WO2007058570A1

WO2007058570A1 - Computer-aided image processing method to determine tissu viscoelasticity

Info

Publication number: WO2007058570A1
Application number: PCT/SE2005/001734
Authority: WO
Inventors: Gert Nilsson; Chris Anderson
Original assignee: Wheelsbridge Ab
Priority date: 2005-11-17
Filing date: 2005-11-17
Publication date: 2007-05-24

Abstract

This invention relates to the non-invasive assessment and determination of tissue viscoelastic properties including water content, which may be uncomfortably low (dehydration) or high (hyperhydration) following medical procedures, certain diseases and during residence and work in extreme environments as well as tissue elasticity and stiffness, which may be altered in e.g. prostate gland disease. More specifically, the invention relates to a method of determining the tissue viscoelastic properties from the recovery of tissue following the making of an imprint by a mechanical device, employing computer aided image analysis to produce a representation of how the multi-phase recovery to normal state of impressed tissue is influenced by an increased or reduced water content, elasticity or stiffness.

Description

Computer-aided image processing method to determine tissue viscoelasticity

Field of invention

This invention relates to the non-invasive assessment and determination of tissue vis- coelastic properties including water content, which may be uncomfortably low (dehydration) or high (hyperhydration) following medical procedures, certain diseases and during residence and work in extreme environments as well as tissue elasticity and stiffness, which may be altered in e.g. prostate gland disease.

Background of the invention

In medical practice and during residence and work under extreme environmental conditions, the maintenance of adequate body water content is of paramount importance for maintaining the comfort and well-being of the human body. Dehydration (too low water content) or hyperhydration (too high water content) may both result in serious and even life-threatening conditions that need immediate medical care. Practical methods for recording especially the water content of the skin are consequently urgently needed. The transepidermal water loss (evaporative exchange of water with the envi- ronment) can readily be recorded by use of available non-invasive technologies (Nils- son, G.E., Measurement of water exchange through skin, Medical & Biological Engineering & Computing, 15, p209-218, 1977 ), while assessment of tissue water content and more generally tissue viscoelasticity, have been more difficult to perform, due to a lack of suitable methods. Attempts have been made to study the reduction in force ap- plied to the skin for maintaining a certain depression of the tissue, which can be demonstrated to be related to water translocation in tissue, and thereby to the degree of oedema formation (Mannan, M. Characterization of cutaneous oedema — modelling and measurement methods. Thesis. Linkδping Studies in Science and Technology, No.. 227, 1990)). Practical methods based on this method are, however, still bulky (Zdolsek, H.J., Lindahl, O.A., and Sjoberg, F., Non-invasive assessment off fluid volume status in the interstitium after haemodialysis, Physiological Meas. 2000, May;21(2);p211-20), and neither suitable for embedding in e.g. a wearable transducer nor adopted for advanced computer-aided analysis. More recently, methods based on the measurement of the dielectric properties of the skin (Nuutinen, Jet al., Validation of a new dielectric device to assess changes of tissue water in skin and subcutaneous fat, Physiol. Meas. 25, p447-454, 2004), recording of the bioelectrical impedance at a single frequency (Ward et al, Method and device for measuring tissue oedema, US patent 6, 760, 617- 2004) and impedance (Kun et al, Impedance spectroscopy system for ischemia monitoring and detection, US patent 5,807,272 - 1998) have been presented, but have not yet resulted in an easily applicable, specific and inexpensive technique for the evaluation of tissue water content. Alternative methods based on making an imprint in the tissue, the recovery of which is followed in time (Fi? 2 840 795 - Al) have been sug- gested but no method for the detailed analysis using computer-aided methods and image processing technologies have been advised to date. SMn surface observation cameras (JP 2002 112970 A) demonstrate how polarized light can be utilized to block the diffusely backscattered light from tissue, but gives no guidance on how computer-aided image processing methods can be used to extract an output signal representing the tis- sue viscoeiastic properties.

