In-vivo fluorescence and reflectance imaging of human cervical tissue

Gustafsson, Ulf; McLaughlin, Elisabeth; Jacobsen, Ellen; Hakansson, Johan; Troy, P.; DeWeert, Michael J.; Svanberg, Katarina; Pålsson, Sara; Thompson, Marcelo Soto; Svanberg, Sune; Vaitkuvienė, Aurelija

doi:10.1117/12.480413

Cited by 10 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, this study was based on single point detection of tissue spectra and did not involve global imaging information of the tissue. Gustafsson et al acquired the hyperspectral reflectance and fluorescence data cube of the entire cervical tissue separately, and demonstrated the spectral difference between different lesion types [19]. However, such a multimodal imaging system was expensive and the image coregistration was time consuming, inappropriate for deployment in developing countries and rural regions of limited resources.…”

Section: Introductionmentioning

confidence: 99%

Development of a Multimodal Colposcopy for Characterization of Cervical Intraepithelial Neoplasia

Ren

Pei

et al. 2017

Journal of Medical Devices

View full text Add to dashboard Cite

To develop and evaluate the clinical application of a multimodal colposcopy combining multispectral reflectance, autofluorescence, and red, green, blue (RGB) imaging for noninvasive characterization of cervical intraepithelial neoplasia (CIN). We developed a multimodal colposcopy system that combined multispectral reflectance, autofluorescence, and RGB imaging for noninvasive characterization of CIN. We studied the optical properties of cervical tissue first; then the imaging system was designed and tested in a clinical trial where comprehensive datasets were acquired and analyzed to differentiate between squamous normal and high grade types of cervical tissue. The custom-designed multimodal colposcopy is capable of acquiring multispectral reflectance images, autofluorescence images, and RGB images of cervical tissue consecutively. The classification algorithm was employed on both normal and abnormal cases for image segmentation. The performance characteristics of this system were comparable to the gold standard histopathologic measurements with statistical significance. Our pilot study demonstrated the clinical potential of this multimodal colposcopic system for noninvasive characterization of CIN. The proposed system was simple, noninvasive, cost-effective, and portable, making it a suitable device for deployment in developing countries or rural regions of limited resources.

show abstract

Section: Introductionmentioning

confidence: 99%

Development of a Multimodal Colposcopy for Characterization of Cervical Intraepithelial Neoplasia

Ren

Pei

et al. 2017

Journal of Medical Devices

View full text Add to dashboard Cite

show abstract

“…We conducted a technology exploration project for the development of a CAD system for colposcopy, prototyping image processing algorithms for glare removal, anatomic features detection, acetowhite region detection, lesion margin shape analysis, and blood vessel mosaic and punctuation structure detection using RGB images from 111 human subjects participating in a clinical study of our Hyperspectral Diagnostic Imaging (HSDI TM ) instrument 10,11 . Here, we present the details and preliminary results of the glare removal algorithm for single RGB color images of the cervix that can be used as a pre-processing step in CAD systems.…”

Section: Introductionmentioning

confidence: 99%

Automatic glare removal in reflectance imagery of the uterine cervix

Lange¹

2005

SPIE Proceedings

View full text Add to dashboard Cite

Colposcopy is a diagnostic method used to detect cancer precursors and cancer of the uterine cervix. Computer-AidedDiagnosis (CAD) for colposcopy is a new field in medical image processing. Colposcopists analyze glare (glint or specular reflection) patterns on the cervix to assess the surface contour (3D topology) of lesions, an important feature used to evaluate lesion severity. However, glare in the imagery presents major problems for automated image analysis systems. Glare eliminates all information in affected pixels and can introduce artifacts in feature extraction algorithms, such as acetowhite region detection. Although cross-polarization filters can be used to eliminate glare, the reality is that we have to deal with glare when we want to use existing cervical image databases or use an instrument that does not provide cross-polarized imagery. Here, we present the details and preliminary results of a glare removal algorithm for RGB color images of the cervix that can be used as a pre-processing step in CAD systems. The algorithm can be extended to multispectral and hyperspectral imagery. The basic approach of the algorithm is to extract a feature image from the RGB image that provides a good glare to background ratio, to detect the glare regions in the feature image, to extend the glare regions to cover all pixels that have been affected by the glare, and to remove the glare in the affected regions by filling in an estimate of the underlying image features. In our current implementation we use the green (G) image component as the feature image, given its high glare to background ratio and simplicity of calculation. Glare regions are either detected as saturated regions or small high contrasted bright regions. Saturated regions are detected using an adaptive thresholding method. Small high contrasted bright regions are detected using morphological top hat filters with different sizes and thresholds. The full extent of the glare regions is estimated by using a morphological constraint watershed segmentation to find the contour of the glare regions and adding a constant dilatation. The image features are estimated by interpolating the R,G,B color components individually from the surrounding regions based on Laplace's equation and modifying the intensity (I) component of the HSI color space transformed image. As the glare pattern is important to the physician, we embed it as a just visible intensity texture that does not affect the image processing. The performance of the algorithm is demonstrated using human subject data.

show abstract

“…These tradeoffs, albeit successful in form and with great diagnostic potential, introduce certain drawbacks, namely the extended time necessary to image the whole scene and the resultant low resolution image with limited intuitive value to a human interpreter (such as a physician for medical applications). The HyperSpectral Diagnostic Imaging System (HSDI ® ) instrument [1] is an example of an HSI system exploiting the spatial resolution vs. spectral and FOV vs. scanning tradeoff. HSDI ® -type systems have been found useful when applied to colposcopy, the inspection of the female lower genital tract, especially the cervix, for indications of cancerous or precancerous tissue [2], albeit with relatively low spatial resolution and long scan time.…”

mentioning

confidence: 99%