and liquid crystal polarization modulators, enabling simultaneous acquisition and processing of multiple spectral bands, and employing artificial series of measurements to increase image frequency and reduce computational burden, allowing for real-time visualization and efficient polarimetric parameter calculation.
This approach enhances the frame rate of polarimetric imaging, facilitating real-time observation and reducing the duration of clinical examinations by increasing the frequency of image acquisition and processing, while maintaining image quality and reducing the need for complex calculations.
Smart Images

Figure 1.1
Abstract
Description
[0001]Description Title: Real-time imaging method Technical field The present invention relates to the field of polarimetric imaging, and more particularly but not exclusively to polarimetric imaging applied to colposcopy. State of the art Colposcopy consists of examining the cervix using a device called a colposcope. The colposcope allows remote observation of the cervix without any contact between the optics or other components of the colposcope and the tissue to be explored. The colposcope includes an illumination system for illuminating the cervix through a speculum introduced into the vagina and for observing the cervix from a distance. Illumination is conventionally carried out using white light. A green filter can be added to the illumination system and used optionally to highlight the areas most absorbed by hemoglobin.The colposcope's optical system can be connected to a color camera that allows color images and videos to be recorded (or monochromatic images using the green filter) on a computer's hard drive. In addition, the image of the cervix can be displayed in real time on the computer screen. In some cases, the colposcope's optical system is also connected to eyepieces that allow the user to directly observe the cervix with, for example, stereoscopic vision.The colposcope may comprise a head mounted on an articulated arm, and may comprise two eyepieces, more precisely a left eyepiece and a right eyepiece, communicating with two respective light input ports present on the front face of the head, oriented towards the area to be examined, used to directly visualize the area of interest, and next to these two input ports, a light output port used to illuminate the area of interest; the light is produced, for example, by a lamp arranged outside the head and conveyed to the latter by a light guide comprising a bundle of silica optical fibers. Furthermore, Mueller polarimetric imaging consists of measuring the Mueller matrix of a target sample and can provide various useful information on the nature of the sample, in particular by making it possible to analyze its optical anisotropy and its light scattering properties.Mueller polarimetry imaging is of interest for the ex vivo and in vivo study of various biological tissues, particularly the cervix. Patent EP 1738 682 describes means for implementing a polarimetric image in colposcopy. Indeed, the knowledge provided by polarimetric information can improve the quality of medical diagnosis for different types of pathologies, particularly for the early detection of cervical cancer. In particular, Mueller polarimetric imaging is a technique that allows the complete polarimetric characterization of a sample by measuring its Mueller matrix. A Mueller polarimeter is generally composed of a light source, a polarization state generator (PSG), a polarization state analyzer (PSA), and a detector. A Mueller polarimeter can provide a point measurement. In this case, a photodetector can be used as a detector.In the case of a point Mueller polarimeter, only one Mueller matrix is measured. However, a Mueller polarimeter can also work advantageously as an imager. In this case, a CCD or CMOS camera is used as a detector. Each pixel of the camera is the equivalent of a photodetector for a point Mueller polarimeter. In the case of an imaging Mueller polarimeter, a Mueller matrix is measured for each pixel. Mueller polarimetric imaging generally requires the acquisition of several intensity images to measure the Mueller matrix of a sample. This technique can be slow, especially if the intensity images needed to obtain the Mueller matrix are acquired consecutively. In the case of an imaging Mueller polarimeter, the light intensity measurement is done simultaneously for all pixels of the camera used. For each pixel, the light signal is transformed into photoelectrons.Among other things, the modality with which the electrical signal produced by the photoelectrons is processed, as well as the number of pixels in the region of interest (ROI) of the selected sensor, determine the number of frames per second (FPS). If several wavelengths must be acquired to explore a biological tissue at different depths, Mueller polarimetric imaging can take even longer if these wavelengths are acquired in succession, which is the case for most multispectral Mueller polarimeters currently in use. However, multispectral analysis is crucial for exploring a biological tissue at different depths. Indeed, the shorter wavelengths of the visible spectrum, corresponding for example to the blue and green colors, are strongly absorbed by hemoglobin and mainly allow exploring the most superficial layers of biological tissues.On the contrary, red / near-infrared light, which is much less absorbed by hemoglobin, allows biological tissues to be explored more deeply. The penetration length of light in biological tissues for wavelengths in the visible / near-infrared spectrum depends on the microscopic absorption and scattering properties of the tissue in question. In general, the penetration length of light in the red / near-infrared part of the electromagnetic spectrum increases with wavelength. Many different types of Mueller polarimeters exist in the literature. The most widespread are time-sequence Mueller polarimeters that successively acquire the intensity images necessary to obtain the Mueller matrix. Mueller polarimetric imaging is done in three main steps: 1) Measurement of the intensity matrix (B = AMW according to the established notation).This matrix can be acquired several times in order to increase the signal to noise ratio with an averaging process, thus improving the reliability of the measurements, which can be called the measurement step. 2) Obtaining the Mueller matrix M=A. -1 BW -1, which can be called the processing step. 3) Determination of the relevant polarimetric parameters from M, which can be called the post-processing step, using for example different algebraic methods, such as Mueller matrix decompositions. For biomedical applications, step 1) should be accomplished in a maximum time of the order of a second to reduce blurring effects due to involuntary patient movements (breathing, heartbeat, etc.) during the measurement. Steps 2) and 3) should ideally be accomplished very quickly to restore images useful to practitioners as soon as possible after the measurement, with a maximum delay of a few seconds. Step 2) can generally be fast and can be accomplished quite easily, especially if the matrices A -1 and W -1have been calculated beforehand. Step 3) can, on the contrary, be quite slow and require high computing power, in particular if it requires, for example, the calculation of the eigenvalues of the matrix M, which can often be the case for decompositions of Mueller matrices or other algebraic treatments. To determine the Mueller matrix of a sample, using a time-sequence Mueller polarimeter, it is necessary to carry out a succession of light intensity measurements, at least 16 in number for a 4x4 Mueller matrix,using a PSG and a PSA to modulate the polarization of the light sent to the sample and to analyze the polarization of the light returned by the sample, respectively. Different combinations of PSG and PSA make it possible to acquire the 16 intensity images necessary to obtain the Mueller matrix of the sample. For 4x4 Mueller polarimetric imaging (complete Mueller polarimetric imaging), at least 16 intensity images must be acquired, which are grouped into the intensity matrix B with 16 real components. The use of polarized cameras allows the acquisition of 12 intensity images, which give access to a 3x4 Mueller matrix. For example, for biological tissues with particular polarimetric properties,It is possible to go back to the 4x4 Mueller matrix with algebraic calculations starting from the measured 3x4 Mueller matrix. However, several factors may currently be limiting for the use of polarized cameras. Polarized cameras are generally monochromatic. Several cameras are therefore necessary to reconstruct a color image, which significantly increases the system footprint. In addition, the images obtained with the different cameras are difficult to superimpose pixel by pixel. In addition, very fine optical adjustments, quite complex to implement, are necessary to obtain images, acquired with the different cameras, which are superimposable pixel by pixel. In addition, with these cameras, each pixel is divided into 4 sub-pixels, two of which correspond to two different linear polarization states. The other two pixels generally correspond to the same linear polarization state between them,which is different from the polarization states corresponding to the first two pixels mentioned above. Pixels divided into four subpixels can produce a loss of resolution in the images. This loss of resolution can be recovered by numerical methods, which can, however, generate artifacts in the images, or by rather complex computational methods, which can considerably increase the image acquisition time. Finally, there can be crosstalk between different subpixels because their polarization states are not completely separated. For biomedical applications, it is necessary that the B matrix be acquired within a maximum time of about one second and that the relevant polarimetric parameters, calculated during the post-processing step,be restored with a maximum delay of a few seconds. The acquisition of the intensity matrix B in about one second is necessary to reduce the blurring effects due to involuntary movements of the patient during the measurement. In one second, the matrix B can be acquired several times to increase the signal-to-noise ratio, with an averaging process. The restitution of the most relevant polarimetric parameters from the Mueller matrices in a few seconds, after post-processing using, for example, the decompositions of the measured Mueller matrices, is thus crucial to allow the practitioner to have useful information very quickly during the visit. The ideal would be to be able to restore the measured Mueller matrix in real time, or even to restore the relevant polarimetric parameters (after post-processing of the Mueller matrix) in real time. However,It is very difficult to obtain an image stream returned to the user with a sufficient frequency for a comfortable observation, substantially in real time by the practitioner during the clinical examination. The first limitation is the acquisition speed of the intensity images needed to obtain the intensity matrix B and therefore the Mueller matrix M. The second limitation is related to obtaining the main polarimetric parameters of M with a post-processing process. Most Mueller matrix decompositions, for example, require the calculation of the eigenvalues of the Mueller matrix M, which represents a computational process that can hardly be parallelized and accomplished in real time. Indeed, in a conventional white light imaging system, for a smooth real-time tracking of an image, the stream should generally be at least 8 FPS if the target moves very slowly. Thus,In the case of Mueller polarimetric imaging, the number of images to be acquired and processed is at least 8*16 = 128 if 16 measurements are to be made between two displayed images, which can be a rather high frequency for high-resolution images. Some polarimetric imaging methods allow for a reduction in the number of measurements but do not allow for the acquisition of images in full Mueller polarimetry (4x4). Others rely on relatively complex specialized sensors such as polarized cameras, which allow for faster acquisition but which may have, at the current state, other important limitations described above. Real-time imaging becomes difficult to accomplish if a multispectral approach is required and if the different wavelengths are acquired in succession. One possibility would be to use several monochromatic CCD or CMOS cameras simultaneously,coupled with spectral filters to select the wavelength ranges of interest, which would make the polarimetric system bulky and difficult to integrate, in a compact manner, on an existing imaging system or within a completely new system to be used in medical practice. Several problems therefore arise in adapting Mueller polarimetry, for example, to colposcopy or other optical imaging techniques for biomedical applications such as endoscopes or even the microscope (or exoscope) for neurosurgery. The realization of Mueller polarimetric imaging in real time is a first challenge. It involves the restitution in real time of the Mueller matrix after the measurement and the processing step, and even better of the most relevant polarimetric parameters after the post-processing step of the Mueller matrix, in order to allow the practitioner to have immediate access,for example during or a few seconds after the visit, to information relating to the microstructure of the tissue. A second challenge is the realization of step 1) above simultaneously for several wavelengths of the visible / near infrared spectrum to allow the practitioner to explore the microstructure of the tissue at different depths. A third challenge is to also provide the reference color imaging in real time, which is necessary to clearly identify the area analyzed with polarimetry, thus allowing the practitioner to have well-known spatial landmarks, which would not be possible with monochromatic light intensity images or by directly displaying the polarimetric images. Another challenge is the possibility of being able to superimpose pixel by pixel different types of image, such as for example, a color image,light intensity images at different selected wavelength ranges and polarimetric images at different selected wavelength ranges, with all possible combinations, which is crucial for the efficient analysis of the images using, for example, image processing algorithms or