We employed a Pix2Pix generative adversarial network to translate the multispectral fluorescence images into colored brightfield representations resembling H&E staining. The model underwent training using 512x512 pixel paired image patches, with a manually stained image serving as the reference and the fluorescence images serving as the processing input. The baseline model, without any modifications, did not achieve high microscopic accuracy, manifesting incorrect color attribution to various biological structures and the addition or removal of image features. However, through the substitution of simple convolutions with Dense convolution units in the U-Net Generator, we observed an increase in the similarity of microscopic structures and the color balance between the paired images. The resulting improvements underscore the potential utility of virtual staining in histopathological analysis for veterinary oncology applications.
Skin and subcutaneous tumors are common in companion animals, that can be difficult to diagnose and treat. Raman spectroscopy shows high diagnostic accuracy in identifying malignant tumors and benign lipomas in dogs and cats. However, the traditional single-point raster scanning approach is not ideal for a large-field-of-view Raman imaging due to its time-consuming nature when scanning areas larger than a square centimeter. Additionally, focusing the excitation spot can lead to high levels of light fluency (J/cm2), potentially causing damage to tissue biomolecules. Furthermore, the resulting raster-scan image often lacks sufficient spatial resolution to effectively compare it with tissue morphology findings. In this study, we focused on implementing EMCCD camera-based Raman imaging to accurately capture Raman spectral band signatures and overcome autofluorescence interference in veterinary cancer samples ex vivo. By utilizing the tunable band-pass filters set-up, our system enables large-field-of-view imaging of specific Raman bands, such as the 1437 cm-1 band or 1652 cm-1 band in biological tissue, proposing a more efficient, accurate and safe approach for Raman imaging in the veterinary field.
Skin and subcutaneous tumors are widespread in dogs and cats. Current tumor diagnostics (e.g., biopsy, fineneedle cytology) is invasive and labor-consuming. In this work, we studied ex vivo the most common canine and feline tumor OCT images using sliding window analysis and linear SVC classification, and we compared different sliding window sizes to determine the most optimal window sizes when differentiating between skin, mast cell tumours and soft tissue sarcomas. Sensitivities and specificities of all tissue classes saw an increase with increasing window size at small window size values and plateaued at around 60-80 μm, indicating the most significant tissue structures for differentiation via SWA likely lay here. Our work is the first veterinary OCT study on multiple canine and feline skin tumors to optimize the sliding window size for image pattern analysis.
Cancer is one of the leading causes of companion animal mortality. Up to 30% of all canine and feline tumors appear on or directly under the skin. To date, only a limited number of studies applied biophotonics techniques for optical characterization and detection of tumors in pets. In this work, we acquired ex vivo optical coherence tomography (OCT) images and Raman spectra of native skin and the most common canine and feline skin and subcutaneous tumors; lipomas, mast cell tumors, and soft tissues sarcomas. Lipomas exhibited the most distinctive tissue morphology (i.e., honeycomb structure) and biochemistry (lipid-related Raman peaks of 1063, 1301, and 1652 cm-1). Moreover, lipomas had significantly higher values of coefficient of variation (CV) retrieved from OCT images. On the other hand, all other tissues exhibited signal-dense and highly scattering OCT images. Despite the similar Raman spectra, we detected the malignant tumors with the sensitivity and specificity of 100% and 88.2%, respectively. Additionally, malignant tumor types were distinguished with an accuracy of 78.6%. Our results showed the potential of OCT and Raman techniques for ex vivo optical characterization of common canine and feline tumors and native skin.
Canine atopic dermatitis (CAD) is a common inflammatory and pruritic skin disease associated with allergy to exogenous allergens. The regular monitoring of skin lesions is essential to execute the anti-allergic therapy successfully. Erythema is one of the most important CAD-related lesions since it represents acute skin inflammation. Previously, we studied two optical systems (i.e., multispectral and dermatoscopic devices), which could objectively estimate erythema severity. However, we did not investigate, which image sampling method for selecting erythema-representing pixels and erythema index (EI) are correlated with the visual erythema assessment the most. In this paper, we tested three image sampling methods and four EIs for erythema severity estimation in 43 dogs. We discovered that all studied sampling methods and EIs were strongly correlated (r> 0.58) with the visual CADESI-4-based erythema severity assessment. However, the highest average Spearman’s correlations coefficient r of 0.77 was achieved when the average pixel value from the two small squared image sections without the hair and pigment was considered. On the other hand, EI, which was calculated from all three RGB values, achieved the highest r of 0.78. In this study, we identified a reliable image sampling method and erythema index (EI), which are well correlated with the visual erythema estimation.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.