If such methods can be made available, they would find applications in

1) recording the hydration in infants and young children, 2) assessing the risk of oedema formation following active hydration instituted to avoid dehydration in patients treated in intensive care units and during surgical procedures,

3) dialysis,

4) skin testing, 5) development of skin care products,

6) home health care settings, giving an early warning to patients at risk of oedema formation,

7) situations of elevated risk for oedema formation such as transatlantic flights and during military and work-related operations in extreme climates, and 8) evaluation of altered elasticity and stiffness of tissues including the prostate gland. Obviously there are many applications in which there is a need for an easy-to-use and reliable method for assessment of the water content, elasticity and stiffness of tissue.

Description of invention

It is an object of the present invention - using computer aided image analysis - to provide for a method employing mechanical depression of the tissue and an imaging device, utilizing polarized or non-polarized light, to produce a representation of how the multi-phasic recovery to normal state of impressed tissue is influenced by an increased or reduced water content, elasticity or stiffness.

It is another object of the present invention to provide for a reliable, quick and cost- effective method to determine if a subject suffering from a disease or condition known to produce impairments in the tissue water content balance, tissue elasticity or tissue stiffness is susceptible to acquire such complications and thereby need a specific regiment or therapy or changes thereof.

These and other objects will be apparent from the following specifications and its ap- pended claims.

In it most general form, the recovery of a tissue to its normal state and shape - following the formation of a mechanical imprint in the tissue - is analysed by use of computer-aided image processing techniques. The initial formation of an imprint in the tis- sue is performed typically by use of a mechanical device forming an integral part of the complete transducer system or independently by a separate device. The mechanical device, or part thereof, for depressing the tissue, may consist of disposable material in order to avoid contamination of the tissue under investigation. When the mechanical device, having depressed the tissue for a predetermined amount of time, is removed or in other ways prevented from making contact with the tissue, the imprint of the tissue persists for a certain period of time. The tissue returns to its original shape and state through a multi-phasic process at a rate determined by the viscoelastic properties of the tissue. Examples of such tissue imprints at various points in time following release of the device forming the imprint are shown in the following exemplifying part of the specification.

In accordance with the present invention, these imprints are investigated by illuminat- ing the target tissue, typically the body surface, with polarized or non-polarized light within a specific wavelength interval and from an angle that generates a shadow effect on the tissue surface. Backscattered light is collected with or without a polarized filter and a lens system, and detected by means of a photosensitive array sensitive to a specific wavelength interval. Arrangements to generate polarized light and polarizing fil- ters are well known and will not be further discussed herein. A photo-sensitive array may be implemented as a digital camera that is capable of converting the incoming light having passed through the polarizing filters and lens system, to digital values. Alternatively, conventional types of sensors producing analogue signals are conceivable and may be connected to a digital analogue/digital converter of conventional na- ture. The person skilled in this technology may readily envision suitable arrangements.

By successively collecting information, using said photo-sensitive array, of said imprint of the tissue, a series of data matrixes may be generated that carries information about the rate of return to the original shape and state of the tissue and thus to its water con- tent, elasticity and stiffness. The so collected, digitalized information is transferred to a computing device, wherein it is separated by the data processor into digitalized information suitable for further image processing measures. The computing device is further adapted to generate an output signal by processing the data matrix values by use of a suitable algorithm.

According to one embodiment of the invention each data matrix is processed in such a way that values below a certain threshold value in the original image, appear as dark areas in the processed image. By use of threshold techniques, the image of the imprint as represented in the data matrix is enhanced. The number of black individual data points in the resulting data matrix in relation to the total number of data points in the same data matrix represents a measure of the extent and the depth of the imprint at a specific point in time during the tissue recovery period, hi order to improve the signal- to-noise ratio of the calculations an area coinciding with the site of the skin that is im- printed may be used in the analysis. This area can be defined by a region of interest entirely enclosing the imprint as displayed by the first data matrix. Alternatively this region of interest can be made identical to the mask formed by the black data points in the first or any subsequent data matrix. This mask is then applied as the region of inter- est in the analysis of the subsequent data matrixes. By forming an index based on the relation between the number of black data points and the entire number of data points within the region of interest for each data matrix, a time serious of indexes is generated, the decay of which represents the recovery to original state of the tissue and thereby its water content, elasticity and stiffness. Typically the time constants of this recovery curve can be used for the assessment of the tissue water content, tissue elasticity and tissue stiffness. The initial and frequently rapid recovery phase represents the elasticity and stiffness of the tissue, while the second recovery phase represents the translocation of water in tissue and thus its water content. In another embodiment of the invention, the image processing steps may consist of equalizing the image and calculating its Fou- rier-transform in order to evaluate the areas produced by the mechanical depression of the tissue. The person skilled in this technology may readily envision suitable arrangements and other methods of processing the data matrixes. The above discussed process are further illustrated in the detailed part of the specification and its appended drawings.

