PL-TARMI: A deep learning framework for pixel-level traffic crash risk map inference.

A citywide traffic crash risk map is of great significance for preventing future traffic crashes. However, the fine-grained geographic traffic crash risk inference is still a challenging task, mainly due to the complex road network structure, human behavior and high data requirements. In this work, we propose a deep-learning framework PL-TARMI, which leverages easily accessible data to achieve accurate fine-grained traffic crash risk map inference. Specifically, we integrate the satellite image and road network image, combine with other accessible data (e.g., point of interest distribution, human mobility data, traffic data, etc.) as input, and finally obtain the pixel-level traffic crash risk map, which could provide more reasonable traffic crash prevention guidance with a lower cost. Extensive experiments on real-world datasets demonstrate the effectiveness of PL-TARMI.

A preliminary study on classification and therapeutic strategies for spontaneous perirenal hemorrhage.

BACKGROUND: The aim of our study was to report our experience in the classification and therapeutic management strategies for spontaneous perirenal hemorrhage (SPH). METHODS: From September 2005 to April 2015, 20 patients with SPH were newly diagnosed in our hospital. Their clinical features, image findings, identification of underlying causes, and therapeutic management were retrospectively analyzed, and relevant literature was reviewed. In this study, patients were classified according to the degree of severity of the disease or emergency imaging diagnosis of underlying causes. On the basis of the former, patients were classified as critical and noncritical, and on the basis of the latter, patients were classified as renal cell carcinoma (RCC), undefined solid neoplasm, angioleiomyolipoma (AML), and unknown cause. RESULTS: In the acute stage, contrast-enhanced computed tomography (CT) was superior to ultrasonography for both diagnostic accuracy of SPH (p = 0.02) and etiology discovery power (p = 0.004). The results of contrast-enhanced magnetic resonance imaging (MRI) were identical to those of contrast-enhanced CT. We summarized a flowchart in the whole classification and therapeutic strategies of SPH. According to the imaging diagnosis of underlying causes, all the patients with undefined solid neoplasm or RCC underwent emergency operation. Patients with AML or unknown cause underwent selective arterial embolization (SAE) or conservative management according to the critical degree. Acute hemorrhage was controlled in 19 cases, of which 14 were cured by the operation and only one critical patient with severe shock died shortly despite rescue efforts. CONCLUSIONS: Contrast-enhanced CT or MRI is the first choice of imaging examination, which could not only accurately diagnose SPH but also detect the underlying causes. Choice of therapeutic strategies for SPH should vary according to the identification of critical patients and imaging diagnosis of underlying cause.

The feasibility and safety of high-intensity focused ultrasound combined with low-dose external beam radiotherapy as supplemental therapy for advanced prostate cancer following hormonal therapy.

The aim of this study was to investigate the feasibility and safety of high-intensity focused ultrasound (HIFU) combined with (+) low-dose external beam radiotherapy (LRT) as supplemental therapy for advanced prostate cancer (PCa) following hormonal therapy (HT). Our definition of HIFU+LRT refers to treating primary tumour lesions with HIFU in place of reduced field boost irradiation to the prostate, while retaining four-field box irradiation to the pelvis in conventional-dose external beam radiotherapy (CRT). We performed a prospective, controlled and non-randomized study on 120 patients with advanced PCa after HT who received HIFU, CRT, HIFU+LRT and HT alone, respectively. CT/MR imaging showed the primary tumours and pelvic lymph node metastases visibly shrank or even disappeared after HIFU+LRT treatment. There were significant differences among four groups with regard to overall survival (OS) and disease-specific survival (DSS) curves (P = 0.018 and 0.015). Further comparison between each pair of groups suggested that the long-term DSS of the HIFU+LRT group was higher than those of the other three groups, but there was no significant difference between the HIFU+LRT group and the CRT group. Multivariable Cox's proportional hazard model showed that both HIFU+LRT and CRT were independently associated with DSS (P = 0.001 and 0.035) and had protective effects with regard to the risk of death. Compared with CRT, HIFU+LRT significantly decreased incidences of radiation-related late gastrointestinal (GI) and genitourinary (GU) toxicity grade ≥ II. In conclusion, long-term survival of patients with advanced PCa benefited from strengthening local control of primary tumour and regional lymph node metastases after HT. As an alternative to CRT, HIFU+LRT showed good efficacy and better safety.