Digital pathology platforms with integrated artificial intelligence have the potential to increase the efficiency of the nonclinical pathologist’s workflow through screening and prioritizing slides with lesions and highlighting areas with specific lesions for review. Herein, we describe the comparison of various single- and multi-magnification convolutional neural network (CNN) architectures to accelerate the detection of lesions in tissues. Different models were evaluated for defining performance characteristics and efficiency in accurately identifying lesions in 5 key rat organs (liver, kidney, heart, lung, and brain). Cohorts for liver and kidney were collected from TG-GATEs open-source repository, and heart, lung, and brain from internally selected R&D studies. Annotations were performed, and models were trained on each of the available lesion classes in the available organs. Various class-consolidation approaches were evaluated from generalized lesion detection to individual lesion detections. The relationship between the amount of annotated lesions and the precision/accuracy of model performance is elucidated. The utility of multi-magnification CNN implementations in specific tissue subtypes is also demonstrated. The use of these CNN-based models offers users the ability to apply generalized lesion detection to whole-slide images, with the potential to generate novel quantitative data that would not be possible with conventional image analysis techniques.

Background Peptidylarginine deiminase enzymes (PADs) convert arginine residues to citrulline in a process called citrullination or deimination. Recently, two PADs, PAD2 and PAD4, have been linked to hormone signaling in vitro and the goal of this study was to test for links between PAD2/PAD4 and hormone signaling in vivo. Methods Preliminary analysis of Padi2 and Padi4 single knockout (SKO) mice did not find any overt reproductive defects and we predicted that this was likely due to genetic compensation. To test this hypothesis, we created a Padi2/Padi4 double knockout (DKO) mouse model and tested these mice along with wild-type FVB/NJ (WT) and both strains of SKO mice for a range of reproductive defects. Results Controlled breeding trials found that male DKO mice appeared to take longer to have their first litter than WT controls. This tendency was maintained when these mice were mated to either DKO or WT females. Additionally, unsexed 2-day old DKO pups and male DKO weanlings both weighed significantly less than their WT counterparts, took significantly longer than WT males to reach puberty, and had consistently lower serum testosterone levels. Furthermore, 90-day old adult DKO males had smaller testes than WT males with increased rates of germ cell apoptosis. Conclusions The Padi2/Padi4 DKO mouse model provides a new tool for investigating PAD function and outcomes from our studies provide the first in vivo evidence linking PADs with hormone signaling.

A 12-y-old castrated male domestic longhaired cat had progressive paraparesis and neurolocalization of L4–S3. MRI revealed a circumscribed intradural-extraparenchymal mass from L5 to S1 that was T2 and short tau inversion recovery hyperintense and strongly contrast-enhancing. Cytologic interpretation of a blind fine-needle aspirate obtained through the L5–L6 space was a tumor of probable mesenchymal origin. A pair of suspect neoplastic cells was seen on a cytocentrifuged preparation of the atlanto-occipital CSF sample, despite a normal nucleated cell count (0 × 106/L) and total protein (0.11 g/L) with only 3 RBCs × 106/L. Clinical signs progressed despite increasing doses of prednisolone and cytarabine arabinoside. Repeat MRI on day 162 demonstrated tumor progression from L4 to Cd2 vertebral segments with intraparenchymal extension. Surgical tumor debulking was attempted, but an L4–S1 dorsal laminectomy revealed diffusely abnormal neuroparenchyma. Intraoperative cryosection favored lymphoma, and the cat was euthanized intraoperatively 163 d following presentation. Postmortem examination was performed, and the final diagnosis was a high-grade oligodendroglioma. This case illustrates the cytologic, cryosection, and MRI features of a unique clinical presentation of oligodendroglioma.

Granulosa cell tumors (GCTs) are common ovarian neoplasms in the mare and bitch that can be challenging to diagnose on histopathology. Inhibin has long been the standard immunohistochemical (IHC) marker for GCTs; however, anti-Müllerian hormone (AMH) has not been evaluated widely as an IHC marker in the bitch and mare. We compared the efficacy of AMH and inhibin as IHC markers in canine and equine GCTs. We selected retrospectively 18 equine and 15 canine cases. All equine tumors were dominated by a cystic pattern; canine tumors often had solid patterns. Both inhibin and AMH had similar punctate cytoplasmic patterns of immunolabeling, although labeling intensity was variable; distribution and intensity of labeling were unrelated to the histomorphologic pattern. Labeling for AMH occurred in 12 of 15 canine and 18 of 18 equine cases. Labeling for inhibin occurred in 15 of 15 canine and 18 of 18 equine cases. AMH in equine GCTs often had stronger immunolabeling than inhibin, and granulosa cells were labeled more extensively. Inhibin and AMH performed comparably in bitches, but AMH had more diffuse immunolabeling than inhibin in mares.