learning algorithms. This step can allow, among other things, to: 1) select the most relevant polarimetric and non-polarimetric parameters; 2) determine the most relevant combination of polarimetric and non-polarimetric parameters for diagnosis; 3) merge together, using, for example, different types of image processing algorithms,polarimetric images (e.g. Mueller matrix elements and / or polarimetric parameters obtained after using algebraic processing such as Mueller matrix decompositions) and non-polarimetric images (e.g. color image and / or monochromatic intensity images) of interest at different wavelengths in a single image or in a limited number of images to provide practitioners with simplified and optimized information for diagnosis that is not directly observable in conventional images or in the initial unfused polarimetric images; 4) merge together the polarimetric images obtained for different wavelength ranges to obtain color polarimetric images. Another challenge may be the restitution of the combined images substantially in real time. Finally, a last challenge is to have a very compact and easily adaptable Mueller polarimeter,ergonomically, to different existing imaging systems, such as, for example, a colposcope or an endoscope or a microscope (or an exoscope) for neurosurgery, or easily usable for the realization of a new ergonomic and space-saving imaging system to be used in medical practice according to the intended application. Disclosure of the invention Consequently, there is a need to benefit from high-performance polarimetric imaging systems, in particular polarimetric colposcopes, in order to allow multispectral and / or color, as well as comfortable multispectral and / or color polarimetric visualization of the area examined, in particular in real time. More generally, there is a need for a solution to increase the frame rate for any imaging technique such as Mueller polarimetry, requiring a large number of measurements to be carried out before an image can be generated,so as to make the viewing of images by the practitioner more fluid during the examination, thus facilitating decision-making, and / or reducing the duration of the examination, among other things. Furthermore, independently of the above, there is also an interest in: - facilitating the post-processing of the Mueller matrix in order to reduce the need for computing resources, and reduce the time required for calculating polarimetric images, - facilitating the production of an imaging system, in particular colposcopy, capable of both performing multispectral and / or color imaging, as well as multispectral and / or color polarimetric imaging, and in particular allowing the easy transformation of a conventional observation system such as a conventional colposcope into an observation system allowing multispectral and / or color imaging, as well as multispectral and / or color polarimetric imaging, - improving the illumination system of a colposcope,in particular with a view to improving the quality of the images, - generally improving the ergonomics of an observation system such as a colposcopy system, and / or - facilitating the development of the system, in particular for carrying out calibration. The invention aims to meet all or part of the needs identified above. Summary of the invention Multispectral and / or color polarimetric imaging According to a first of its aspects, the subject of the invention is a multispectral and / or color polarimetric imaging system, in particular polarimetric colposcopy, comprising: - an illumination system comprising at least one light source, this illumination system emitting in at least two, better still at least three, spectral bands, - a polarization state generator (PSG) arranged downstream of the light source and upstream of a target to be imaged, - a polarization state analyzer (PSA) arranged downstream of the target to be imaged,- a multi-sensor camera comprising at least two, better at least three, sensors, for recording respectively at least two, better at least three, images in said spectral bands. The polarimetric imaging system can be wide-field (macroscopic) or microscopic, and work in transmission or reflection, in free space or not. It is also possible to have an endoscope between the target and the PSA and a liquid guide between the PSG and the target, as described in EP2021 / 052647. Preferably, it is wide-field and works in reflection, in particular in free space (such as a colposcope). The imaging system can also be used for other biomedical applications, for example brain surgery or endoscopy, and non-biomedical applications,for example in the field of cosmetics or microelectronics. The use of a light source emitting in several spectral bands in combination with the use of a multi-sensor camera makes it possible to generate images in different wavelength ranges that are perfectly superimposable, since by construction, the different sensors of the camera produce perfectly superimposable images, that is to say that for all the pixels of the image, the same point on the image corresponds to the same pixel on each of the sensors. In particular, it is possible, among other things, to limit parallax effects between the different types of images produced by the camera, for example between color images (conventional RGB images) and polarimetric images,since the same camera is used to produce the color images and the polarimetric images and the system produces images of the same size on each of the sensors and positioned identically on each of the sensors, thus allowing their pixel-to-pixel superimposability. Parallax effects could also be limited by combining several cameras, however the alignment of several cameras can be complicated and time-consuming, the solution proposed previously makes it possible to limit parallax effects more easily and thanks to the use of a single camera. Preferably, the spectral bands are three in number and go, for example, from 445 nm to 475 nm for the first, from 510 nm to 550 nm for the second, and from 600 nm to 660 nm for the third, being for example centered respectively on 460 nm, 530 nm and 630 nm. Preferably, the camera is a bi-CCD, bi-CMOS, tri-CCD, tri-CMOS, 4-CMOS or 4-CCD camera, preferably tri-CCD, tri-CMOS,4-CCD or 4-CMOS. Also preferably, the spectral bands are three in number and respectively in the red or near infrared, green and blue. For example, for a 2-CCD camera, the spectral bands are preferably visible and near infrared, for a 3-CCD camera, the spectral bands are blue, green, and red / near infrared, and for a 4-CCD camera, the spectral bands are blue, green, red and near infrared. This allows the use of, for example, high-resolution, high-speed cameras manufactured on a large scale. A multi-sensor camera provides an independent grayscale intensity image for each sensor,these grayscale images can be used on the one hand to measure the coefficients of the intensity matrix B in the case of Mueller polarimetry and on the other hand they can be combined to reconstruct a color image of the observed area. The camera can comprise at least two, better three, dichroic prisms, to separate the wavelengths towards the different sensors. A tri-CCD camera, for example, comprises three dichroic prisms, to separate the wavelengths towards the different sensors. The sensor for detecting blue can be fixed on the first prism, the sensor for detecting red can be fixed on the second prism, itself can be fixed on the first prism, and the sensor for detecting green can be fixed on the third prism, itself fixed to the second prism, in a manner known per se. Other arrangements are possible, this type of camera being well known. Preferably,the polarization state generator (PSG) comprises electrically controllable liquid crystal polarization modulators, which allows for relatively fast switching, good compactness, suitable bandwidth and does not disturb the superposability of the images. Liquid crystals also allow the implementation of full-field polarimetric imaging for the analysis of surfaces on a macroscopic scale as well as microscopic polarimetric imaging. Alternatively, the polarization state generator comprises, for example, rotating wave plates, as well as rotating polarizers and / or coupled systems of rotating wave plates and rotating polarizers. Alternatively, the polarization state generator may comprise systems coupling rotating wave plates and fixed polarizers,or fixed wave plates and rotating polarizers or rotating delay plates and rotating polarizers. The polarization state generator can also be realized using spatially separated polarization systems (polarizers, delay plates, etc.), which can produce different polarization states of light. The polarization state generator can also be realized using spatially separated polarization systems (polarizers, delay plates, etc.), which can produce different polarization states of light. The spatially separated polarization systems can also be fixed or modulated in time with a liquid crystal system or rotating systems. In this configuration, the light beam can be deflected using fast switching systems to pass consecutively through the different polarization systems before illuminating the target. Similarly,the polarization state analyzer preferably comprises electrically controllable liquid crystal depolarization modulators. Alternatively, the polarization state analyzer comprises, for example, rotating wave plates, as well as rotating polarizers or systems coupling wave plates and polarizers in all the combinations described above for the polarization state generator. Preferably, the liquid crystal polarization modulators are ferroelectric liquid crystal modulators, which allows a high switching frequency, with a control voltage of a few volts. Alternatively, the liquid crystal polarization modulators are nematic liquid crystal modulators. The polarization state analyzer can also be implemented using spatially separated polarization systems (polarizers, delay plates, etc.),which allow the production of different polarization states of light. The polarization systems, just like for the polarization state generator, are spatially separated and can also be fixed or modulated in time with a liquid crystal system or rotating systems. In this configuration, the light beam can be deflected using fast switching systems to pass consecutively through the different polarization systems before illuminating the target, in particular before reaching the detector. The light source is preferably a white light source, in particular a xenon lamp. Such a lamp makes it possible to emit in a broad spectrum and to produce both good multispectral and / or color images, as well as multispectral and / or color polarimetric images. Alternatively, it is a halogen type lamp, at least one LED or a source of any other type,incoherent or coherent (laser) When using a xenon lamp, the illumination system may include a single-band, dual-band, or better still, a tri-band dichroic filter, downstream of the light source placed before the target to be analyzed or placed after the target to be analyzed and before the detector. This filter may block UV and IR, for example, and allow white light to pass, including the red, green, and blue bands of the visible spectrum. This filter may be removable, particularly mounted on a filter wheel, to allow, for example, acquisition in the IR or near IR, or even the near UV. Alternatively, depending on the type of source used, the illumination system may include a dual-band, better still, a tri-band, or at least a single-band filter. A quad-band filter may be used with a 4-CMOS or 4-CCD camera, for example, to allow simultaneous acquisition of images in the blue, green,red and near infrared. It is also possible to use no spectral filter. For example, the illumination system may comprise several light sources emitting respectively in the spectral bands of interest, for example blue, green and red, or blue and yellow LEDs. The imaging system may comprise a polarimetric system according to another aspect of the invention, as defined below. The imaging system may comprise a filter wheel arranged directly upstream of the camera, as well as possibly a linear polarizing filter of the analyzer, as described below. Alternatively, the filter wheel is arranged directly downstream of the source. The filter wheel may carry a tri-band filter, as mentioned above, to allow for example the acquisition of images at wavelengths in spectral ranges centered around 460nm, 530nm and 630nm by the camera,as well as one or more monochromatic dichroic filters, to allow for example the acquisition of images in spectral ranges around 650nm and 700nm. The sensor intended to image red can also allow imaging of the near infrared. The filter wheel can allow the selection of different spectral ranges in blue, green and red / near infrared by changing the tri-band filter, removing the filter, replacing the type of filter used, for example replacing a tri-band filter with at least one monochromatic filter. The filter wheel can allow certain wavelength ranges to be explored in more detail. The imaging system can comprise a processor for generating a Mueller polarimetric image of the target in each of the spectral bands. A Mueller polarimetric image is understood to mean an image comprising at least one piece of information extracted from the Mueller matrix,for example a polarimetric property such as diattenuation, polarization, birefringence, depolarization, etc., as well as any information at least partially deduced from the knowledge of one of these properties. In the case of colposcopy for the analysis of the uterine cervix, knowledge of the birefringence and depolarization properties are particularly useful for diagnosis. In the case of use in colposcopy, polarimetric parameters particularly useful for diagnosis may include linear phase retardation, the degree of randomness of the azimuth of the slow (or fast) axis of the linear phase retardation, the combined image of the linear phase retardation and the azimuth of the slow (or fast) axis of the linear phase retardation, depolarization, the linear correlation between depolarization and linear phase retardation, in particular at 530 nm. The processor may be configured to at least partially superimpose,to an image corresponding to the non-polarimetric observation by the camera (for example the basic color image), an image containing at least one polarimetric information. This polarimetric information can