The polarizing filter in front of the photo-sensitive device preferably has a polarization direction parallel to that of said illuminating light, in order to produce an image of the surface layer of the tissue under investigation alone, thereby making the representation of the tissue impression in the data matrixes more distinct. However, also other ar- rangements would be conceivable to persons skilled in the art.

According to one embodiment of the invention, translocation of the tissue water is attained by attaching a separate device of suitable geometry under occlusion to the tissue surface for a predetermined period of time (Fig.l., upper part), thus producing an im- print in the tissue readily observable when the occlusion is terminated and the device is removed. Said device or part thereof may preferably be made of a disposable part in order to avoid tissue contamination. According to another embodiment of the invention, the device making the imprint in tissue forms an integral part of a transducer system. The complete transducer may be attached to the tissue by double-adhesive tape or coupled mechanically to a holder attached to the tissue by double-adhesive tape. The mechanical device, making the im- print in the skin, may be operated by use of a motor embedded in the transducer system controlled by the computing device. The period of time said mechanic device depresses the tissue and the successive collection of data can be controlled by the user through a user-interface forming an integral part of the computing device. The said mechanic device or part thereof may be fabricated from disposable material in order to avoid contamination.

The present invention is also directed at a system for determining how tissue water content is influenced by dehydration, hyperhydration and environmental factors. The system generally comprises a light source combined with a filter for illuminating a tis- sue surface with polarized light and a mechanical device, part of which may be produced by disposable material, for producing an imprint in the tissue. Said mechanical device may form an integral part of a transducer and be controlled by an embedded motor under user-control through the user interface of the computing device. Further, a polarization filter is used for collecting the backscattered light and a photosensitive array detects the backscattered and polarized light and converts the detected signal to collected information of digital values. A computing device receives the collected information and separates it into specific data matrixes. Threshold techniques may be used to enhance the representation of said imprint in tissue as represented in the data matrixes. These threshold techniques may include the use of regions of interest in the data matrix in order to improve the overall signal-to-noise ratio. An example of a suitable region of interest is the use of the mask formed by the values below threshold in one of the data matrixes. A sequence of indexes is formed by the number of values within the region of interest below threshold in relation to all values within the same region of interest, each representing this information extracted from the individual data matrixes. The multi-phase decay curve of this sequence represents tissue elasticity and stiffness, respectively tissue water translocation and thus water content. The system can further be adapted to cooperate with a mobile communication terminal capable of transmitting the output information over a telecommunication network, such as a mo- bile network or a public fixed network, the Internet. The system can be integrated with a mobile communication terminal as details of a mobile telephone, or form separate units combined with local communication links. Accordingly the output information from the system can be adapted for direct communication with clinical care centres or for immediate analysis by the patient or by health care professionals.

According to a specific aspect of the invention, above mentioned methods are used to determine if a patient or other subject is susceptible to acquire such complications from cardiovascular disease, malignancies, intensive care treatment, surgical procedures, dialysis or by presence in extreme environments that arrive from tissue viscoelasticity property imbalances. A patient or other subject is thereby subjected to controlled administration of fluid, whereupon any of the aforementioned methods are conducted. The resulted output information is compared with a reference obtained from healthy individuals, or from the same subject prior to the controlled administration of fluid.

The following part of the description demonstrates in further detail a preferred way of conducting present invention by way of examples. The skilled person will realize the possibility to deviate from what is exemplified and still operate the inventive scope as earlier outlined.

