Visual counting, a task that aims to estimate the number of objects from an image/video, is an open-set problem by nature, i.e., the number of population can vary in [0, +∞) in theory. However, collected data and labeled instances are limited in reality, which means that only a small closed set is observed. Existing methods typically model this task in a regression manner, while they are prone to suffer from an unseen scene with counts out of the scope of the closed set. In fact, counting has an interesting and exclusive property-spatially decomposable. A dense region can always be divided until sub-region counts are within the previously observed closed set. We therefore introduce the idea of spatial divide-and-conquer (S-DC) that transforms open-set counting into a closed-set problem. This idea is implemented by a novel Supervised Spatial Divide-and-Conquer Network (SS-DCNet). Thus, SS-DCNet can only learn from a closed set but generalize well to open-set scenarios via S-DC. SS-DCNet is also efficient. To avoid repeatedly computing sub-region convolutional features, S-DC is executed on the feature map instead of on the input image. We provide theoretical analyses as well as a controlled experiment on toy data, demonstrating why closed-set modeling makes sense. Extensive experiments show that SS-DCNet achieves the state-of-the-art performance on three crowd counting datasets (ShanghaiTech, UCF_CC_50 and UCF-QNRF), a vehicle counting dataset (TRANCOS) and a plant counting dataset (MTC), with a 7.7% relative improvement on the UCF-QNRF, 33.1% on the TRANCOS, and 26.4% on the MTC. SS-DCNet also reports the state-of-the-art cross-domain performance on crowd counting datasets. Particularly in the task from UCF-QNRF to ShanghaiTech Part_A, SS-DCNet even beats most existing models trained directly on the target domain. Code and models have been made available at: https://tinyurl.com/SS-DCNet.

Although the roles of Delphian lymph node (DLN) metastasis in papillary thyroid cancer (PTC) have been previously reported, there are still limited data on correlations of clinicopathologic factors with DLN metastasis and unique patterns of cervical node subsite metastasis in PTC patients with DLN metastasis. We retrospectively reviewed medical records of 320 patients with a diagnosis of PTC who underwent primary surgery. Clinicopathologic features and DLN metastasis patterns were analyzed for predicting extensive cervical lymph node metastasis. Both univariate and multivariate Cox regression analyses were used to identify independent factors for cervical lymph node metastasis. DLN metastasis was significantly associated with multifocality, tumor size > 1 cm, extrathyroid extension, BRAFV600E mutation, central neck node metastasis (CNNM), and lateral neck nodes metastases. Patients with DLN metastasis had more lymph node metastases in the central compartment. CNNM number and tumor size > 1 cm were independent risk factors for DLN metastasis. DLN metastasis was highly predictive of lateral lymph node metastasis with moderate sensitivity and high specificity. DLN metastasis is associated with several poor prognostic factors, including extensive cervical lymph node metastasis, and can serve as a predictor of advanced PTC. The presence of DLN metastasis should prompt surgeons to perform an aggressive surgery approach.

No abstract

Abstract. C-X-C motif chemokine receptor (CXCR) 4/CXCL12 is associated with tumor invasion and metastasis in pancreatic cancer. The present study aimed to investigate the possible mechanisms behind this process by studying the association between the expression of CXCR4 and numerous molecular markers. A total of 60 patients with pancreatic cancer who had been treated with radical surgery between

Purpose Achondroplasia is the most common short stature skeletal dysplasia (1:20,000–30,000), but the risk of adverse health outcomes from cardiovascular diseases, pain, poor function, excess weight, and sleep apnea is unclear. A multicenter retrospective natural history study was conducted to understand medical and surgical practices in achondroplasia. Methods Data from patients with achondroplasia evaluated by clinical geneticists at Johns Hopkins University, A.I. duPont Hospital for Children, McGovern Medical School UTHealth, and University of Wisconsin were populated into a REDCap database. All available retrospective medical records of anthropometry (length/height, weight, occipitofrontal circumference), surgery, polysomnography (PSG), and imaging (e.g., X-ray, magnetic resonance imaging) were included. Results Data from 1,374 patients (48.8% female; mean age 15.4 ± 13.9 years) constitute the primary achondroplasia cohort (PAC) with 496 subjects remaining clinically active and eligible for prospective studies. Within the PAC, 76.0% had a de novo FGFR3 pathologic variant and 1,094 (79.6%) had one or more achondroplasia-related surgeries. There are ≥37,000 anthropometry values, 1,631 PSGs and 10,727 imaging studies. Conclusion This is the largest multicenter achondroplasia natural history study, providing a vast array of medical information for use in caring for these patients. This well-phenotyped cohort is a reference population against which future medical and surgical interventions can be compared.

No abstract

No abstract

No abstract