HK1158482B - Locating and analyzing perforator flaps for plastic and reconstructive surgery - Google Patents

Description

Locating and analyzing a perforator flap for plastic and reconstructive surgery

Cross Reference to Related Applications

This application claims priority to U.S. provisional application serial No. 61/044,779, filed on 14/4/2008, incorporated herein by reference in its entirety.

Background

Reconstructive plastic surgery often requires the positioning and clinical evaluation of flaps of skin and subcutaneous tissue that are supplied by isolated perforator vessels and that are potentially suitable for implantation into another part of the body. Perforator branches (perforators) pass from their source vessels to the skin surface through or between deep muscle tissue. Well-distributed skin flaps (flaps) are good candidates for transplantation.

For example, the abdominal donor area skin flap has become the standard for autologous breast reconstruction since the 80's of the 18 th century. Within the abdomen, free fat is selectively distributed in a range from a complete Transverse Rectus Abdominis (TRAM) flap to an isolated perforator flap such as a deep inferior abdominal aorta (DIEA) perforator flap. The perforator flap also allows for the transplantation of the patient's own skin and fat in other areas of tissue reconstruction in a reliable manner with minimal donor area morbidity. Flaps that rely on a random pattern of blood supply are quickly replaced by pedicle shaft type flaps that can reliably transplant large amounts of tissue. The advent of free tissue transplantation allows an even greater range of possibilities for properly matching donor and recipient areas. The increasing use of perforator flaps has led to a rapidly increasing need for preoperative familiarity with the individual DIEA and the specific anatomical features of its perforator, especially given the significant variation in anatomical models of vascular supply to the abdominal wall (anatomy).

The positioning and evaluation of the perforator is a laborious and time consuming process. Preoperative Computed Tomography Angiography (CTA) imaging is often performed for localization. Such an approach requires considerable expense and has the additional complexity that the surgeon must mentally associate images from a previously acquired 3D modality (modality) with the current 2D view of the patient lying on the operating table at that time. Thus, the search for more favorable imaging modalities continues, and recently of interest is the use of indocyanine green (ICG) fluorescence imaging, in which blood circulation through the skin is assessed based on fluorescence signals. Fluorescence with an emission peak in ICG of approximately 830nm occurs as a result of excitation by radiation in the near infrared spectral range. The excitation light having a wavelength of about 800nm may be generated, for example, by a diode laser, a Light Emitting Diode (LED), or other conventional illumination sources such as arc lamps, tungsten halogen lamps with suitable band pass filters. The skin is transparent to this wavelength.

ICG binds strongly to blood proteins and has been used previously for cardiac output measurements, liver function assessment, and ocular angiography with little adverse reaction. Evaluation of ICG fluorescence signals can be used to locate the perforator. Since the skin surface near the perforator normally accumulates more blood and at a faster rate than the surrounding tissue, the perforator, once injected with ICG, tends to fluoresce brighter and faster than the surrounding tissue. This rapid, high intensity fluorescence allows visual localization of the perforator. However, surgeons are often interested not only in positioning, but also in evaluation and comparison to support making good clinical decisions. The surgeon needs to determine which of several penetrations is the best implant candidate. Here, simple visual observation while fluorescence is rapidly accumulated and dissipated is not sufficient. For example, the tendency of residual ICG from successive injections to accumulate in the tissue and gradually increase background brightness with each injection further confounds the easy visual discrimination of the best candidate perforator. In addition, ICG sometimes moves extremely slowly over several minutes, making such dynamic analysis very challenging and subjective. The surgeon will evaluate by proposing the following questions:

1) how much ICG-bound blood is in the tissue?

2) How long ICG-bound blood stays in the tissue?

3) How quickly blood bound to ICG moves through tissue?

4) In what order the anatomical areas light up after the bolus?

These questions are difficult to answer on a subjective basis. Accordingly, there is a need for more advanced image processing and display methods to apply objective criteria to locate and assess punch-outs.