Detailed and exemplifying part of the description

Fig. 1 shows an arrangement of a system with separate mechanical device for tissue depression (upper part) and device for recording of tissue impression by way of an op- tic technique (lower part). A ) impression in tissue, B) mechanical device for tissue impression, C) occlusive cover, I) illuminating device, J) incident light beam, K) first polarizing filter, L) backscattered light beam, N) second polarizing filter, P) computing device, Q) user interface, R) tissue surface and U) photosensitive array.

Fig. 2 shows an arrangement of a system with integrated mechanical device for tissue depression and device for recording of tissue impression by way of an optic technique. D) controlling mechanical device for tissue impression in lower position, E) housing, F) controlling mechanical device for tissue impression in upper position, G) motor, H) ring-shaped disposable impression device, O) photosensitive array, R) tissue, S) holder and T) tissue impression.

Fig. 3 shows an example of a sequence of images of skin including depressed skin ar- eas captured by use of a digital camera device.

Fig. 4 shows the corresponding sequences of processed data matrixes following the application of threshold techniques. In the leftmost data matrix a suitable region of interest is outlined.

Fig. 5 shows a multi-phase decay curve representing a sequence of indexes, each representing the fraction of areas indicating depressed skin in relation to the entire image.

Fig. 6 shows three representative decay curves with initial slopes for skin of different biological age. A) 7 years, B) 24 years and C) 80 years.

In investigation of tissue elasticity, stiffness and water content, the imprint in the tissue (Fig. 1, item A) is made by applying a mechanical device (Fig 1, item B) to the tissue under occlusion (Fig. 1. item C), for a certain period of time, typically 1 - 2 minutes. Alternatively the imprint in the tissue can be made by way of a mechanical device (Fig 2., item D) forming an integral part of a transducer (Fig 2, item E) applied to the tissue. When the imprint has been made, the mechanical device making the imprint is automatically pulled back to its resting position (Fig. 2, item F) by use of an embedded motor (G), or if being a separate device (Fig 1, item B), removed from the tissue sur- face, after which the capturing of images of the tissue can commence. The mechanical device (Fig 1, item B and Fig 2, item H), which forms the imprint in the tissue is preferably fabricated from a disposable material, thereby avoiding contamination of the tissue. Light from an illuminating device (Fig.l, item I) illuminates (Fig. 1. item J) the tissue surface through a polarizing filter (Fig. 1, item K). The backscattered light (Fig. 1., item L), reaches the photo-sensitive array (Fig. 1., item U) by passing a second polarizing filter (Fig. 1., item I) that has a direction of polarization parallel to that positioned in front of the illuminating device (Fig. L, item T). By this arrangement, light that is diffusely scattered in tissue, thereby becoming random polarized, is effectively blocked from reaching the photo-sensitive array. This implies that in principle only light from surface reflections reaches the photo-sensitive array. In the embodiment of the invention, where the photo-sensitive array forms an integral part of the housing (Fig 2), the photo-sensitive array (Fig.2., item O) is to be understood to incorporate also the means for illuminating the tissue by polarized light and means for detecting only light of a specific polarization direction. These items have been omitted in Fig.2 for the sake of clarity. Typically a digital camera (e.g. Canon IXUS V3, Canon Svenska AB, Solna, Sweden) is used to capture the images of the depressed tissue. The information from the photo-sensitive device is directed to a computing device (Fig. 1, item P) in which the further signal processing is performed. In the example given in Fig 3 - 5,

MATLAB® 6.5.1 (The Math Works, Inc.) was used to process the images. In the first step, an image was colour separated and only the blue colour data matrix was used for further processing. By use of threshold and image processing techniques, individual matrix elements where given the value 255 (white) if they where above threshold, and the value zero (black) if below threshold (Fig. 4.). The layout of the black dots of the first generated matrix was used as a mask for the subsequent data matrixes in the entire sequence of recorded images. In this way only black and white dots within the area defined by this mask were accounted for in the calculation of a data matrix specific index composed of the number of white dots divided by the total number of dots within the mask. As the imprint in the tissue diminishes, the number of data points above threshold (white dots) increases as indicated in Fig. 4 and the index value increases. By using the formula:

Index = A/B

where A = Total number of data points above threshold within the region of interest and B = Total number of data points within the region of interest,

an index could be assigned to each data matrix in Fig. 4. scaling with the number of values below threshold, and thus to the extent of the black areas in corresponding data matrix or mask area within the data matrix. Temporal changes in this index from one data matrix to another reflect the tissue elasticity and stiffness (first part of the multi- phase curve) respectively the tissue water content (second part of the multi-phase curve).