be displayed in false color. The system can be arranged to carry out, in particular with each of the sensors of the camera and in parallel for all the different sensors, successive real series of measurements on a given sample corresponding to the acquisition, for example, of at least 9 (Mueller 3x3 polarimetry), in particular 12 (Mueller 3x4 polarimetry), better 16 (Mueller 4x4 polarimetry) coefficients of the intensity matrix B, the processor being arranged to: - generate series of n additional measurements, also called “artificial” series, from a grouping of measurements coming from a series μ of n measurements of a given rank, and from at least one series of a different rank, in particular of the following rank,and a permutation P(μ) of the measurements within each artificial series so as to respect the predefined order of the n measurements within each series, then - generating a stream of polarimetric images at a frequency higher than that which would be allowed by said real series of measurements without the series of artificial measurements, from the polarimetric images produced from the real series of measurements and the polarimetric images generated from the artificial series interposed between the real series. The processor can be arranged to carry out a decomposition of the Mueller matrix into a non-depolarizing component and a depolarizing component. The depolarizing component and the non-depolarizing component can have several forms. The processor can thus be arranged to carry out an additive decomposition of the Mueller matrix M in the form ^ = ^^, ^^ + ^^ ^ where ^ ^^ is the non-depolarizing component and ^^ is the depolarizing component. The parameters ^ and ^ are the weights of the components ^ ^^ and ^ ^ of the Mueller matrix M, respectively. According to this model, the measured intensity matrix ^ can be written as the sum of a polarizing contribution and a non-depolarizing contribution. ^= ^^^ + ^^ = ^^^^^^ + ^^^)^Then ^= ^^^^ + ^^^ = ^ ^^ ^^^^ + ^^)^ ^^ For a 4*4 Mueller matrix, we have B=AMW, où and W = ^S ^^ , S ^^ , S ^^ , S ^^ ^ the measurements being obtained using the PSG polarization state generator which produces four independent states described by four Stokes vectors S ^^ ^ i = 1,2,3,4 ) and the PSA polarization state analyzer which generates analysis states described by four Stokes vectors S ^^^i = 1,2,3,4). This increases the frame rate, which improves the comfort of viewing the images, and allows real-time polarimetric imaging, for example with a stream of at least 8 frames per second of polarimetric images. Thus, when the practitioner changes the observed area, he quickly benefits from the corresponding polarimetric image. Increasing the frame rate makes it possible to artificially increase the number of frames per second without changing the exposure time required for acquisition, and therefore offers the possibility of obtaining images in real time. The rate of 8 frames per second is sufficient if the observed object is static or moves slowly. If the object moves more quickly, the required frame rate may be higher.The processor can be arranged to, before performing the permutation P, perform a spatial registration of the images corresponding to the measurements in order to take into account possible displacements of the areas observed in the field of vision of the camera. This limits the risk of blurring in the polarimetric images, and increases the precision of the polarimetric information. The processor can be arranged to perform a decomposition of the Mueller matrix into a non-depolarizing component and a depolarizing component. The depolarizing component and the non-depolarizing component can have several forms. The processor can thus be arranged to perform an additive decomposition of the Mueller matrix M in the form M = qMnd+pMd where Mnd is the non-depolarizing component and Md is the depolarizing component. The parameters q and p are the weights of the two components of the Mueller matrix. In particular M. ndcan be the Mueller matrix of a given linear phase retarder par : M ^^ = ^&'(^), *) and M8 can be the Mueller matrix of a pure depolarizer given by: 1 0 0 0 ^ ^ = + 0 0 0 0 0 0 0 0 7 0 0 0 0 ) is the azimuthal orientation of the fast (or slow) axis in degrees and * the linear phase retardation in degrees (ranging from 0° to 180°) with The Mueller matrix M can be normalized with respect to its unnormalized and unpolarized intensity coefficient m̂ ^^ = ^ + ^ and can be written in the form: ^2.F=GH1 / éJ =^^^^ + ^^^ ^^^ + B^^^ + ^ = 1 + B Or The matrix ^2.F=GH1 / éJ can then be written in the form: the coefficients mij (i,j=1,2,3,4) being those of the normalized Mueller matrix M, m̂ ^^ being the unnormalized and unpolarized intensity coefficient of M, the other coefficients mij (i,j = 1,2,3,4) being normalized with respect to m̂ ^^. The processor may be arranged to calculate the depolarization by performing the following calculation: D epolarization The system can be arranged to generate the parallel display of at least one non-polarimetric image in at least one of the spectral bands, in particular the color image of the observed area, and at least one polarimetric image. The invention also relates to a method for training an artificial intelligence system, for example comprising at least one convolutional neural network, in which the artificial intelligence system can receive as input non-polarimetric color and / or multispectral images, as well as polarimetric color and / or multispectral images. The fact that these images come from a multi-sensor camera avoids additional spatial registration operations of the images at different wavelengths acquired simultaneously by different sensors, and reduces the computing time, which leaves more resources available for training as such.The images obtained by the different sensors can be superimposed pixel by pixel. Increase in the image frequency According to another of its aspects, preferably taken in combination with the above, but which can be applied generally to other types of imaging, the invention relates to a real-time imaging method, comprising the steps of: - Acquiring, per given time interval, original real series μ of n measurements. following one another in a predefined order within the series, each series of n measurements allowing, by processing these measurements, to generate at least one original result X= G(μ), at a given frequency f, G being a function giving the result from the measurements, - generating artificial series from a grouping of n measurements coming from a real series μ of a given rank, and from at least one real series of a different rank, in particular the following rank, and from a permutation of the measurements within each artificial series so as to respect the predefined order of the n measurements within each series, - generating a flow of results at a frequency greater than f from the original results and from the results generated from the artificial series intercalated between the original real series.This imaging method is advantageously applied to polarimetric imaging, for example to polarimetric colposcopy, and better, to multispectral and / or color polarimetric colposcopy as defined above. This method can be applied to colposcopy, microscopy, brain exoscopy, endoscopy, preferably to polarimetric colposcopy, and better, to multispectral polarimetric colposcopy. The result is then a Mueller matrix or a physical property calculated from this matrix. The series of measurements are then polarimetric measurements. The invention makes it possible, according to this aspect, to artificially increase the flow of images and to enable, or improve, real-time imaging despite the number of measurements to be carried out and the calculations necessary to generate the desired images.The invention, according to this aspect, in other words makes it possible to artificially increase the number of images per second, while maintaining a fixed number of acquisitions. By "real time" is meant a delay between acquisition and visualization that is relatively small, for example 2 or 3 seconds, and compatible with visualization of the information generated by the practitioner during the examination; for example, the delay between the moment when the measurements are carried out and the moment when the information resulting from these measurements is displayed is less than or equal to 2s, better than 1s. The flow of results generated can be greater than 8 per second. The imaging to which this method is applied is advantageously polarimetric imaging and the series of n measurements can then correspond to the measurements of the coefficients of the intensity matrix B used to calculate the Mueller matrix.In particular, the measurement series may correspond to the acquisition, for example, of at least 9, notably 12, better 16, coefficients of the intensity matrix B. As indicated above, for a 4*4 Mueller matrix, with B=AMW,. où and W = ^S ^^ , S ^^ , S ^^ , S ^^ ^ the measurements are obtained using a PSG polarization state generator which produces four independent states described by four Stokes vectors S ^^ ^i = 1,2,3,4) and a PSA polarization state analyzer that generates analysis states described by four Stokes vectors S ^^ ^ i = 1,2,3,4 ) . According to this model, the measured intensity matrix ^ can be written as the sum of a polarizing contribution and a non-depolarizing contribution. ^= ^^^ + ^^ = ^^^^^^ + ^^^) ^ Preferably, as mentioned above, before performing the permutation, a spatial registration of the images corresponding to the different intensity measurements that represent the components of the matrix B (obtained for different configurations of the PSG and the PSA) is carried out in order to take into account possible displacements of the areas observed in the field of view of the camera. The method of increasing the frame rate is general and can be applied for example to systems using multi-sensor CCD and CMOS cameras, as well as monochromatic or polarized CCD and CMOS cameras. Post-processing of the Mueller matrix To accelerate the calculation of the Mueller matrix, one can implement for example a particular post-processing which avoids proceeding via a so-called Lu-Chipman decomposition but nevertheless allows to obtain a close result.The invention thus also relates, according to another of its aspects, independently or in combination with the above, to a polarimetric imaging method in which intensity images are acquired using a polarimetric imaging system (preferably that defined above) comprising an illumination system, a polarization state generator (PSG) placed on the light path between the illumination system and the area to be observed, a polarization state analyzer (PSA) placed on the light path between the area to be observed and at least one image acquisition system, then an additive decomposition of the Mueller matrix M is carried out in the form M = qMnd+pMd where Mnd is the non-depolarizing component and Md is the depolarizing component. In particular M. nd can be the Mueller matrix of a linear phase retarder given by:M ^^ = ^&'(^), *) and ^ ^can be the Mueller matrix of a pure depolarizer given by: ) is the azimuthal orientation of the fast axis in degrees and * the phase of the retardance in degrees (ranging from 0° to 180°) with The parameters q and p are the weights of the two components of the Mueller matrix. The Mueller matrix M can be normalized with respect to its unnormalized and unpolarized intensity coefficient m̂ ^^ = ^ + ^ and can be written in the form: where B = ^ ^ The matrix ^2.F=GH1 / éJ can then be written in the form: The coefficients mij (i,j=1,2,3,4) are those of the normalized Mueller matrix M, m̂ ^^ being the unnormalized and unpolarized intensity coefficient of M, the other coefficients mij (i,j = 1,2,3,4) being normalized with respect to m̂ ^^ The processor may be arranged to calculate the depolarization by performing, for example, the following calculation: D epolarization Such an additive decomposition of the Mueller matrix allows to parallelize the calculations and to extract useful parameters more quickly. The Mueller matrix and the relevant polarimetric parameters can be obtained for a single wavelength range or simultaneously for several wavelength ranges, including at least two, in particular in the visible and infrared. Such a method thus makes it possible to reconstruct the image, in particular the color image, in real time from the combination of the unnormalized and unpolarized intensity coefficientsm̂^^, of the three Mueller matrices obtained simultaneously in the spectral rangescorresponding to the blue, green and red part of the visible spectrum. Such a method makes it possible to avoid calculating the eigenvalues of the Mueller matrix, which is computationally expensive,and allows the extraction of the desired parameters with much shorter calculations. This method is based on the assumption that in certain tissues, such as the uterine cervix, the non-depolarizing effects are rather linked to the most superficial layers of the tissue while those of depolarization are linked to the volume of the tissue. Observation system with two input ports equipped with a polarimetry system According to another of its aspects, independently or in combination with the above, the invention also relates to an observation system with two input ports, in particular a binocular system, for example for colposcopy, comprising a head comprising an optical system having a light output port for illuminating an area to be examined and left and right input ports directed towards the area to be observed,the observation system further comprising a polarimetric system comprising a polarization state generator arranged in front of the output port and a polarization state analyzer of which at least some of the optical elements are arranged in front of one of the input ports, this analyzer comprising at least one optical element held by at least one support applied to only part of the contour of the optical element. This can make it possible to provide a free edge for the optical element, this free edge partially overlapping the other input port. The optical element can only partially overlap this other input port, without having this input port hidden by the support which would hinder observation. The optical element is, for example, a liquid crystal polarization modulator. This aspect of the invention takes advantage of the fact that the free edge of the optical element, which is preferably of circular