Simultaneous sparsity model for histopathological image representation and classification.

The multi-channel nature of digital histopathological images presents an opportunity to exploit the correlated color channel information for better image modeling. Inspired by recent work in sparsity for single channel image classification, we propose a new simultaneous sparsity model for multi-channel histopathological image representation and classification (SHIRC). Essentially, we represent a histopathological image as a sparse linear combination of training examples under suitable channel-wise constraints. Classification is performed by solving a newly formulated simultaneous sparsity-based optimization problem. A practical challenge is the correspondence of image objects (cellular and nuclear structures) at different spatial locations in the image. We propose a robust locally adaptive variant of SHIRC (LA-SHIRC) to tackle this issue. Experiments on two challenging real-world image data sets: 1) mammalian tissue images acquired by pathologists of the animal diagnostics lab (ADL) at Pennsylvania State University, and 2) human intraductal breast lesions, reveal the merits of our proposal over state-of-the-art alternatives. Further, we demonstrate that LA-SHIRC exhibits a more graceful decay in classification accuracy against the number of training images which is highly desirable in practice where generous training per class is often not available.

Development and evaluation of an automated histology classification system for veterinary pathology.

A 2-stage algorithmic framework was developed to automatically classify digitized photomicrographs of tissues obtained from bovine liver, lung, spleen, and kidney into different histologic categories. The categories included normal tissue, acute necrosis, and inflammation (acute suppurative; chronic). In the current study, a total of 60 images per category (normal; acute necrosis; acute suppurative inflammation) were obtained from liver samples, 60 images per category (normal; acute suppurative inflammation) were obtained from spleen and lung samples, and 60 images per category (normal; chronic inflammation) were obtained from kidney samples. An automated support vector machine (SVM) classifier was trained to assign each test image to a specific category. Using 10 training images/category/organ, 40 test images/category/organ were examined. Employing confusion matrices to represent category-specific classification accuracy, the classifier-attained accuracies were found to be in the 74-90% range. The same set of test images was evaluated using a SVM classifier trained on 20 images/category/organ. The average classification accuracies were noted to be in the 84-95% range. The accuracy in correctly identifying normal tissue and specific tissue lesions was markedly improved by a small increase in the number of training images. The preliminary results from the study indicate the importance and potential use of automated image classification systems in the histologic identification of normal tissues and specific tissue lesions.

Blinded, controlled field trial of two commercially available Mycoplasma bovis bacterin vaccines in veal calves.

Mycoplasma bovis is an etiologic agent of pneumonia, arthritis, and otitis in young calves, such as those found in the special-fed veal industry. We conducted a blinded, controlled trial of two commercially available M. bovis bacterin vaccines for the prevention of respiratory disease in calves associated with M. bovis infection. Calves were randomly assigned to a subcutaneous treatment of vaccine A (n=50), adjuvant A (n=50), vaccine B (n=50), or 0.9% sterile saline solution (n=50) beginning at 27 days of age. Upper-respiratory tract colonization was not impacted by vaccination status. Vaccine A significantly reduced the presence of lung lesions (p=0.0325), however there was no significant reduction of M. bovis in lung lesions. Vaccine B did not significantly reduce total lung lesions or M. bovis-specific lung lesions. The relative risk was determined to be 0.56, 1.0, and 1.36 for vaccine A, adjuvant A, and vaccine B, respectively. There was no association between the total specific antibody isotype (IgM, IgG1, IgG2, IgA) concentrations or M. bovis antibodies and the M. bovis-associated morbidity in the veal calves. Under the field conditions of this study, observed vaccine efficacy for vaccine A and vaccine B was 44% and less than 1%, respectively.

Real-time multiplex polymerase chain reaction assay for rapid detection of Clostridium difficile toxin-encoding strains.

Clostridium difficile is considered an important emerging pathogen capable of causing disease in humans and animal species. In our study, we developed and evaluated a multiplex real-time polymerase chain reaction (PCR) assay for the detection of C. difficile genes encoding toxin A (tcdA), toxin B (tcdB), and binary toxin (cdtA and cdtB). The standardized real-time PCR assay for toxin genes of C. difficile was used to screen for toxigenic C. difficile in fecal samples from 71 preweaned calves, 53 retail ground meat samples, and 27 pasteurized milk samples. All samples were also examined for C. difficile using traditional culture techniques to validate the PCR assay. A total of 24 fecal samples (33.80%) were positive for toxigenic C. difficile using either multiplex real-time PCR or culture. Toxin-encoding C. difficile was detected in 23 enriched fecal samples using the multiplex real-time PCR assay and only 15 samples using culture techniques. C. difficile was not detected in ground meat or pasteurized milk by traditional culture or real-time PCR assay. Eleven fecal samples were positive for all 4 toxin genes, suggesting that preweaned calves may be a likely source for toxigenic C. difficile. On the basis of findings of our study, it can be concluded that multiplex real-time PCR carried out on samples enriched for C. difficile is a reliable, sensitive, and specific diagnostic tool for rapid screening and identification of samples contaminated with C. difficile.