Although the above example of the invention discusses an algorithm based on threshold techniques, several other algorithms may be designed to attain the desired result. In order to avoid the adverse effects of alternating illumination from the illuminating device (Fig 1., item I), a reference-area positioned in the field view of the photo-sensitive device may be used, the backscattered light from which can be used for compensating for alterations in illumination. The skilled person will realize such and other possibi- lities to deviate from what is exemplified and still operate the inventive scope as earlier outlined.

By way of an example (Fig. 6), the initial slope of the multi-phase decay curves for skin recorded from test subjects at different age (7, 24 and 80 year), demonstrates how the skin looses is elasticity with increasing age.

Claims

1. A method of determining the water content, elasticity and stiffness of living tissue based on analysis of the recovery of the tissue following the formation of an mechanical imprint, characterized by :

(i) illuminating a tissue surface with polarized light;

(ii) collecting the backscattered light through a polarizing filter;

(iii) detecting the backscattered and polarized light by a photo-sensitive array;

(iv) transferring the collected information in digital form to a computing device;

(v) processing the collected information by an image feature extraction algorithm to form a data matrix representation of the tissue imprint; (vi) generating a multi-phase output signal by processing corresponding values in said data matrix by an algorithm, wherein said output signal represents a data reduction of said data matrix, thereby obtaining a representation of the tissue water content, elasticity and stiffness.

2. A method according to claim 1 , wherein said polarization filter provides a polarization direction parallel with that of said illuminating light.

3. A method according to claim 1, including producing values for normalization of the data elements of said data matrix by simultaneously illuminating a reference area.

4. A method according to claim 1, wherein said feature extraction algorithm is applied to at least one of the colour planes in said collected information.

5. A method according to claim 1, wherein said feature extracting algorithm includes threshold calculation.

6. A method according to claim 5, wherein said threshold calculation is performed on data elements within a selected region of interest to form the output signal.

7. A method according to claim 6, wherein said region of interest is formed by the below-threshold data elements in one of the said data matrixes to form the output signal.

8. A method according to claim 6, wherein said output signal is calculated from the number of data elements in said region of interest below said threshold in relation to the total number of data elements in said region of interest.

9. A method according to claim 7, wherein said output signal is calculated from the number of data elements in said region of interest below said threshold in relation to the total number of data elements in said region of interest.

10. A method according to claim 1, wherein said feature extracting algorithm includes image equalization and Fourier transform calculation.

11. A method according to claim 10, wherein the energy in the low frequency re- gion of said Fourier transform in relation to the energy in the high frequency region of said Fourier transform is calculated.

12. A system for determining the water content, elasticity and stiffness of living tissue based on analysis of the recovery of the tissue following the formation of an mechanical imprint, characterized by :

(i) a light source (I) and a filter (K) capable of illuminating a tissue surface with polarized light, (ii) a polarizing filter (N) for collecting the backscattered light; (iii) a photo-sensitive array (U) capable of detecting the backscattered and polarized light and converting the detected light to a collected information of digital values; (iv) a computing device (P) receiving said collected information and adapted to separate it into three data matrixes, of which at least one is utilized for feature extraction and to apply an algorithm that generates an output signal representing tissue water content, elasticity and stiffness.

13. A system according to claim 12 comprising means (Q) for presenting said output signal as a representation of the tissue water content, elasticity and stiffness.

14. A system according to claim 12, wherein said polarization filter provides a polarization direction parallel with that of said illuminating light.

15. A system according to claim 12, comprising a reference area for producing a measurement value for normalisation of the data elements of said data matrix.

16. A system according to claim 12, wherein said feature extraction comprises means for calculating a threshold value and the number of data elements in the data matrix above respectively below said threshold value.

17. A system according to claim 16, wherein said means for feature extraction comprises means for generating a region of interest in the data matrix.

18. A system according to claim 17, wherein said means for generating a region of interest comprises means for generating a mask the shape of which relating to said region of interest.