Disclosure of Invention

In accordance with one aspect of the present invention, a method for preoperative identification of perforator vessels for plastic and/or reconstructive surgery using ICG fluorescence angiography imaging is disclosed that includes time-resolved image processing for highlighting perforator locations and allowing visual discrimination among candidate perforators through various calculated metrics. The surgeon can select and compare the results of the algorithm used to analyze the time series and output the metrics according to at least one of the following processing actions: the time-integrated fluorescence is determined on a pixel-by-pixel basis.

The average fluorescence is calculated by dividing the time-integrated fluorescence by the elapsed time.

The rate of increase/elution (wash-out) in fluorescence was determined.

The elapsed time was determined to obtain the peak fluorescence.

The various image processing steps process the image pixels independently and compute a unique numerical measure for each pixel in the input sequence observed across the entire acquisition time or selected temporal subrange. Thus, each image output is an array of values having the same dimensions (i.e., number and arrangement of pixels) as the frames in the input image sequence. Thus, the processed image may be displayed, for example, as a three-dimensional representation of pixel values calculated across the imaged region, such as a contour map, or as a color-coded two-dimensional image or relief map. In such a case where the punch-through location is under the skin, such an image representation facilitates a quick understanding of the image features and a comparison between the areas on the image.

These and other features and advantages of the present invention will be more readily understood from the following detailed description of the invention.

Drawings

The following figures depict certain illustrative embodiments of the present invention, in which like reference numerals refer to like elements. These depicted embodiments should be understood as illustrative of the invention and not as limiting the invention in any way.

FIG. 1 schematically shows a camera system for observing ICG fluorescence;

FIG. 2 shows an ICG fluorescence image of a skin region, in which pixel values are integrated over time;

FIG. 3 shows an ICG fluorescence image of a skin region integrated over time, with pixel values inversely weighted by elapsed time;

FIG. 4 shows an ICG fluorescence image of a skin region, where the pixel values are determined by the rate of increase of fluorescence;

FIG. 5 shows an ICG fluorescence image of a skin region where the pixel values are determined by the elapsed time to maximum fluorescence; and

FIG. 6 shows an ICG fluorescence image of a skin region where the pixel values are determined by the peak fluorescence;

FIG. 7 shows an overlay of a fluorescence image processed with a variable contrast transfer function; and

FIG. 8 shows an overlay of a fluorescence image processed with another variable contrast transfer function.

Detailed Description

The present invention is directed to preoperative determination of the location of a perforator vessel in a perforator flap by a non-invasive method prior to any incision.

Fig. 1 schematically shows an apparatus for non-invasive, percutaneous determination of tissue perfusion in surgical applications, particularly preoperative applications, by ICG fluorescence imaging. An infrared light source, such as one or more diode lasers or LEDs, having a peak emission of approximately 780-800nm for exciting fluorescence in ICG is located inside the housing 1. The fluorescence signal is detected by the CCD camera 2 having sufficient sensitivity to near infrared rays; such cameras are commercially available from several vendors (Hitachi, Hamamatsu, etc.). The CCD camera 2 may have a viewfinder 8, but during operation may also observe images on an external monitor, which may be part of an electronic image processing and evaluation system 11.

The light beam 3 may be a diverging beam or a scanned beam, which is emitted from the housing 1 to illuminate a region of interest 4, i.e. a region where a flap with a suitable perforator is expected to be located. The region of interest may be about 10cm x 10cm, but may vary based on surgical requirements and available illumination intensity and camera sensitivity.

A filter 6 is typically placed in front of the camera lens 7 to block excitation light from reaching the camera sensor while allowing fluorescence light to pass. The filter 6 may be a NIR long-wave pass filter (cut-off filter) transparent only to wavelengths greater than about 815nm, or preferably a band-pass filter, which transmits at a peak wavelength between 830 and 845nm and has a Full Width Half Maximum (FWHM) transmission window between about 10nm and 25nm, i.e. outside the excitation wavelength band. The camera 2 may also be designed for acquiring a color image of the region of interest, to allow real-time correlation between the fluorescence image and the color image.

In the context of the present invention, the device shown in fig. 1 is used to identify/locate a perforator vessel prior to surgery-this will help the surgeon select the best flap or flap partition for use during reconstruction.