Failure to detect bovine viral diarrhea virus in necropsied farm-raised white-tailed deer (Odocoileus virginianus) in Pennsylvania.

Between January 1 and December 31, 2005 gross and histologic examinations were performed on carcasses of 61 farm-raised white-tailed deer originating from Pennsylvania. Single-tube real-time reverse transcription polymerase chain reaction (real-time RT-PCR) for the detection of bovine viral diarrhea virus type 1 (BVDV-1) and type 2 (BVDV-2) was performed on each animal. Virus isolation was performed on tissue samples from 25 of 61 animals. Immunohistochemical (IHC) staining of ear-notch skin to identify BVDV antigen was performed on each animal. All tissues samples tested negative for both BVDV-1 and BVDV-2 by real-time RT-PCR, virus isolation, and IHC. Gross or histopathologic lesions suggestive of BVDV infection were not detected. Results of this study suggest that BVD is not a common cause of mortality in farm-raised white-tailed deer in Pennsylvania.

Mortality in Pennsylvania captive elk (Cervus elaphus): 1998-2006.

The necropsy records of 65 elk ranging from 1 day to 13(1/2) years of age were reviewed to determine the primary cause of death of each animal. The animal carcasses were submitted from 22 separate Pennsylvania elk farms over an approximately 8-year period. The most common causes of mortality in animals in which the cause of death was determined were gastrointestinal parasitism (21 cases), pneumonia (7 cases), enterocolitis (5 cases), malnutrition (5 cases), and rumenitis/acidosis (5 cases). The cause of death was undetermined in 4 of the 65 elk. Nematode organisms (20 cases) were the primary parasites identified in cases of mortality owing to gastrointestinal parasitism. Arcanobacterium pyogenes (3 cases), Escherichia coli (3 cases), and Streptococcus sp. (2 cases) were the most commonly isolated bacteria from the lung in the cases of pneumonia. Fungal organisms, consistent with Aspergillus sp. were present histologically within the pulmonary lesions in 5 cases. Bacterial agents associated with enterocolitis included Clostridium perfringens (2 cases), E. coli (1 case), Salmonella Newport (1 case) and Salmonella Thompson (1 case). The majority (56.3%) of the death loss in elk of known ages occurred in animals 1 year of age or less, with 71.4% of the gastrointestinal parasitism, 60.0% of the malnutrition, 60.0% of the enterocolitis, and 57.1% of the pneumonia cases occurring in animals within this age range. In conclusion, gastrointestinal parasitism, pneumonia, malnutrition, enterocolitis and rumenitis/acidosis were considered the most common causes of death in captive elk in this study.

A retrospective study of mortality in Pennsylvania captive white-tailed deer (Odocoileus virginianus): 20000--2003.

The postmortem records of 160 white-tailed deer (Odocoileus virginianus) submitted for necropsy examination from 59 separate Pennsylvania captive deer farms over a 3.5-year period were reviewed to determine the primary cause of death of each animal. The most common causes of death were bronchopneumonia (39 cases), enterocolitis (30 cases), malnutrition (13 cases), and trauma (11 cases). Other causes of mortality included severe gastrointestinal parasitism (6 cases), cellulitis with septicemia (5 cases), degenerative myopathy (4 cases), ruminal acidosis (4 cases), and nephritis (4 cases). The cause of death was undetermined in 13 of the 160 animals. Arcanobacterium pyogenes (19 cases), Fusobacterium necrophorum (10 cases), Escherichia coli (7 cases), and Mannheimia haemolytica (4 cases) were the most commonly isolated bacteria from the pneumonic lungs. Bacterial agents associated with enterocolitis included Clostridium perfringens (15 cases), E. coli (12 cases), and Mycobacterium avium subsp. paratuberculosis (2 cases). The majority (52.2%) of the death loss in white-tailed deer of known ages occurred in animals 1 year of age or less, with 46.2% of the bronchopneumonia cases and 50.0% of the enterocolitis cases occurring during this time period. Cases of degenerative myopathy, myocardial degeneration, hepatic necrosis, meningoencephalitis, peritonitis, and urolithiasis considered severe enough to be the primary cause of death appeared early in life, affecting deer 6 months of age or less in all cases. In conclusion, bronchopneumonia, enterocolitis, malnutrition, and trauma were considered the most common causes of death in confined white-tailed deer in this study.

Additional evaluation of undiagnosed bovine abortion cases may reveal fetal neosporosis.

