Clustering Aided Weakly Supervised Training to Detect Anomalous Events in Surveillance Videos.

Formulating learning systems for the detection of real-world anomalous events using only video-level labels is a challenging task mainly due to the presence of noisy labels as well as the rare occurrence of anomalous events in the training data. We propose a weakly supervised anomaly detection system that has multiple contributions including a random batch selection mechanism to reduce interbatch correlation and a normalcy suppression block (NSB) which learns to minimize anomaly scores over normal regions of a video by utilizing the overall information available in a training batch. In addition, a clustering loss block (CLB) is proposed to mitigate the label noise and to improve the representation learning for the anomalous and normal regions. This block encourages the backbone network to produce two distinct feature clusters representing normal and anomalous events. An extensive analysis of the proposed approach is provided using three popular anomaly detection datasets including UCF-Crime, ShanghaiTech, and UCSD Ped2. The experiments demonstrate the superior anomaly detection capability of our approach.

Stabilizing Adversarially Learned One-Class Novelty Detection Using Pseudo Anomalies.

Recently, anomaly scores have been formulated using reconstruction loss of the adversarially learned generators and/or classification loss of discriminators. Unavailability of anomaly examples in the training data makes optimization of such networks challenging. Attributed to the adversarial training, performance of such models fluctuates drastically with each training step, making it difficult to halt the training at an optimal point. In the current study, we propose a robust anomaly detection framework that overcomes such instability by transforming the fundamental role of the discriminator from identifying real vs. fake data to distinguishing good vs. bad quality reconstructions. For this purpose, we propose a method that utilizes the current state as well as an old state of the same generator to create good and bad quality reconstruction examples. The discriminator is trained on these examples to detect the subtle distortions that are often present in the reconstructions of anomalous data. In addition, we propose an efficient generic criterion to stop the training of our model, ensuring elevated performance. Extensive experiments performed on six datasets across multiple domains including image and video based anomaly detection, medical diagnosis, and network security, have demonstrated excellent performance of our approach.

IgG4-Related Systemic Disease Can Be Easily Mistaken as a Uroepithelial Tumor.

Immunoglobulin G4-related disease (IgG4-RD) is a newly recognized systemic syndrome characterized by elevated serum IgG4 concentrations and tumefaction or tissue infiltration by IgG4-positive plasma cells. We experienced a case of IgG4-RD involving multiple organs in a 64-year-old female who was referred for a suspected uroepithelial tumor. A mass biopsy confirmed dense lymphoplasmacytic infiltration with an increased number of IgG4-positive plasma cells. We discuss this case and review the literature to bring IgG4-RD to the attention to clinicians because it responds dramatically well to steroid therapy and should be kept in mind as a differential diagnosis to avoid unnecessary surgery.

Sequential combination of intravenous recombinant tissue plasminogen activator and intra-arterial urokinase in acute ischemic stroke.

BACKGROUND AND PURPOSE: Combined intravenous (IV) and intra-arterial (IA) thrombolytic therapy may be faster and easier to initiate than monotherapy, and its recanalization rate may be better as well. The sequential combination of recombinant tissue plasminogen activator (rTPA) and urokinase (UK) has synergistic and complementary effects on clot lysis. We prospectively evaluated the effectiveness and safety of sequential combination of IV rTPA and IA UK in acute ischemic stroke. METHODS: IV rTPA was administered to patients with acute stroke within 3 hours of onset. Those whose condition had not improved at the end of rTPA infusion were further treated with selective IA UK. We evaluated baseline and 30-day National Institutes of Health Stroke Scale (NIHSS) scores and 90-day modified Rankin Scale scores. RESULTS: Thirty patients were initially treated with IV rTPA; 24 were further treated with IA UK. Four patients who had rapid reocclusion following initial successful IA therapy received IV abciximab. Fourteen of 24 patients who underwent angiography had an effective perfusion state of Thrombolysis in Myocardial Infarction grade 3 flow. Median baseline and 30-day NIHSS scores were 18 and 2, respectively. Eighteen patients improved to a modified Rankin scale score of 0 or 1 after 90 days. Symptomatic hemorrhage developed in two patients. CONCLUSION: The strategy of using conventional-dose IV rTPA and the sequential combination of IA UK in patients without an early clinical response to IV treatment was safe and feasible. This strategy achieved high complete arterial recanalization rates and good functional outcomes.