In other post-operative applications, the device may be used to:

the patency of the anastomosis and the arterial-to-venous blood flow were verified-this could potentially improve the access (outome) to eliminate skin flap necrosis that could be the result of poor arterial blood flow and inadequate perfusion, as well as poor venous blood return leading to hyperemia.

Visualization and confirmation of complete tissue perfusion, since microvascular perfusion of the entire flap and native tissue is critical to flap survival.

With the present invention, the locations of the perforator branches are visualized through image processing and visualization techniques to allow easy and objective visual discrimination among candidate perforator branches. ICG was injected and the entire ICG fluorescence perfusion and elution cycle was captured by the imaging device. After image acquisition, the entire sequence or a certain time sub-range of images is processed with an image processing algorithm, which may be selected by the surgeon.

For example, the processed fluorescence measurements may be visualized as false color images or contour maps to allow rapid visual evaluation according to the applied algorithmic metrics. For example, the fluorescence intensity of each pixel may be rendered into a spectral color that changes from blue (a "cold" spot or low fluorescence intensity or rate) to red (a "hot" spot or high fluorescence intensity or rate). Other spectral associations can also be easily complied with. The output may be presented as a semi-transparent overlay over the original anatomical image. This allows visual association of the "hot" point with the underlying anatomical model. The meaning of a "hot" spot varies with the algorithm employed, such as integrated intensity (weighted or unweighted), rate of increase or elution.

The user is given interactive control over the "hot" to "cold" color mapping and can change it in real time to explore the finer or coarser subranges of the dynamic range of the output metrics of each algorithm. When the color window is widened, the hottest areas are highlighted first, followed by the cooler areas. This type of adjustment may be made by changing a mapping of luminosity or contrast between the captured pixels and the pixels in the displayed image. Such a mapping function may be included in a standard imaging procedure. This windowing process based on currently employed metrics helps to discern punch-outs and enhance perception and improve the surgeon's understanding of the dynamics of the applied ICG.

The present invention also supports the simultaneous display and evaluation of two sequences from two different locations on the skin of a patient. This allows comparison of candidate skin flaps separated by a distance greater than the field of view of the imaging system.

Fig. 2 shows an image of a region of the skin of a patient where a suitable perforator vessel is to be identified. Each pixel represents the time integral of the fluorescence intensity over the exposure time of the image sequence. This mode is commonly referred to as "integration mode" in image processing, and many image processors provide this mode as a standard feature. In practice, for example in an image processor, the pixel intensities (collected charge in the CCD) acquired during each frame in the image sequence are summed on a pixel-by-pixel basis and divided by the number of frames, and the sum may then be normalized to a fixed dynamic range, for example from 1 to 255(8 bits). The concept is that brighter pixels in the image represent an area of skin perfused with a greater amount of blood carrying ICG over a preset period of time. In fig. 2, a perforator 24 exhibits the highest integrated fluorescence intensity, while another perforator exhibits a weaker fluorescence intensity, as shown at 26.

It should be noted that the transparency of the image has been set so that the doctor's indicium 22 is visible on the upper right of the screen through the transparent color overlay of the ICG fluoroscopic image.

Fig. 3 shows ICG fluorescence images of the same skin area integrated over time, with pixel values inversely weighted by elapsed time. This image processing algorithm is similar to the integration described previously, but instead of adding the measured intensity of each pixel directly, the measured intensity values are first divided by the elapsed time after the start of the observation of the ICG fluorescence before being added. In this way, earlier fluorescence signals are given greater effectiveness than later acquired fluorescence signals. The "hottest" pixels are those that fluoresce earlier in the sequence of image frames than other pixels where the ICG bolus arrives at a later time. The same perforator vessel 34 as in fig. 2 is identified, while the other vessel 36 is hardly identifiable.

Fig. 4 again shows an ICG fluorescence image of the same skin area, where the pixel values in this image are determined by the rate of increase of the fluorescence intensity. In this image processing algorithm, the slope of the pixel intensity versus elapsed time is calculated for each pixel in the image. For example, each pixel may have a given lowest intensity value (baseline) and a given highest intensity value (or another relatively high intensity value). For each pixel, the time at which the pixel intensity crosses the baseline and the time at which the pixel intensity crosses the high intensity value are recorded. From this information, the image processing algorithm calculates an increase rate for each pixel in the image, where the "hotter" pixels have a greater slope, i.e., they reach a high intensity value faster than the "colder" pixels. Thus, this embodiment of the image processing algorithm highlights the speed at which the ICG bolus reaches the perforator. In fig. 4 the transparency is closed so that the surgeon's tool is not visible in the image.