contour,although partially superimposed on the input port, remains substantially transparent in the eyepiece associated with this input port due to the focusing distance, which is typically of the order of 20 to 40 cm, therefore much greater than the distance separating the input port from the optical element, which is at most a few cm. In addition, since the optical elements of the analyzer are relatively transparent, except possibly for the linear polarizer which can be placed in front of the camera downstream of the prism for reflecting towards the eyepiece, these elements do not cause any significant loss of brightness for the observer. It is thus possible to use a conventional colposcope head or a head of another observation system with two input ports, in particular a binocular observation system, by attaching all or part of the polarimetric system to it on the front face,which limits the manufacturing costs of the system and makes it easy to add a polarimetric imaging function to the observation system. Each support of an optical element of the analyzer may extend, for example, over an angular extent of between 180° and 300° in contact with the optical element, preferably having a general C shape open in a downward oblique direction, substantially at 45°. The polarimetric system may also comprise at least one other support for holding at least one optical element of the polarization state generator placed in front of the light output port of the colposcope head. This other support may have a general C shape open upwards,which limits the size of the state generator in the vertical direction and makes it possible not to encroach on the input ports present near the output port on the front face of the colposcope head. Each support may comprise a succession of support pieces holding between them optical elements of the polarization state generator (or of the analyzer), and two supports used to hold two consecutive optical elements may share an intermediate support piece. The polarimetric system may thus comprise at least two consecutive support pieces assembled against each other, each of these support pieces having on one face a housing for receiving a respective optical element and one of the support pieces serves to retain the optical element of the other piece. It is thus possible to have a relatively compact polarimetric system,which does not unduly hinder the manipulation of the head of the observation system by the practitioner. The polarization state generator and the analyzer preferably each comprise similar optical elements which are placed in reverse orders relative to the direction of propagation of the light. The generator thus comprises, for example, in the direction of propagation of the light, a linear polarizing filter, a QFLC quarter-wave liquid crystal polarization modulator, a QWP half-wave plate, and a HFLC half-wave liquid crystal polarization modulator. The analyzer may then comprise, in the direction of propagation of the light, a HFLC half-wave liquid crystal polarization modulator, a QWP half-wave plate, and a QFLC quarter-wave liquid crystal polarization modulator. The polarizer associated with the analyzer is preferably arranged after returning the light from the input port to the associated eyepiece,upstream of the camera used to collect the intensity images used to generate the Mueller matrix, which makes it possible to limit the difference in brightness between the images delivered to the observer by the left and right eyepieces, as mentioned above. The housings receiving the electrically controllable optical elements may include at least one passage for an electric cable. This passage may include a channel which follows the shape of the contour of the element, in particular a semi-circular channel. The C-shaped supports of the optical elements allow, before tightening, manual rotation of the optical elements around their axis,in order to carry out orientation adjustments. The optical elements of the generator are preferably oriented perpendicular to the direction of propagation of the light leaving the head of the observation system. When this light is emitted by a prism in a direction forming an angle with the direction of observation of the input ports, the mounting of the optical elements of the generator is preferably carried out at an angle relative to those of the analyzer, so as to orient them perpendicular to the axis of propagation of the light leaving the output port. This makes it possible to limit parasitic reflections. Preferably, the polarimetric system comprises a fan configured to blow air onto the optical element of the polarization state generator closest to the light output port of the head of the observation system. This ensures better and faster stabilization of the temperature,and the corresponding drifts are limited. The polarimetric system may include a movable filter holder in front of the analyzer, allowing the system to be calibrated. This filter holder is, for example, fixed to one of the support parts used to hold the optical elements. The filter holder may be movable in a horizontal direction, generally perpendicular to the direction of observation. The filter holder may, for example, take four positions, three of which are for placing a predefined optical element in front of the analyzer, and the last one to free the field. The movement of the filter holder is preferably motorized, and controlled by the aforementioned controller. The system may include a housing protecting the supports and the filter holder. The invention also relates to a polarimetry system intended to be attached to a head of an observation system, in particular a head of a colposcope,comprising an optical system having a light output port for illuminating an area to be examined and left and right input ports directed towards the area to be observed, the polarimetric system comprising a polarization state generator placed in front of the output port and a polarization state analyzer of which at least some of the optical elements are placed in front of one of the input ports, this analyzer comprising at least one optical element held by at least one support applied to only part of the contour of the optical element. This may make it possible to provide a free edge for this optical element, this free edge being arranged so as to partially overlap the other input port. The polarimetry system may comprise means for fixing to the head of the observation system. These fixing means may comprise screws, which are screwed into the frame of the head, for example under the output window. Alternatively,the fixing is carried out otherwise, for example, by clamping. The polarimetry system may have all or part of the characteristics given above. Compact and modular polarimetry system The invention also relates to a polarimetry system intended to be attached to a head of an observation system, in particular a colposcope head, comprising an optical system having a light output port for illuminating an area to be examined and at least one input port, the polarimetric system comprising a polarization state generator (PSG) placed in front of the output port and a polarization state analyzer (PSA) of which at least some of the optical elements are placed in front of the input port, at least one of the analyzer and the generator comprising optical elements held by at least one support,each support comprising a succession of parts holding between them optical elements of the polarization state generator or of the analyzer, at least two supports serving to hold two consecutive optical elements sharing an intermediate support part, each of these parts preferably having on one face a housing for receiving a respective optical element and one of the parts serving to retain the optical element of the other part. Such a polarimetry system is particularly compact, because it comprises support parts which serve as a housing for receiving an optical component and as a closing cover for the next support part. Such a system is also modular, because it is easy to replace support parts with others. It can have any of the characteristics of the observation system, in particular of the colposcope,described elsewhere. The polarimetry system may in particular comprise an analyzer comprising at least one optical element held by at least one support applied to only part of the contour of the optical element so as to provide a free edge for this optical element, this free edge being arranged so as to partially overlap the other input port, the polarimetry system comprising means for fixing to the head of the observation system described elsewhere. Illumination system for colposcope According to another of its aspects, independently or in combination with the above, the invention also relates to a colposcopy system comprising: - an illumination system comprising o at least one light source, o a liquid light guide, to be connected at one end to the light source, this light guide preferably having a core diameter less than or equal to 5 mm, - a colposcope head,comprising an optical system having an input port to be connected to the other end of the liquid light guide, and a light output port for illuminating an area to be observed, the optical system preferably comprising between the input port and the output port an aspherical lens and a deflecting prism. This configuration of the illumination system makes it possible to obtain a well-collimated light beam at a working distance preferably of approximately 30 cm. The choice for the light guide of a liquid guide and for the core diameter of a diameter preferably less than or equal to 5 mm, makes it possible to reduce the divergence of the illumination beam and its size, and to increase the intensity of the light at the center of the area to be observed, i.e. the cervix, and thus to reduce parasitic reflections on the surrounding surfaces (vaginal wall, speculum, etc.),and consequently significantly improve the quality of the acquired images. The liquid light guide makes it possible to obtain illumination with satisfactory uniformity, in particular a uniformity greater than that obtained with the silica fiber bundles generally used in colposcope illumination systems. Preferably, the inner diameter of the liquid optical guide is between 2.5 and 3.5 mm. The illumination system may comprise an adapter configured to receive the optical guide and to be mounted on the source with the possibility of adjustment in the three directions X, Y and Z relative to the source. Ergonomic colposcopy system According to another of its aspects, independently or in combination with the above, the invention also relates to a colposcopy system comprising: - a rolling base, - a mast carried by the rolling base, - at least one reference reflector carried by the mast,used for calibrating a polarimetry system, - a colposcope head carried by an articulated arm connected to the base. The colposcopy system may also include: - a workstation comprising a computer, carried by the rolling base, - a keyboard carried by an articulated arm connected to the mast, and - a screen carried by an articulated arm connected to the mast at a height greater than that of the arm supporting the keyboard. Such an arrangement makes it possible to keep the screen and keyboard close to the user, who can thus easily control the computer himself for image acquisition. The mobility of the system is also improved, since the assembly can be easily moved as a single unit on the floor. Preferably, the system includes a housing housing the light source(s), placed on the workstation,as well as a controller which allows to control the various electro-optical and electronic components of the polarimetric system such as the controllable liquid crystal filters, the filter wheel, and to process the images coming from the camera, if necessary, or even to generate the polarimetric images, etc. The reference reflector can be arranged in a light shield, at the top of the mast. This reflector is for example articulated around a vertical axis, and comprises for example on one face a frosted metal wall, for example aluminum, and on the opposite side a reference surface having known spectral properties. The rotating mounting of the reflector allows to easily replace one of the faces by the other without moving the head of the colposcope, which facilitates the calibration operations. During these operations, the head is for example placed approximately 30cm from the reflector,by maneuvering the articulated arm carrying the head. The colposcopy system advantageously comprises a pedal which allows the user to trigger a predefined action, for example to start a polarimetric acquisition. The head of the colposcope may also comprise a button allowing another predefined action to be triggered, for example the start and stop of video recording. Description of the figures The invention may be better understood by reading the detailed description which follows, of a non-limiting example of implementation of the different aspects of the invention, and by examining the appended drawing, in which: [Fig 1] Figure 1 is a block diagram of an example of a polarimetric multispectral imaging system according to the invention, [Fig 2] Figure 2 is a view similar to Figure 1 illustrating the possibility of delivering an image facilitating diagnosis,[Fig 3] Figure 3 illustrates the possibility of using artificial intelligence to generate diagnostic assistance, [Fig 4] Figure 4 illustrates different modes of displaying information in a colposcopy system according to the invention, [Fig 5] Figure 5 is a diagram illustrating the acquisition of successive series of measurements, [Fig 6] Figure 6 illustrates the generation of artificial series from the series of the example of Figure 5, [Fig 7] Figure 7 illustrates the reordering of measurements within the artificial series of Figure 6, [Fig 8] Figure 8 illustrates the spatial registration of areas of the image corresponding to successive measurements, [Fig 9] Figure 9 illustrates the application of the method to measurements carried out in the context of Mueller polarimetry, [Fig 10] Figure 10 illustrates the generation of artificial series of measurements in the case of the example of Figure 9,[Fig 11] Figure 11 illustrates the assumptions used for the accelerated calculation of the parameters of interest, [Fig 12] Figure 12 represents comparative images obtained by implementing the Lu-Chipman decomposition on the one hand, and the additive decomposition