Surgery – Diagnostic testing – Detecting nuclear – electromagnetic – or ultrasonic radiation
Reexamination Certificate
1999-10-19
2002-01-29
Lateef, Marvin M. (Department: 3737)
Surgery
Diagnostic testing
Detecting nuclear, electromagnetic, or ultrasonic radiation
C600S478000, C606S001000, C606S002000, C606S017000, C606S018000, C607S001000, C607S088000, C607S089000, C359S483010, C359S485050, C359S489040, C356S005140, C356S323000, C356S324000, C356S327000, C356S364000
Reexamination Certificate
active
06343228
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to a method and apparatus for the imaging of tissue using fluorescence techniques, and in particular for the endoscopic imaging of tissue in the diagnosis of various diseases such as cancer,
BACKGROUND OF THE INVENTION
Light induced fluorescence (LIF) imaging techniques have been proposed as an effective and non-invasive method to diagnose diseased tissue, especially for example the early stages of cancer. Autofluorescence from tissue is emitted from endogenous fluorophores when they are excited by low power short wavelength light. Such tissue autofluorescence can be used to determine biochemical and bio-morphological changes in the tissue. In particular studies have shown that the fluorescence yields of lesions in their early stages are normally lower than for surrounding healthy tissue. As a consequence LIF techniques have considerable potential for use as diagnostic tools.
The practical application of LIF techniques as effective diagnostic tools has, however, proved harder to achieve. One commonly used technique for in vivo study of tissue fluorescence uses a multiple optical fiber sensor to deliver the excitation light and to collect the fluorescence signal. The distal tip of the sensor is gently touched to the tissue surface to ensure that the fluorescence excitation and collection geometry is the same at different sites. The sensor is then moved from site to site to take the necessary readings. This point-by-point diagnosis has been successfully used for the detection of early cancers at many organ sites based on the contrast in fluorescence yield between healthy tissue and lesions. However, it is also time consuming and not practical for the examination of large tissue areas in clinical practice.
An imaging technique that enabled a relatively large tissue area to be imaged would be desirable. However, in the examination of a large area by an imaging device, the recorded fluorescence power will be strongly affected by the geometry of the excitation and collection system. For example, the separation of the source-sample-detector, the incident/emission angles and any irregularities in the sample surface will all have a large effect on the measurements making it very difficult to identify the fluorescence variations caused by biological changes in the tissue itself. These difficulties are particularly severe when internal tissues are being imaged in in vivo techniques using, for example, an endoscopic system.
A number of attempts have been made to overcome this geometrical difficulty. In one non-imaging approach the fluorescence signals were normalized to reflection signals taken from exactly the same sites. The results of this technique show that geometrical effects can be corrected, but that artifacts caused by the specular reflection of the tissue surface cause many false positives.
Another approach is to create a mathematical model of the geometrical effects which can then be used to correct the measured readings. However, the fluorescence power is a function of the emission angel and with a large image area of an irregular surface the emission angle will vary over a wide range and this cannot be compensated for mathematically without knowing the precise nature of the tissue surface, which is clearly completely impractical.
Another possibility is a digital image processing method in which the raw fluorescence image is normalized to a reference image which is the raw image processed by a moving average algorithm. This method can correct for geometrical errors providing that the moving average algorithm is carefully chosen. However, the validity of this method is dependent on the ratio of the size of the lesion over the imaged area and the degree of non-uniformity of the fluorescence excitation and collection geometry. To ensure the filtering out of lesions and the keeping of the excitation and collection nonuniformities in the reference image area, the algorithm requires that the lesion width is much smaller than the imaged area. The variation of fluorescence caused by geometrical effects must also be assumed to be a slowly varying function over the tissue surface because the algorithm cannot distinguish a lesion from geometrical artifacts of which the spatial frequency distribute in the same region as the lesion.
SUMMARY OF THE INVENTION
According to the present invention there is provided apparatus for imaging the autofluorescence yield of a sample, comprising: (a) means for illuminating and exciting an area of said sample to stimulate autofluorescence, (b) means for forming a fluorescence image of the illuminated area, (c) means for forming a cross-polarized reflection image of the area, and (d) means for producing an output image by normalizing the fluorescence image to said cross-polarized image.
Preferably the output producing means forms a ratio image of the fluorescence image and the cross-polarized image.
In a preferred embodiment the apparatus comprises means for detecting the fluorescence image and the cross-polarized image and means for digitally processing the detected images. For processing the images the apparatus may comprise computer means provided with a frame grabber. Preferably the computer means may take an average of multiply detected images.
The apparatus is preferably incorporated as an endoscope and preferably comprises two optical channels, a first channel for providing light for illuminating and stimulating an area of tissue, and a second channel for collecting fluorescence and reflected light. The apparatus may include means for linearly polarizing the illuminating and stimulating light, and may further comprise imaging optics for forming fluorescence and reflection images from light collected by the second channel. The apparatus most preferably includes a cross-polarizer linearly polarized at an angle of 90° to the illumination to collect the cross-polarized reflection image.
In one embodiment the first optical channel may extend along the central optical axis of an endoscope and the second channel may be annular in cross-section and may surround the first channel. Alternatively the first and second channel may be formed adjacent one another extending parallel to the central axis of an endoscope.
Viewed from another broad aspect the present invention provides a method for imaging the autofluorescence yield of a sample, comprising: (a) illuminating and exciting an area of said sample to stimulate autofluorescence, (b) forming a fluorescence image of said illuminated area, (c) forming a cross-polarized reflection image of the said area, and (d) producing an output image by normalizing the fluorescence image by the cross-polarized image.
Preferably the normalizing step comprises forming a ratio image of the fluorescence image and the cross-polarized image. In an embodiment of the invention digital images may be formed and are digitally processed. The images may be processed by a computer provided with a frame grabber. Preferably multiple images of the same area are obtained and are averaged.
REFERENCES:
patent: 6104939 (2000-08-01), Groner et al.
patent: 6104945 (2000-08-01), Modell et al.
Lateef Marvin M.
Lin Jeoyuh
The Hong Kong University of Science and Technology
LandOfFree
Method and apparatus for fluorescence imaging of tissue does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for fluorescence imaging of tissue, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for fluorescence imaging of tissue will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2827082