The previously identified perforator vessel is shown here with reference 44 and is much better defined, as is the vessel 46 (previously shown as 26 and 36) and the further vessel 48.

Fig. 5 shows an ICG fluorescence image of the same skin area, with the pixel values determined by the elapsed time to maximum fluorescence. Unlike fig. 4, which shows the time rate of change, the image processing algorithm of fig. 5 shows the time at which pixels reach their maximum intensity, with "hotter" pixels reaching their respective peak fluorescence intensities earlier than colder pixels. Thus, the algorithm highlights regions of the image in order of their penetration to their peak intensity. In this image, the previously identified perforator vessels 24, 34, 44 are also clearly discernable, as are vessels 56 and 58 corresponding to vessels 46 and 48 of fig. 4.

Fig. 6 shows an ICG fluorescence image of the same skin area, with the pixel values determined by the maximum peak fluorescence value at each pixel. A higher ("hot") fluorescence intensity value 64 may indicate a higher ICG concentration or may be caused by a perforator vessel located closer to the skin surface, which may reduce the absorption/fluorescence response of the excitation light. The blood vessels 66, 68 clearly visible in fig. 4 and 5 are hardly distinguishable from the background.

While images such as those shown in fig. 2 and 6 are rendered with a linear contrast transfer function that provides a 1: 1 mapping of pixel values processed with the various algorithms described above to displayed pixel intensities, images may also be rendered with a variable contrast transfer function (either as a contour map or a false color overlay) to enhance parallax in the image. In addition, a label may be placed in the overlay image, referred to hereinafter as an ACR (cumulative or time-integrated intensity ratio) label, which facilitates quantitative comparisons between two or more regions of the anatomical model.

Since the absolute pixel values in the image change as the dynamic range and slope of the variable contrast transfer function are modified, the ACR label allows the user to compare the relative perfusion in different image regions measured according to any of the selected overlay techniques (e.g., cumulative/time-averaged intensity, etc.).

The following method is used to calculate the ACR tag value. For clarity, we assume that the cumulative intensity is chosen as the coverage technique, but the same method can be used with any of the available coverage techniques.

1) The cumulative intensity of all pixels of all images in the sequence of images is calculated over a time window.

2) The cumulative intensity is averaged over the area of the selected label (e.g., a square matrix of 5 x 5 pixels).

3) The averaged intensities are normalized to the maximum of the cumulative intensity in the entire image.

4) The normalized average value is scaled (scale), wherein the maximum value of the transfer function represents 100%.

By following this approach, the relative ratio of two different ACR tags remains unchanged even if the slope of the transfer function is modified. Fig. 7 and 8 show, for two different contrast functions, a fluorescence image (upper part of a grayscale image) from a sequence of images that has been processed with one of the aforementioned algorithms and a false color overlay image that renders the integrated intensity from the sequence in color (from blue for low values to red for high values). Processing the pixel values in fig. 7 with the first contrast transfer function, two regions with intensities of 52% and 72%, respectively, are known, corresponding to a ratio between these two labeled regions of 52/72 ═ 0.72. The second overlay image in fig. 8 shows the same pixel values processed with different contrast transfer functions, where the intensities in the two regions are now labeled 99% and 71%, respectively. However, their relative ratio remains approximately constant at 71/99-0.72.

The user may modify the transfer function such that the control region is labeled with 100%, where all other regions may then be compared to the control region.

Fig. 9 shows that the overlay is transparent where the pixel has a cumulative intensity value less than the point where the bottom of the transfer function slope intersects the horizontal pixel value axis. Furthermore, this indicates that in this example, 12% of the image area (with the coverage number at the bottom right of the bottom window) has a cumulative intensity that is greater than 52% of the maximum cumulative intensity. The diagram shows several regions bounded by their 52% contours.