used for the accelerated calculation of the polarimetric parameters on the other hand, [Fig 13] Figure 13 is a schematic and partial view of a colposcopy system according to the invention, [Fig 14] Figure 14 represents more particularly the head of the colposcope and the polarimetry and acquisition systems, [Fig 15] Figure 15 represents more particularly the lower part of the colposcopy system, [Fig 16] Figure 16 illustrates a detail of the system at the level of the head of the colposcope, [Fig 17] Figure 17 is a top view of the colposcopy system, arranged near an examination table, [Fig 18] Figure 18 is a view in partial elevation of the colposcopy system,[Fig 19] Figure 19 represents the head of the colposcope with part of the polarimetry system fixed on its front face, [Fig 20] Figure 20 is an exploded view showing different optical elements of the polarization state generator and the analyzer and their support parts, [Fig 21] Figure 21 illustrates the assembly of the parts supporting the optical elements of the polarization state generator, [Fig 22] Figure 22 illustrates the assembly of the parts supporting the optical elements of the polarization state analyzer, [Fig 23] Figure 23 represents in isolation one of the support parts, used for fixing the filter holder used for calibration, [Fig 24] Figure 24 illustrates the assembly of two support parts of the polarization state generator, [Fig 25] Figure 25 represents in front view one of the support parts of the generator and the received optical element in the accommodation of this room,[Fig 26] Figure 26 is a front view of the front face of the colposcope head equipped with the polarimetric system, [Fig 27] Figure 27 is a partial and schematic sectional view of the illumination system, [Fig 28] Figure 28 is a schematic and partial perspective view with axial section of the illumination system of Figure 27, [Fig 29] Figure 29 is a partial and schematic longitudinal section of the colposcope head at the connection of the light guide, [Fig 30] Figure 30 is an exploded perspective view of elements used for mounting the light guide to the colposcope head, [Fig 31] Figure 31 illustrates the spatial distribution of the light intensity at the level of the observed zone as a function of the diameter of the active part of the light guide used, [Fig 32] Figure 32 represents the calibration reflector and its cap, [Fig 33] the Figure 33 shows the reflector in isolation, without its cover,[Fig 34] Figure 34 schematically and partially represents the calibration filter holder, [Fig 35] Figure 35 represents the part of the polarimetry system which is fixed on the front face of the colposcope head, equipped with its protective cover, [Fig 36] Figure 36 is a schematic and partial side view illustrating the mounting of the acquisition system on the colposcope head, [Fig 37] Figure 37 represents the filter wheel of the acquisition system and its drive motor, [Fig 38] Figure 38 partially and schematically represents the acquisition system from another viewing angle, and [Fig 39] Figure 39 schematically and partially represents an example of a multi-sensor camera optical system. Detailed description Multispectral and / or color polarimetric imaging system Figure 1 shows a multispectral polarimetric imaging system 1 according to the invention, for example a Mueller polarimetric colposcope,it being understood that this aspect of the invention is not limited to a colposcope. The system 1 comprises an optical illumination system 2 for illuminating a target T, for example the uterine cervix, and an optical system 3 for direct observation of the target and / or reconstruction of the image of the target T on an electronic medium. The illumination and observation of the target T are carried out through a polarimetry system 4 which comprises a polarization state generator (PSG) crossed by the light coming from the illumination system and a polarization state analyzer (PSA) crossed by the light coming from the target T. Preferably, the PSG and the PSA comprise liquid crystal polarization modulators,in particular chosen from those whose control frequency can be at least 60 Hz. Ferroelectric liquid crystal polarization modulators are preferred. The system 1 also comprises an acquisition system comprising a multi-sensor camera 5, constituted for example by a tri-CCD or tri-CMOS camera, having for example sensors dedicated respectively to the red, green and blue domain of the visible spectrum. The system 1 comprises computer means 6 for, among other things, managing the acquisition of intensity images for each sensor, controlling the operation of the liquid crystal polarization modulators, synchronizing the camera acquisitions with the modulation of the liquid crystals, processing the images obtained by each of the sensors of the camera 5 and calibrating the system for each of the chosen spectral bands, as will be detailed later. These computer means 6 comprise for example one or more processors,one or more microcontrollers, specialized circuits such as FPGAs, or microcomputers, and the associated human-machine hardware interfaces, and can be programmed to perform the acquisition of intensity images in the different spectral bands of interest, for example red, green and blue in the example considered, then the calculation, for all or part of the pixels of the images, of the Mueller matrices in these different spectral bands, and the calculation and display of the polarimetric parameters of interest. The calculations and the different types of image processing for the processing and post-processing steps of the Mueller matrices can be carried out very quickly using a graphics card (GPU) or several graphics cards. Different types of programming languages (C, C++, Python, etc.) can be used for the calculations and for the processing of the images,as well as for image acquisition management. Different types of programming languages (C, C++, Python, etc.) can be used for image calculation and processing, as well as for image acquisition management. Using the same programming language for image acquisition and for the image processing and post-processing phase could make it possible to completely eliminate the delay between image acquisition and their real-time restitution. The computer means 6 can thus comprise a controller for controlling the PSG and the PSA and for processing the images coming from the camera, said controller comprising for example one or more FPGAs, and a workstation comprising a computer equipped, where appropriate, with at least one graphics card, and a human-machine interface comprising for example a screen, a keyboard, and one or more control buttons or pedals,as described further. The illumination system is capable of emitting in each of the observation spectral bands. In the case where the spectral bands are in the red, green and blue respectively, the illumination system may comprise a white light source such as a xenon source associated with spectral filters, or a set of LED diodes emitting respectively in the red, green and blue, or LED diodes emitting in the blue and yellow. The use of a xenon source may be preferred for the simplicity of use it provides and its power and because it is already widely used for endoscopic systems. The camera of the acquisition system being multi-sensor, preferably a tri-CCD or tri-CMOS RGB camera,it can acquire several images at the same time without significant loss of light and without interference problems ("cross talking") between the channels. The image obtained on each of the sensors can be a grayscale intensity image for the corresponding spectral band. The grayscale intensity images of the different sensors are perfectly superimposable by construction, which simplifies the reconstruction of the final images after processing, in particular the addition of polarimetric information to color images or the combination of polarimetric images at different wavelengths. Images obtained by the multispectral polarimetric imaging system, applied to Mueller polarimetry in the different spectral bands, in particular in red R, green G and blue B, can be generated by calculation in the different spectral bands, as illustrated in Figure 1. These images can, if necessary,be combined to generate a Mueller polarimetry RGB color image. In parallel, the intensity images acquired by the three camera sensors can be combined to form an RGB color image. This image is useful to allow the practitioner to clearly identify the area where the polarimetric analysis is carried out. The determination of the Mueller matrices for each of the R, G and B spectral bands makes it possible to calculate in each of these bands intensity images translating the value of a polarimetry parameter (for example the retardance) for each of the pixels of the image, as illustrated in Figure 2. Thus, for example, for each spectral band, a depolarization image determined from the knowledge of the Mueller matrix as well as an image of the retardance is generated,also determined from knowledge of the Mueller matrix. This or these retardance and / or depolarization images can be merged with the color image to generate an image I in which the hue and / or contrast of certain areas is modified in order to provide additional information to the practitioner, helping with diagnosis. The combination within the same imaging system of a polarimeter with liquid crystal polarization modulators, a multispectral illumination source, in particular a xenon lamp, and a tri-CCD or tri-CMOS camera makes it possible to obtain a particularly compact and efficient imaging system, well suited to colposcopy in particular. The fact of being able to generate multimode images perfectly spatially aligned with each other facilitates deep learning, for example by means of a convolutional neural network 7, as illustrated in Figure 3,this network receiving as input conventional color images, monochromatic intensity images and polarimetric images for the selected wavelength ranges and delivering as output one or more simplified images to aid diagnosis, for example in the form of increased contrast. It is possible, in particular during learning, which may be supervised learning, to provide the artificial intelligence system with non-polarimetric RGB color images as well as representative polarimetric images, for example images relating to depolarization, linear phase retardation and azimuth of the linear phase retardation, etc.,the artificial intelligence system producing one or more images containing information aiding diagnosis. The computing means 6 may be configured to allow all or part of the following visualizations: - a parallel multimode visualization (figure 4A) of different polarimetric and non-polarimetric images, with for example display on the same screen of the RGB color image obtained by the camera (on the left), an image of the linear phase retardation (in the middle), and an image of the azimuth of the slow axis (on the right) - a parallel multispectral visualization (figure 4B), with for example on the left the reference RGB color image and on the right three series of three images in each of the spectral bands of interest, the first line representing the linear phase retardation,the second line representing the azimuth of the slow axis of the linear phase retardation and the third line representing the depolarization; - an interactive visualization (figure 4C) in which the overall RGB color image is displayed but a restricted display area, in this case circular, for example centered on a pointer that can be moved on the image by the user, displays the information relating to a polarimetric parameter, in this case the linear phase retardation, partially superimposed with the displayed RGB image, - a visualization (figure 4D) with display in the left column of the RGB color image (top), the color polarimetric image (middle), and the image of the azimuth of the slow axis of the linear phase retardation at the bottom, as well as the enlarged image (right) of the three images in the left column merged together. It is also possible to display: - color images of the Mueller matrices,- color images of depolarization, - color images of linear phase retardation, - color images of azimuth of linear phase retardation, etc. Such an imaging system is advantageously applied to colposcopy, as will be detailed later, but can also be applied to other types of imaging systems, for example endoscopy, as well as the microscope and exoscope for neurosurgery. Reminder on Mueller polarimetry Generally speaking, the polarized state of the incident light T, U^ is related to the light exciting a TVW sample (by the relation: T VW( = ^ ∗ TU^ with M the Mueller matrix. In order to obtain a Mueller matrix M of dimensions 4x4, the measurement of intensity coefficients is necessary. These intensity coefficients can be obtained through the use of the polarization state generator (PSG) which produces four independent polarization states characterized by four Stokes vectors S^^ ^i = 1,2,3,4). Each represents a column of the modulation matrix W:W= ^S^^ , S^^ , S^^ , S^^^After interaction with the observed area, each polarization state produced by the polarization state generator is analyzed by four polarization configurations of the polarization state analyzer (PSA). These four configurations are also described by four Stokes vectorsS^^ ^i = 1,2,3,4), which represent the rows of the analysis matrix A: After these steps we obtain the intensity matrix B:^ = ^^^ d’où : ^ = ^^^^^^^ B corresponds to a series of n measurements, in this case 16 measurements in the case of a 4x4 Mueller matrix. According to this model, the measured intensity matrix ^ can be written as the sum of a polarizing contribution and a non-depolarizing contribution. ^= ^^^ + ^^ = ^^^^^^ + ^^^)^^ = ^^^^ + ^^^ = ^ ^^ ^^^^ + ^^)^ ^^The Mueller matrix can be processed according to various known methods in order to calculate polarimetric properties, namely depolarization, linear phase retardation, etc. In the case of the polarimetric imaging system according to the invention, 16 successive measurements can be carried out for each pixel of the image, for each of the sensors of the multi-sensor camera, corresponding to the different coefficients of the intensity matrix B. Preferably, the Mueller matrix is the classic 4x4 matrix, i.e. the complete Mueller matrix, but it is not outside the scope of the invention to use an incomplete 3x3, 3x4, or other version. The invention can also be used for simplified