Diagnosis of Neospora caninum associated bovine abortion is often made by histological examination of fetal tissues, although this procedure is not very sensitive. One hundred and forty-four undiagnosed bovine abortion cases were evaluated for N. caninum associated bovine abortion using a revised diagnostic protocol. As a result, 12 (8.3%) of these previously undiagnosed bovine abortion cases were definitively diagnosed as N. caninum positive. The 12 new positive cases included 7 that had exhibited histological lesions, and 5 that had not exhibited histological lesions when examined prior to this study. None of the 12 cases had been immunohistochemically (IHC) stained prior to this study. Also, IHC staining implemented during this research revealed tachyzoites without associated histological lesions in at least 1 tissue from 69 (47.9%) of 144 aborted bovine fetuses, and positive IHC staining of cytoplasmic N. caninum antigen in macrophages in at least 1 tissue from 44 (30.6%) of 144 aborted bovine fetuses. These results demonstrate the necessity for more aggressive evaluation of bovine fetuses for neosporosis.

STK receptor tyrosine kinase regulates susceptibility to infection with Listeria monocytogenes.

We have previously identified the STK receptor tyrosine kinase as a key regulator of macrophage activation and cell-mediated immune responses. Here we demonstrate that, although MSP activation of STK inhibits NO production by macrophages in response to heat-killed Listeria monocytogenes, STK-deficient mice exhibit increased susceptibility to infection with Listeria.

Canine Cutaneous Neosporosis: Clinical Improvement with Clindamycin.

Résumé- Une dermite pyogranulomateuse provoquée par le protozoaire Neospora caninum a été diagnostiquée chez un Golden Retriever de 12 ans. Le nombreux nodules fistulisés étaient localisés au niveau de la tête et du thorax. De nombreux tachyzoites de N. caninum ont été retrouvés dans les biopsies des nodules et le diagnostic a été confirmé pas immunohistologie et examen en microscopie électronique. Le chien avait un titre de sérologie Neospora caninum de 3200 par immunofluorescence indirecte. Après un traitement à base de clindamycine pendant 45 jours, les lésions cutanées ont disparu. Le chien est probablement mort à cause d'un lymphosarcome. Il existait aussi une infection latente àToxoplasma gondii. les Neospora caninum n'ont pas pu être retrouvés par des techniques biologiques ni en culture ou par inoculation de souris à partir de prélèvements nécropsiques. Seuls des tachyzoites dégenérés ont pu être observés histologiquement. Ces observations montrent que la néosporose peut être envisagée dans le diagnostic différentiel des dermites pyogranulomateuses du chien et que la clindamycine est un médicament efficace pour traiter la néosporoe canine. [Dubey, J. P., Metzger, F. L., Hattel, A. L., Lindsay, D. S., Fritz, D. L. Canine cutaneous neosporosis: clinical improvement with clindamycin (Néosporose cutanée canine: amélioration clinique par la clindamycine). Resumen- Se diagnosticó una dermatitis piogranulomatosa causada por el protozoo parásito Neospora caninum en un perro de raza Golden Retriever de 12 años. El animal presentaba varios nódulos en la piel de la cabeza y tórax. Se observaron numerosos taquizoitos de N. caninum en los cortes histológicos de tejido obtenido mediante biopsia de dichos nódulos y el diagnóstico fue confirmado por tinción inmunohistológica y por microscopia electrónica. El perro mostró un titulo de anticuerpos contra N. caninum de 1:3,200 en la prueba de fluorescencia indirecta. Las lesiones cutáneas se resolvieron tras un tratamiento con hidroclorido de clindamicina durante 45 dias. El perro murió posteriormente a causa de un linfoma y presentaba también una infestación latente por Toxoplasma gondii. No se pudo demostrar la presencia de Neospora caninum mediante bioensayos en cultivos celulares ni en ratones inoculados con tejido canino obtenido en la necrospia. Tan solo se pudieron observar taquizoitos degenerados de N. caninum en tejido cutáneo obtenido en la necrospia. Estos hallazgos indican que se debe incluir neosporosis en el diagnóstico diferencial de dermatitis piogranulomatosas en el perro y que la clindamicina puede ser un fármaco eficaz para el tratamiento de la neosporosis cutánea. [Canine cutaneous neosporosis: clinical improvement with clindamycin (Neosporosis cutánea canina: mejora clinica con clindamicina). Abstract- Pyogranulomatous dermatitis caused by the protozoan parasite Neospora caninum was diagnosed in a 12-year-old Golden Retriever dog. Multiple draining nodules were located in the skin of the head and thorax. Numerous tachyzoites of N. caninum were found in histologic sections of the biopsy tissue from the cutaneous nodules and the diagnosis was confirmed by immunohistochemieal staining and by electron microscopic examination. The dog had a 1:3200 serum antibody titer to N. caninum in the indirect fluorescent antibody test. The cutaneous lesions resolved after a 45-day treatment with clindamycin hydrochloride. The dog eventually died because of lymphosarcoma and also had a latent infection with Toxoplasma gondii. Neospora caninum could not be demonstrated by bioassays in cell culture or mice inoculated with canine tissue obtained at necropsy. Only degenerating N. caninum tachyzoites were seen in skin tissue taken at necropsy. These observations indicate that neosporosis should be considered in the differential diagnosis of pyogranulomatous dermatitis in dogs and that clindamycin may be an effective drug for treating cutaneous neosporosis.