The described embodiments detect the fluorescent signal emitted by ICG through the skin after excitation in the near infrared spectral range. However, it will be appreciated by those skilled in the art that other dyes that can be excited and emit fluorescence in the spectral range of tissue transmitted light may also be used.

Although the present invention has been described with reference to an example of arterial blood flow (i.e., the blood supply to the perforator vessel (s)), the method may also detect graft failure due to venous congestion by quantifying and displaying the rate of change from peak intensity down to baseline. This will highlight venous return in the perfused area.

While the invention is susceptible to various modifications and alternative forms, specific examples thereof have been shown in the drawings and are herein described in detail. It should be understood, however, that the invention is not to be limited to the particular forms or methods disclosed, but to the contrary, the invention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the appended claims.

Claims (9)

1. A method for evaluating a perforator in a candidate perforator flap, the method comprising the steps of:

detecting an ICG fluorescence response from blood-borne ICG in a perforator vessel in the candidate perforator flap;

acquiring a sequence of time images of the fluorescent response over a predetermined time;

processing the sequence of images to generate a time-integrated intensity of pixel values in the sequence of images or a time derivative of the time-integrated intensity corresponding to an ICG carried by blood in a perforator vessel in the candidate perforator flap; and

displaying the time-integrated intensity or the time-derivative of the time-integrated intensity as a color or black-and-white image;

wherein the method further comprises applying a contrast transfer function to the time-integrated intensity or a time derivative of the time-integrated intensity;

wherein the contrast transfer function represents a linear or non-linear function that converts the time-integrated intensity or the time-derivative of the time-integrated intensity into an overlaid image representing different perfusion characteristics in different colors;

wherein the contrast transfer function is a non-linear function of regions having different slopes, and wherein the different slopes and a transformation between the different slopes may be selected by a clinician during a procedure of assessing perfusion of the blood vessel.

2. The method of claim 1, wherein the sequence of images is processed by dividing the time-integrated intensity value for each pixel by an elapsed time after the fluorescence response from the tissue is detected.

3. The method of claim 1, wherein the sequence of images is processed by determining an elapsed time after a fluorescence response from tissue is detected until a time-integrated fluorescence intensity for each pixel reaches a peak.

4. The method of claim 1, wherein the sequence of images is processed by determining a peak in fluorescence intensity for each pixel.

5. The method of claim 1, further comprising displaying the time-integrated intensity or a value of a time derivative of the time-integrated intensity in the overlay image.

6. The method of claim 1, further comprising calculating a time-integrated intensity ratio of different image regions of an anatomical feature.

7. The method of claim 6, wherein the time-integrated intensity ratio is calculated by:

calculating the time integral intensity of all pixels;

averaging the time-integrated intensity over a predefined area in the overlay image;

normalizing the time-integrated intensity to a maximum value of the time-integrated intensity in the entire coverage image; and

scaling the normalized intensity with a maximum value of the contrast transfer function.

8. The method of claim 7, wherein the predefined area in the overlay image is a square matrix of 5 x 5 pixels.

9. A device for evaluating a perforator in a candidate perforator flap, the device comprising:

means for detecting a fluorescence response from blood-borne ICG in a perforator vessel in the candidate perforator flap;

means for acquiring a temporal image sequence of the fluorescence response over a predetermined time;

means for processing the sequence of images to generate a time-integrated intensity of pixel values in the sequence of images or a time derivative of the time-integrated intensity corresponding to an ICG carried by blood in a perforator vessel in the candidate perforator flap; and

means for displaying the time-integrated intensity or the time-derivative of the time-integrated intensity as a color or black-and-white image;

wherein the apparatus further comprises means for applying a contrast transfer function to the time-integrated intensity or a time derivative of the time-integrated intensity;

wherein the contrast transfer function represents a linear or non-linear function that converts the time-integrated intensity or the time-derivative of the time-integrated intensity into an overlaid image representing different perfusion characteristics in different colors;

wherein the contrast transfer function is a non-linear function of regions having different slopes, and wherein the different slopes and a transformation between the different slopes may be selected by a clinician during a procedure of assessing perfusion of the blood vessel.