polarimetric imaging techniques such as Stockes polarimetric imaging or other. According to one of its aspects, the invention makes it possible to increase a flow of results obtained for a physical quantity (or a set of physical quantities) X relating to a physical object O per unit of time from consecutive series µ of n real measurements, called original series, succeeding one another in a predefined order at each given time interval, by generating additional series, called artificial series, of the original series. These measurements can be the 16 intensity measurements of the matrix B in the case of Mueller polarimetry, but this aspect of the invention is more general and can be applied to other types of imaging different from Mueller polarimetric imaging.Thus, X can be other than a Mueller matrix obtained by a Mueller polarimetry system, and in particular a non-square matrix of real numbers, a square matrix of real numbers, a vector or even a real number, or more generally any physical property or set of physical properties whose determination requires several successive measurements to be carried out. Otherwise, X can be obtained for a point array (imaging). This array can be two-dimensional (2D) or three-dimensional (3D). This is the case for 2D or 3D imaging. For 2D imaging, each measurement point corresponds to a pixel. For 3D imaging, each measurement point corresponds to a voxel. For the result X=G(µ) to be calculated correctly, the n measurements of the series µ must respect a predefined order µ0, µ1, µ2, … µ. n-1. Any other order, for example such as µ' = (µ2, µ3, …, µn-1, µ0), generally gives a value X'=G(µ') where X' is not the correct result. If the measuring device is capable of performing l times (l∈N) the set of n measurements given by μ = ^μ Q , … , μ S^^ ) in 1 second, in total it provides a set of s discrete measurements m = ^m Q , … , m Z^^) in 1 second, where s=ln. These measured values are, for example, the coefficients of the intensity matrix B in the case of Mueller polarimetry. Figure 5 represents, as an example, three real series µ of measurements, each comprising four measured values (n=4). We denote µ0 as the first value, µ1 as the second, µ2 as the third and µ3 as the last, this order being predefined. We first estimate that the result X (for example the Mueller matrix) does not vary or varies little over time. We can therefore assume that each series µ comprises n measurement values similar to those of the counterparts of the following series obtained in the same order. We can then artificially form new combinations of measurements allowing us to obtain new series of artificial measurements by associating measurements from different series, as illustrated in Figure 6. In the example in Figure 6, we associate, for example, for the series μ[ > the µ values 1, µ2, µ3 of the first series to the value µ0 of the second series. We repeat these groupings on all n measurements. The series μ[ ? , μ[ \ and μ[ ^^ are identical to the µ series in Figure 5. The order of the measurements is preserved. For these series it is not necessary to swap the measurements. For the µ series[ > , μ[ ] , μ[ ^ , μ[ _ , μ[ ` and μ[ ^Q the predefined order of the measurements is not respected; we then apply a permutation function P, this permutation function being illustrated in Figure 7, in order to respect the predefined order of the measurements within each series. Following the permutation of the series μ[ > , μ[ ] , …in order to preserve the predefined order, the function G can be applied to the set of new permuted series P(μ[ > ), P(μ[ ]), …, to calculate X. In the general case of n measurements, if Tn is the time needed to carry out the n measurements, Tt the time to make a measurement, we can carry out 1 / Tn series of n measurements in one second; T p is the time required to perform a permutation, and T cis the computation time needed to calculate X=G(µ), then we preferably have Tc< <Tt et Tp<<Tt. Comme indiqué précédemment, les mesures peuvent être des valeurs successives d’intensité de chaque pixel d’une image enregistrée par un capteur donné de la caméra du système d’acquisition, et le procédé qui vient d’être décrit peut s’appliquer à chaque pixel de cette image. Or, cette image peut évoluer au cours du temps, par exemple en raison des mouvements de la cible. Sur la figure 8 (à gauche) on a considéré à nouveau le cas simplifié présenté précédemment avec n=4. Chaque petit carré représente une partie de l'image dans le champ de vision total représenté par le grand carré. Cette partie de l'image peut se déplacer dans le champ de vision au cours du temps.It is then useful to apply a spatial registration function R to re-register the images acquired at different successive times, as illustrated (right), so that the successive intensity measurements used in the calculation of the Mueller matrix concern the same area. The registration function R can be applied immediately after the acquisition of a new image, starting for example for a series of measurements of the image corresponding to m4 in Figure 8. The registration makes it possible to spatially coincide the pixels corresponding to the values m1, m2 and m3 with that corresponding to m4, those corresponding to the values m2, m3 and m4 with that corresponding to m5, those corresponding to the values m3, m4 and m5 with that corresponding to m. 6,etc., as illustrated in Figure 8. The image registration function can involve any suitable image registration algorithm; an example of a registration function is described in the article M. Irani and S. Peleg “Improving Resolution by Image Registration”, 1991, CVGIP Graphical models and image processing, Elsevier. The registration function may seek to determine the registration values a in x and b in y between two images, so as to minimize a loss function between them. An example of a loss function is given in the article MBA Haghighat, A Aghagolzadeth, and H. Seyedarabi, “A non-reference image fusion metric based on mutual information of image features,” Computers & Electrical Engineering, vol. 37, no. 5, pp. 744-756, Sep. 2011, doi:10.1016 / j.compeleceng.2011.07.012.The permutation function P can be applied after the recalibration function R, to put the measurements in the correct order, then the function G can be applied to calculate the result X. If one seeks to have more than eight times the result X calculated per second, in order to have a certain fluidity corresponding to imaging substantially in real time, one needs 1 / Tn >=8. The procedure described above can also be used in a more general case where the object O can be deformed. If Td is the characteristic time of the deformation, one should preferably have Td>>Tn. If Tx is the characteristic duration of a change in the property materialized by the result X, one should also preferably have Tx>>Tn. In the case of Mueller polarimetric imaging, the measured values are the values of the coefficients Bij (i,j = 1,…,4) of the intensity matrix B.The sixteen coefficients of this matrix correspond to the measurements made for a given pixel, in order to then calculate the Mueller matrix for this pixel, and form a series µ of measurements as illustrated in Figure 9. Figure 10 illustrates the application of the permutation function P(µ) to the series μ[. The function G here corresponds to the calculation which makes it possible to obtain from the intensity matrix B the Mueller matrix M. The registration function R is applied before applying the permutation function P; the registration function R can be applied by taking as a reference image, for example, that corresponding to the first image of a sequence of 16 consecutive images (in the case of a 4x4 Mueller matrix).Preferably, Te + Tf + Tg < Ti, in order to have real-time imaging, with Te the time required for the permutation P, Tf the time required to apply the registration function R, Tg the time required to calculate M, Ti the time required for a single measurement. In the case where part of the acquisitions are made simultaneously in spectral ranges corresponding to the blue, green and red part of the visible spectrum, the coefficients of the non-normalized and non-polarized intensity m̂. ^^Mueller matrices obtained respectively in the blue, green and red parts of the visible spectrum, using the above-mentioned method, can be combined together to produce the color image of the target in real time. Post-processing of Mueller matrices Conventional calculation of the Mueller Matrix can involve, for example, a so-called Lu-Chipman decomposition, as described in the article S.-Y. Lu and R.A. Chipman, “Interpretation of Mueller matrices based on polar decomposition”, Journal of the Optical Society of America A, vol. 13, no. 5, p. 1106, May 1996, doi: 10.1364 / JOSAA.13.001106. However, this decomposition is relatively expensive in terms of computation time because it involves the calculation of the eigenvalues of the matrix M. Moreover, when observing the cervix, we mainly observe birefringence and depolarization effects.Here, it is assumed that non-depolarizing effects are associated with the surface area of the observed area, while depolarizing effects are related to the volume of the observed area, as illustrated in Figure 11. This figure schematically illustrates the propagation of light within the tissue. In this case, the Stokes vector S. nd is associated with the undepolarized component originating from the superficial layer of the tissue, while the Stokes vector S dis associated with the depolarized component of light, originating from the tissue volume. Figure 11 also represents the two Mueller matrices Mnd and Md, respectively related to the non-depolarized and depolarized components of light. The factors q and p are intended to estimate the proportions of non-depolarized and depolarized light, and must be determined. The intensity matrix B can be formulated as the sum of the contributions of the non-depolarized and polarized light. We then have:^= ^^^ + ^^ = ^^^^^^ + ^^^)^ et ^= qM + ^M = ^^^ ) ^^S8 8 A BS8 + B8 WThe Mueller matrix of undepolarized light M nd , can correspond to a linear Mueller retardance matrix, such as :): mF1J2nGn1.2 Go1=pnℎGHJ BJ HrGsJ FG^1BJ J2 BJtFé / *: FJnGFB BJ ^ℎG / J J2 BJtFé / The Mueller matrix of depolarized light M d can be equal to: 1 0 0 0 ^ ^ = +0 0 0 00 0 0 0 7 0 0 0 0 We have The Mueller matrix M can be normalized with respect to its unnormalized and unpolarized intensity coefficient =y ^^ = ^ + ^ and can be written in the form: ^2.F=GH1 / éJ =^^^^ + ^^^ ^^^ + B^^^ + ^ = 1 + B Or The matrix ^2.F=GH1 / éJ can then be written in the form: the coefficients ^1, { = 1,2,3,4) being the coefficients of the matrix M normalized with respect to its unnormalized and unpolarized intensity coefficient =y^^ = ^ + ^. The depolarization can be calculated using the following formula:D épolarisation This decomposition method offers significant time savings since it does not require the time-consuming calculation of eigenvalues and eigenvectors. In addition, this method is matrix-based, which allows an algorithm to perform multiple calculations in parallel and not in series, which is a significant time and memory saving. The various polarimetric parameters can thus be obtained more quickly, which is very advantageous in the case of real-time polarimetric imaging. Figure 12 compares polarimetric images obtained on the one hand via a classic Lu-Chipman decomposition, and on the other via the additive decomposition described above.The three top images are obtained by additive decomposition as described above and correspond respectively to the display of the linear phase retardation * (left image), the azimuth of the linear phase retardation ) (middle image) and the depolarization (right image), while the three bottom images represent the same parameters obtained via the Lu-Chipman decomposition. There is a high similarity between the images representing the same parameters, which reflects the performance of the method. The imaging system according to the invention can be arranged to carry out a colposcopy examination, for example using a colposcopy system 10 as illustrated in Figure 13. This system 10 comprises a colposcope 20 having a head 21 carried by an articulated arm 22, itself carried by a rolling base 31. A vertical mast 30 is also carried by the rolling base 31. The colposcope 20 is connected to an illumination system comprising a light source 50 housed in a housing connected to the head 21 by a flexible optical guide 51, not visible in Figure 13 but apparent in Figure 14, the housing housing the source 50 resting for example on a workstation 60 comprising a computer. The colposcopy system 10 comprises a screen 40 connected to the computer 60, which is advantageously a touch screen, preferably liquid-tight. This screen is carried by the mast 30, as well as a keyboard 41, connected to the computer 60.A controller 70 is connected to the computer 60 to control in particular the operation of a polarimetric system 90 carried by the head 21. This controller 70 can also be connected to an acquisition system 110, in particular to the camera 5. Preferably, as seen in Figures 32 and 33 more particularly, the colposcopy system 10 comprises at least one reflector 81 housed in a support carried by the mast 30 and used for calibrating the polarimetry system. It is covered by a light shielding cap 80, preferably carried by the mast 30, in particular at its top. This reflector 81 is for example articulated around a vertical axis of rotation, and comprises for example on one face a frosted metal wall 82, for example made of aluminum, and on the opposite side a reference surface 83 having known spectral properties, for example a substrate with a neutral color surface with high Lambertian reflection called "spectron".The rotatable mounting of the reflector 81 makes it possible to easily replace one of the faces with the other without moving the head 21 of the colposcope, which facilitates calibration operations. During this, the head 21 is for example placed approximately 30 cm from the reflector 81, by maneuvering the articulated arm carrying the head. The colposcopy system 10 advantageously comprises a pedal 100 visible in FIG. 13 which allows the user to trigger a predefined action, for example to start a polarimetric acquisition. The head 21 of the colposcope may also comprise a button 101, as illustrated in FIG. 16, making it possible to trigger another predefined action, for example the start and stop of video