19. A system according to claim 17, wherein said means for generating a region of interest includes means for calculating an output signal from the number of data elements below said threshold value in relation to the total number of data ele- ments within said region of interest.

20. A system according to claim 18, wherein said means for generating a mask includes means for calculating an output signal from the number of data elements below said threshold value in relation to the total number of data elements within said mask.

21. A system according to claim 12, wherein said feature extraction comprises means for equalizing said data matrix and calculate corresponding Fourier transform and its relative energy in the low respectively the high frequency region.

22. A system according to claim 12, wherein said output signal represents data reduction of said processed data matrixes, thereby forming a representation of tissue water content, elasticity and stiffness.

23. A system according to claim 12 adapted to cooperate with a mobile communication terminal capable of transmitting the output signal over a tele- communication network.

24. A system according to claim 23 integrated with a mobile communication terminal.

25. A system according to claim 23 having a separate mobile communication terminal connected to said system with communication link.

26. A method based on analysis of the recovery of a tissue to its normal state following the formation of a mechanical imprint of determining if a patient is susceptible to acquire such complications from dehydration or hyperhydration that arrive from imbalances in the tissue water content comprising or from pathological elasticity and stiffness of a specific tissue characterized by:

(i) illuminating the tissue surface with polarized light; (ii) collecting the backscattered light through a polarizing filter;

(iii) detecting the backscattered and polarized light by a photo-sensitive array; (iv) transferring the collected information in digital form to a computing device; (v) separating the collected information into three data matrixes of which at least one is used for further processing; (vi) extracting features from the data matrix by calculating the number of data elements above or below threshold value or by equalizing the data matrix and calculating its Fourier transform, (vii) calculating an output signal representing the tissue water content, (viii) calculating an output signal representing the tissue elasticity and stiffness.

27. A method according to claim 26 wherein said hyperhydration or dehydration is caused by an inadequate maintenance of the water balance in infants or young children.

28. A method according to claim 26 wherein said hyperhydration or dehydration is caused by an inadequate maintenance of the water balance during surgery or intensive care.

29. A method according to claim 26 wherein said hyperhydration or dehydration is caused by an inadequate maintenance of the water balance during dialysis.

30. A method according to claim 26 wherein said hyperhydration or dehydration is caused by an inadequate maintenance of the water balance in home health care settings.

31. A method according to claim 26 wherein said hyperhydration or dehydration is caused by an inadequate maintenance of the water balance during military and work-related operations in extreme climates.

32. A method according to claim 26 wherein said hyperhydration or dehydration of the tissue is relevant in the development of skin care products.

33. A method according to claim 26 wherein said elasticity and stiffness of the tissue is relevant in the investigation of the prostate gland.

34. A method according to claims 27 - 33, including presenting said output signal as a value representing the tissue water content, elasticity or stiffness.

35. A method according to claim 26, wherein said local polarization filter provides a polarization direction parallel with that of said illuminating light.

36. A method according to claim 26, including producing values for normalization of the data elements of said data matrix by simultaneously illuminating a reference area.

37. A method according to claim 26, wherein said feature extraction algorithm is applied to at least one of the colour planes in said collected information.

38. A method according to claim 26, wherein said feature extracting algorithm includes threshold calculation.

39. A method according to claim 38, wherein said threshold calculation is performed on data elements within a selected region of interest to form the output signal.

40. A method according to claim 39, wherein said region of interest is formed by the below-threshold data elements in one of the said data matrixes to form the output signal.

41. A method according to claim 39, wherein said output signal is calculated from the number of data elements in said region of interest below said threshold in relation to the total number of data elements in said region of interest.

42. A method according to claim 40, wherein said output signal is calculated from the number of data elements in said region of interest below said threshold in relation to the total number of data elements in said region of interest.

43. A method according to claim 26, wherein said equalization of the data matrix is followed by calculation of the Fourier transform of the data matrix, from which the relative energy of the signal in the lower frequency interval in relation to the energy of the signal in the higher frequency interval are calculated.

44. A method according to claim 26, wherein said output signal is extracted from said extracted features to form an output signal that represent the tissue elasticity, stiffness and water content.