recording. The polarimetric system 90, illustrated in Figure 19, comprises the polarization state generator PSG and the polarization state analyzer PSA, almost all of the optical elements of which are illustrated in Figure 20.The PSG polarization state generator comprises a succession of four optical elements placed on the path of the light exiting through a window 210 of the head 21 of the colposcope, namely in the direction going from the outside towards the exit window 210, a first half-wave liquid crystal polarization modulator at 510nm HFLC 191, a quarter-wave plate at 633nm QWP 192, a second quarter-wave liquid crystal polarization modulator at 510nm QFLC 193, and a fixed linear polarizer 194. The PSA polarization state analyzer comprises a succession of three optical elements placed on the path of the light reaching one of the input ports 211a and 211b associated respectively with the left and right eyepieces, in this case the input port 211b associated with the right eyepiece for the practitioner.Starting from the outside and going towards the head of the colposcope, there are successively a first half-wave liquid crystal polarization modulator at 510 nm HFLC 221, a quarter-wave plate at 633 nm QWP 222 and a second quarter-wave liquid crystal polarization modulator at 510 nm QFLC 223. The analyzer also comprises a linear polarizing filter 224 similar to the polarizer 194 of the PSG, arranged downstream of the camera 5 within the acquisition system 110, as illustrated in FIG. 36. Each optical component 191, 192 or 193 of the PSG has, in the illustrated example, a circular shape and is held between two support pieces attached to each other, one having a housing 233 for receiving the element and the other serving to hold the element in its housing. All these support pieces have a C-shaped top part open at the top and leave the upper edge of each optical element clear.The polarizer 194 has a square or rectangular outline and is housed in a support 265 provided with an upwardly open slide. The optical element 191 is placed between a first support piece 261 and a second support piece 262 which has a housing 233 receiving the element 191, the support piece 261 forming the holding cover. The support piece 262 serves as a cover for a third support piece 263 having a housing 233 receiving the optical element 192, as seen in Figure 21. This third support piece 263 serves as a cover for a fourth support piece 264 having a housing 233 receiving the optical element 193. The different support pieces can overlap and are fixed together with the support 265 through the base of the PSA on the front face of the head 21 of the colposcope by screws 242, as illustrated in Figure 19.The optical elements 221, 222 and 223 of the PSA also have a circular shape and are held by support pieces 251, 252, 253 and 254 superimposed on each other. Each support piece 251 to 254 has a base 255 provided with holes for the passage of the screws 242 used for fixing to the front face of the head 21, and an upper part 256 in the general shape of a C oriented substantially at 45° downwards, the upper part 256 being connected by an upright 257 to the base 255, as illustrated in FIG. 22. The support pieces 252 to 254 each have a housing 258 receiving a respective optical element. It is seen in Figure 22 that the optical element 221 is held between the support pieces 251 and 252, the support piece 251 serving as a cover, that the optical element 222 is held between the support pieces 252 serving as a cover and 253, and that the element 223 is held between the support pieces 253 serving as a cover and 254.The support part 252 carries in the upper part two branches 270 for fixing a filter holder 310 shown in figure 34. This filter holder 310 carries in the example considered three filters 311, 312 and 313 which are respectively a phase retarder L30 whose fast axis is oriented at 30° relative to the linear polarizer P0, a linear polarizer P90 whose transmission axis is oriented at 90° relative to the polarizer P0, and a polarizer P0 whose transmission axis is oriented at 0° relative to a reference. The filter holder 310 also comprises a clear zone placed in front of the PSA once the calibration is finished in order to allow the acquisition of the images. The filter holder 310 is movable in translation in the example considered, in a generally horizontal direction. An electronic circuit 320 allows the computer 60 to know the position of the filter holder, and therefore the filter which is active if applicable.The filter holder 310 can be motorized or moved manually, during calibration, so as to successively place the different filters 311 to 313 in front of the PSA. The calibration can be carried out according to the so-called eigenvalue calibration method (ECM). In particular, it can be seen in Figure 25 that the housing 233 is open laterally so as to allow the control wires of the optical element received therein to exit through a passage 233a. The housing 233 also has, opposite the passage 233a, a clearance 233b for receiving the control wire at the level of its connection to the optical element. The wire follows the contour of the optical element in a semicircular channel 233c which matches the profile of the optical element.It can be seen in Figure 23 that the optical elements 221 and 223 of the PSA, of larger diameter, almost entirely overlap the two input ports 211a and 211b, while the optical element 222, of smaller diameter than the elements 221 and 223, entirely overlaps the port 221b but partially overlaps the port 211a. Nevertheless, this does not unduly affect the quality of the observed image, taking into account the focusing distance, because the focusing is carried out at a distance much greater than the distance of the optical element from the port 221a. Figure 27 is a schematic and partial section of the light source 50.This comprises a xenon lamp 52 and a set of filters 53 and 54, i.e. a first band-pass filter (Edmund optics #84-728) which allows only light between 400 and 750 nm to be transmitted with a transmittance of approximately 95% and a second long-pass filter (Edmund optics #84-754) which allows only light at a wavelength greater than 400 nm to be transmitted with a transmittance of approximately 93%. The choice of these filters makes it possible to obtain a relatively high transmittance, around 90% in the 400-750 nm range. The optical guide 51 is held at one end 51a in the axis of the lamp 52 by means of an end piece 56. As illustrated in figure 29, the optical guide 51 is received at its other end 51b in the axis of an aspherical lens 180 of the colposcope, so as to give an image of the liquid guide approximately 30 cm from the head of the colposcope.The lens 180 is placed in front of a prism 181 which reflects the light towards the exit window of the colposcope. The optical guide 51 can be held in place if necessary using a set of supports 191 and 192 whose position relative to the body of the head of the colposcope is adjustable respectively in X and Y, so as to allow the position of the end 51b in the axis of the lens 180 to be precisely adjusted. It is advantageous to use a liquid guide 51 whose core has a diameter of 3 mm, because this allows, as illustrated in Figure 31, to benefit from more intense illumination in the center of the area illuminated by the colposcope, compared to a conventional guide whose core is 5 mm in diameter.In particular, the liquid guide with a diameter of 3mm allows to obtain a higher illumination, compared to that obtained with a liquid guide of 5mm diameter, on a surface of 3cm diameter corresponding approximately to the diameter of the uterine cervix. The acquisition system 110 is fixed, as can be seen in particular in figures 36 to 38, on the head 21 of the colposcope and receives the light reaching one of the eyepieces (the same as the one in front of which the PSA is placed) thanks to one or more deflecting prisms integrated into the head 21. The camera 5 is equipped with an objective 415, and a filter wheel 411 carrying in the example considered three filters 421, 422 and 423. This is for example a tri-band filter adapted to the camera 5, a bandpass filter centered on 650 nm and for example 40 nm of spectral width, and a bandpass filter centered on 700 nm and 50 nm of spectral width. The filter wheel 411 is rotated by a motor 410.An angular positioning system using a contactor can enable the system to know the angular position of the wheel, and therefore of the filter which must be positioned exactly in the path of the light reaching the camera 5. The camera 5 can comprise, as illustrated in figure 39, three dichroic prisms, to separate the wavelengths towards the different sensors. The camera comprises, for example, dichroic prisms 605, 604 and 606, the sensor 603 for detecting blue being fixed to the prism 605, the sensor 601 for detecting red being fixed to the prism 604, itself fixed to the prism 605, and the sensor 602 for detecting green being fixed to the prism 606, itself fixed to the prism 604. To use the system 10, the user can carry out polarimetric calibration in a manner known per se using the two faces of the reflector 81 and the different filters of the filter holder 310.Then, the head 21 can be positioned so as to image the cervix after dilation of the vagina with the use of a speculum. At any time, the practitioner can start recording a video by pressing the button 300 and / or trigger the polarimetric acquisition by pressing the pedal 100. Pressing the pedal triggers the polarimetric acquisition by stopping the video recording, which automatically resumes after the end of the polarimetric acquisition. Pressing the button again stops the video recording. Of course, the invention is not limited to the example just described. In particular, the optical system of the colposcope can be modified, for example by removing the eyepieces, with observation being carried out solely on the screen.
Claims
Claims 1. Real-time imaging method, comprising the steps of: - Acquiring, per given time interval, original real series μ of n measurements μ = ^μ Q , … , μ S^^) following one another in a predefined order within the series, each series of n measurements making it possible, by processing these measurements, to generate at least one original result X= G(μ), at a given frequency f, - generating artificial series from a grouping of measurements coming from a real series μ of a given rank, and from at least one real series of a different rank, in particular of the following rank, and from a permutation P of the measurements within each artificial series so as to respect the predefined order of the n measurements within each series, - generating a flow of results at a frequency greater than f from the original results and from the results generated from the artificial series intercalated between the real series.
2. Method according to claim 1, being applied to polarimetric imaging. 3.Method according to claim 2, being applied to colposcopy, microscopy, brain exoscopy, endoscopy, preferably to polarimetric colposcopy, and better, to multispectral polarimetric colposcopy.
4. Method according to any one of the preceding claims, the result X being a Mueller matrix (M) or a physical property calculated from this matrix, the series of measurements being polarimetric measurements.
5. Method according to the preceding claim, the series of measurements corresponding to the acquisition of the coefficients of the intensity matrix B, preferably at least 9, in particular 12, better 16, coefficients of the intensity matrix B.
6. Method according to claim 5, the Mueller matrix being a 4*4 matrix, with B=AMW,. où and W = ^S ^^ , S ^^ , S ^^ , S ^^ ^ the measurements being obtained using a PSG polarization state generator which produces four independent states described by four Stokes vectors S ^^ ^i = 1,2,3,4) and a PSA polarization state analyzer which generates analysis states described by four Stokes vectors S ^^ ^ i = 1,2,3,4 ) .
7. Method according to claim 6, in which an additive decomposition of the Mueller matrix M is carried out in the form M = qMnd+pMd where Mnd is the non-depolarizing component and M d is the depolarizing component, the parameters q and p being the weights of the two components of the Mueller matrix, Mnd being in particular the Mueller matrix of a linear phase retarder given by: M ^^ = ^&'(^), *) and ^ ^ being, in particular, the Mueller matrix of a pure depolarizer given by ) the azimuthal orientation of the fast axis in degrees and * the phase delay in degrees (ranging from 0° to 180°). with the coefficients m ij (i,j=1,2,3,4) being those of the normalized Mueller matrix M, m̂ ^^ being the unnormalized and unpolarized intensity coefficient of M, the other coefficients mij (i,j = 1,2,3,4) being normalized with respect to m̂ ^^ .
8. A method according to claim 7, the depolarization being given using the following formula: 1Depolarization = 1 − 1 + B9. Method according to one of claims 7 and 8, the Mueller matrix and said parameters being obtained for a single wavelength or simultaneously for several wavelength ranges, in particular at least two, in particular in the visible and infrared spectral range.
10. Method according to any one of claims 7 to 9, applied to the reconstruction of a color image, in real time from the combination of the coefficients of the non-normalized and non-polarized intensity m̂ ^^, of the three Mueller matrices obtained simultaneously in spectral ranges corresponding respectively to the blue, green and red part of the visible spectrum.
11. Method according to any one of the preceding claims, in which before carrying out the permutation P, a spatial registration of the images corresponding to the measurements is carried out in order to take into account possible displacements of the areas observed in the field of vision of the camera.
12. Multispectral polarimetric imaging system (10), in particular for colposcopy, comprising: - an illumination system comprising at least one light source (50), this illumination system emitting in at least two, in particular at least three, spectral bands, - a polarization state generator (PSG) arranged downstream of the light source and upstream of a target (T) to be imaged, - a polarization state analyzer (PSA) arranged downstream of the target to be imaged, - a multi-sensor camera (5) comprising at least two, better at least three, sensors (601, 602, 603) for recording respectively at least two, better at least three, images in said spectral bands.
13. System according to claim 12, the spectral bands being three in number and preferably ranging from 445 nm to 475 nm for the first, from 510 nm to 550 nm for the second, and from 600 nm to 660 nm for the third, these spectral bands being preferably centered on 460 nm, 530 nm and 630 nm respectively.
14. System according to one of claims 12 and 13, the camera (5) being a multi-CCD or multi-CMOS camera, in particular bi-CCD, bi-CMOS, tri-CCD, tri-CMOS, 4-CMOS, 4-CCD, better tri-CCD or tri-CMOS, preferably tri-CMOS.
15. System according to claim 11, the spectral bands being three in number and respectively in the red / near infrared, green and blue. 16.System according to any one of claims 12 to 15, the camera comprising at least two, better at least three, dichroic prisms, for separating the wavelengths towards the different sensors.
17. System according to any one of claims 12 to 16, the polarization state generator (PSG) comprising electrically controllable liquid crystal polarization modulators.
18. System according to any one of claims 12 to 17, the polarization state analyzer (PSA) comprising electrically controllable liquid crystal polarization modulators.
19. System according to one of claims 12 and 18, the liquid crystal polarization modulators being ferroelectric or nematic liquid crystal polarization modulators, preferably ferroelectric.
20. System according to any one of claims 12 to 19, the light source being a white light source, better a xenon lamp. 21.System according to any one of claims 12 to 20, comprising a tri-band dichroic filter downstream of the light source.
22. System according to any one of claims 12 to 21, comprising a filter wheel containing three filters (421, 422, 423), arranged in front of the camera.
23. System according to claim 22, the filter wheel carrying a multi-band filter, in particular tri-band, preferably to allow acquisition at lengths. of waves in spectral bands centered around 460nm, 530nm and 630nm by the camera (5), as well as one or more monochromatic dichroic filters, preferably to allow the acquisition of images at 650nm and 700nm respectively.
24. System according to any one of claims 12 to 23, comprising a processor for generating at least one Mueller polarimetric image of the target in each of the spectral bands.
25. System according to claim 24, the processor being arranged to superimpose at least partially, on an image corresponding to the non-polarimetric observation by the camera, an image containing at least one polarimetric information. 26.System according to one of claims 24 and 25, being arranged to make real series of n measurements corresponding to the acquisition of at least 9, in particular 12, better 16, coefficients of an intensity matrix B, the processor being arranged to - generate series of artificial measurements from a grouping of measurements coming from a real series μ of measurements of a given rank, and from at least one real series of a different rank, in particular following, and from a permutation P of the measurements within each artificial series so as to respect the predefined order of the n measurements within each series, - generate a stream of polarimetric images at a frequency higher than that which would be allowed by said real series of measurements without the series of artificial measurements, from the polarimetric images produced from the real series of measurements and from the polarimetric images generated from the artificial series intercalated between the real series. 27.System according to claim 26, the processor being arranged to, before carrying out the permutation P, carry out a spatial registration of the images corresponding to the measurements in order to take into account possible displacements of the zones observed in the field of vision of the camera.
28. System according to any one of claims 24 to 27, the processor being arranged to carry out an additive decomposition of the Mueller matrix M in the form M = qMnd+pMd where Mnd is the non-depolarizing component and Md is the depolarizing component, the parameters q and p being the weights of the two components of the Mueller matrix, Mnd being the Mueller matrix of a linear phase retarder given by: M^^ = ^&'(^), *). and M8 is the Mueller matrix of a pure depolarizer given by: 1 0 0 0 ^ ^ = + 0 0 0 0 0 0 0 0 7 0 0 0 0) is the azimuthal orientation of the fast (or slow) axis in degrees and delta the phase retardation in degrees (ranging from 0 to 180°) with @ / 122) / 12* * = atan at A 1 + BC = atan < @ an sin^2)) - - 1 + B cos * cos * B = =>> − 1 = -. / * − 1 = 1 + B − 11 + B the coefficients mij (i,j=1,2,3,4) being those of the normalized Mueller matrix, m̂ ^^ being the unnormalized and unpolarized intensity coefficient of M, the other coefficients mij (i,j = 1,2,3,4) being normalized with respect to m̂ ^^ .
29. System according to claim 28, the processor being arranged to calculate the depolarization by performing the following operation: 1Depolarization = 1 − 1 + B 30. System according to any one of claims 12 to 29, being arranged to generate the parallel display of at least one non-polarimetric image in at least one of the spectral bands, and of at least one polarimetric image.
31. System according to any one of claims 12 to 30, being a colposcopy system, this system comprising: - a rolling base (31), - a mast (30) carried by the rolling base, - a colposcope head (21) carried by an articulated arm connected to the base, - at least one reference reflector (81) carried by the mast (30), used for calibrating the polarimetry system, and preferably also: - a workstation comprising a computer (60), carried by the rolling base, - a keyboard (41) carried by an articulated arm connected to the mast, - a screen (40) carried by an articulated arm connected to the mast at a height greater than that of the arm supporting the keyboard. 32.System according to claim 31, the reference reflector (81) being arranged in a light shield (80), at the top of the mast, being articulated around a vertical axis, and comprising on one face a frosted metal wall, preferably made of aluminum, and on the opposite side a reference surface having known spectral properties.
33. System according to one of claims 31 and 32, comprising a pedal (100) which allows the user to trigger a predefined action, in particular to start a polarimetric acquisition, the head (21) of the colposcope comprising a button (101) allowing to trigger another predefined action, in particular the start and stop of the video recording. 34.System according to any one of claims 12 to 33, being a colposcopy system, this system comprising an illumination system comprising o At least one light source (50), o a liquid light guide (51), to be connected at one end to the light source, this light guide having a core diameter less than or equal to 5 mm. the system comprising a colposcope head (21), comprising an optical system having a light input port connected to the other end of the liquid light guide, and a light output port for illuminating an area to be observed.
35. System according to claim 34, the optical system comprising between the light input port and the light output port an aspherical lens (180) and a deflecting prism (181).
36. System according to claim 34 or 35, the inner diameter of the liquid optical guide (51) being between 2.5 and 3.5 mm.
37. Method for training an artificial intelligence system (7), preferably comprising at least one convolutional neural network, in which the artificial intelligence system receives as input non-polarimetric images and polarimetric images generated by the polarimetric imaging system according to any one of 12 to 36. 38.Observation system with two input ports, in particular a binocular system, in particular for colposcopy, comprising a head (21) comprising an optical system having a light output port for illuminating an area to be examined and left and right input ports (221a, 221b) directed towards the area to be observed, the observation system further comprising a polarimetric system (90) comprising a polarization state generator (PSG) arranged in front of the output port and a polarization state analyzer (PSA) of which at least a part of the optical elements is arranged in front of one (211b) of the input ports, this analyzer comprising at least one optical element (222) held by at least one support applied to only a part of the contour of the optical element so as to provide a free edge for the element, this free edge partially overlapping the other input port (211a). 39.System according to claim 38, the support extending over an angular extent of between 180° and 300° in contact with the optical element, preferably having a general C shape open in a downward oblique direction, substantially at 45°.
40. System according to claim 38 or 39, comprising at least one other support for holding at least one optical element of the polarization state generator placed in front of the light output port.
41. System according to claim 38, said other support having a general shape of an upwardly open C.
42. System according to any one of claims 36 to 39, each support comprising a succession of parts holding optical elements of the polarization state generator or of the analyzer between them, at least two supports serving to hold two consecutive optical elements sharing an intermediate support part.
43. System according to claim 40, comprising at least two consecutive support parts assembled against each other, each of these parts having on one face a housing for receiving a respective optical element and one of the parts serving to retain the optical element of the other part.
44. System according to any one of claims 38 to 43,the polarization state generator and the analyzer each comprising similar optical elements which are placed in reverse orders relative to the direction of propagation of the light, the generator preferably comprising in the direction of propagation of the light, a linear polarizing filter P, a quarter-wave liquid crystal polarization modulator QFLC, a half-wave plate QWP, and a half-wave liquid crystal polarization modulator HFLC, the analyzer preferably comprising, in the direction of propagation of the light, a half-wave liquid crystal polarization modulator HFLC, a half-wave plate QWP, a quarter-wave liquid crystal polarization modulator QFLC, the polarizer (224) associated with the analyzer preferably being arranged after returning the light from the input port to an associated eyepiece,upstream of a camera (5) used to collect intensity images used to generate a Mueller matrix.
45. System according to any one of claims 38 to 44, the housings (233) receiving the electrically controllable optical elements comprising at least one passage (233a, 233b, 233c) for an electric cable.
46. System according to any one of claims 38 to 45, being a colposcopy system, this system comprising: - a rolling base (31), - a mast (30) carried by the rolling base, - the head (21) being carried by an articulated arm connected to the base, - at least one reference reflector (81) carried by the mast (30), used for calibrating the polarimetric system, and preferably also: - a workstation comprising a computer (60), carried by the rolling base, - a keyboard (41) carried by an articulated arm connected to the mast, - a screen (40) carried by an articulated arm connected to the mast at a height greater than that of the arm supporting the keyboard.
47. System according to claim 46, the reference reflector (81) being arranged in a light shield (80), at the top of the mast, being articulated around a vertical axis, and comprising on one face a frosted metal wall, preferably made of aluminum, and on the opposite side a reference surface having known spectral properties. 48.System according to one of claims 46 and 47, comprising a pedal (100) which allows the user to trigger a predefined action, in particular to start a polarimetric acquisition, the head (21) comprising a button (101) making it possible to trigger another predefined action, in particular the start and stop of the video recording.
49. System according to any one of claims 38 to 48, being a colposcopy system, this system comprising an illumination system comprising o At least one light source (50), o a liquid light guide (51), to be connected at one end to the light source, this light guide having a core diameter less than or equal to 5 mm, the optical system having a light input port connected to the other end of the liquid light guide. 50.System according to claim 49, the optical system comprising between the light input port and the light output port an aspherical lens (180) and a deflecting prism (181).
51. System according to claim 49 or 50, the internal diameter of the liquid optical guide (51) being between 2.5 and 3.5 mm.
52. Polarimetry system (90) intended to be attached to a head of an observation system, in particular a colposcope head, comprising an optical system having a light output port for illuminating an area to be examined and input ports. left and right directed towards the area to be observed, the polarimetric system comprising a polarization state generator placed in front of the output port and a polarization state analyzer of which at least part of the optical elements is placed in front of one of the input ports, this analyzer comprising at least one optical element held by at least one support applied to only part of the contour of the optical element (222) so as to provide a free edge for this optical element, this free edge being arranged so as to partially overlap the other input port, the polarimetry system comprising fixing means (242) on the head (21) of the observation system.
53. Polarimetry system (90) intended to be attached to a head of an observation system, in particular a colposcope head, comprising an optical system having a light output port for illuminating an area to be examined and at least one input port,the polarimetric system comprising a polarization state generator (PSG) placed in front of the output port and a polarization state analyzer (PSA) of which at least some of the optical elements are placed in front of the input port, at least one of the analyzer and the generator comprising optical elements (222) held by at least one support, each support comprising a succession of parts holding between them optical elements of the polarization state generator or of the analyzer, at least two supports serving to hold two consecutive optical elements sharing an intermediate support part, each of these parts preferably having on one face a housing for receiving a respective optical element and one of the parts serving to retain the optical element of the other part.,