Search

Abstract Background Systemic anaplastic large cell lymphoma (S-ALCL) is a rare disease with a highly variable prognosis and no standard chemotherapy regimen. Anaplastic lymphoma kinase (ALK) has been reported as an important prognostic factor correlated with S-ALCL in many but not all studies. In our study, we retrospectively analyzed 92 patients with S-ALCL from the Peking University Lymphoma Center for clinical and molecular prognostic factors to make clear the role of ALK and other prognostic factors in Han Chinese S-ALCL. Results The majority of Chinese S-ALCL patients were young male patients (median age 26, male/female ratio 1.7) and the median age was younger than previous reports regardless of ALK expression status. The only statistically significant different clinical characteristic in S-ALCL between ALK positive (ALK+) and ALK negative (ALK-) was age, with a younger median age of 22 for ALK+ compared with 30 for ALK-. However, when pediatric patients (≤18) were excluded, there was no age difference between ALK+ and ALK-. The groups did not differ in the proportion of males, those with clinical stage III/IV (49 vs 51%) or those with extranodal disease (53 vs 59%). Of 73 evaluable patients, the 3-year and 5-year survival rates were 60% and 47%, respectively. Univariate analysis showed that three factors: advanced stage III/IV, lack of expression of ALK, and high Ki-67 expression, were associated with treatment failure in patients with S-ALCL. However, ALK expression correlated with improved survival only in patients younger than 14 years, while not in adult patients. In multivariate analysis, only clinical stage was an independent prognostic factor for survival. Expressions of Wilms tumor 1 (WT1) and B-cell lymphoma 2 protein (BCL-2) correlated with the expression of ALK, but they did not have prognostic significance. High Ki-67 expression was also a poor prognostic factor. Conclusions Our results show that ALK expression alone is not sufficient to determine the outcome of ALCL and other prognostic factors must be considered. Clinical stage is an independent prognostic factor. Ki-67 expression is a promising prognostic factor.

Abstract Regulatory T cells (Tregs) are a specialized subpopulation of CD4+ T cells, which act to suppress the activation of other immune cells. Tregs represent important modulators for the interaction between lymphomas and host microenvironment. Lymphomas are a group of serious and frequently fatal malignant diseases of lymphocytes. Recent studies revealed that some lymphoma T cells might adopt a Treg profile. Assessment of Treg phenotypes and genotypes in patients may offer prediction of outcome in many types of lymphomas including diffuse large B-cell lymphoma, follicular lymphoma, cutaneous T cell lymphoma, and Hodgkin's lymphoma. Based on characterized roles of Tregs in lymphomas, we can categorize the various roles into four groups: (a) suppressor Tregs; (b) malignant Tregs; (c) direct tumor-killing Tregs; and (d) incompetent Tregs. The classification into four groups is significant in predicting prognosis and designing Tregs-based immunotherapies for treating lymphomas. In patients with lymphomas where Tregs serve either as suppressor Tregs or malignant Tregs, anti-tumor cytotoxicity is suppressed thus decreased numbers of Tregs are associated with a good prognosis. In contrast, in patients with lymphomas where Tregs serve as tumor-killing Tregs and incompetent Tregs, anti-tumor cytotoxicity is enhanced or anti-autoimmune Tregs activities are weakened thus increased numbers of Tregs are associated with a good prognosis and reduced numbers of Tregs are associated with a poor prognosis. However, the mechanisms underlying the various roles of Tregs in patients with lymphomas remain unknown. Therefore, further research is needed in this regard as well as the utility of Tregs as prognostic factors and therapy strategies in different lymphomas.

The performances of construction supervisors are essential for the monitoring, control, and coordination of the construction process of a project in order to adhere to a predefined schedule, cost, quality and other factors. However, it is challenging to evaluate their performance due to limitations such as data deficiency, human error, etc. Thus, this paper proposes an approach to data-driven quantitative performance evaluation of construction supervisors by integrating an analytic hierarchy process (AHP) and activity tracking. The proposed approach contains three parts, namely, index extraction, weighting, data-driven index calculation, and then validation by case study. Firstly, performance indexes were developed based on a literature review as well as surveys and function analysis of the information system for construction supervision (CSI system). Then, the weights of and relationships among of the indexes are determined by AHP. After that, with daily workflow and inspection activities tracked in the CSI system, a method and a software module for automatic calculation of indexes were developed. Lastly, the proposed approach was validated by a real-world case. The result showed that the proposed approach can quantify the performance of a construction supervisor systematically and automatically, which shed lights on how to evaluate the performance of a worker based on the tracking of daily activities. The data-driven process enhanced our strong interpretation of member actions and evaluation indexes, and can boost the performance of every member in an organization.

Accurate fire load (combustible objects) information is crucial for safety design and resilience assessment of buildings. Traditional fire load acquisition methods, such as fire load survey, which are time-consuming, tedious, and error-prone, failed to adapt to dynamic changed indoor scenes. As a starting point of automatic fire load estimation, fast recognition and detection of indoor fire load are important. Thus, this research proposes a computer vision-based method to automatically detect indoor fire loads using deep learning-based instance segmentation. First, indoor elements are classified into different categories according to their material composition. Next, an image dataset of indoor scenes with instance annotations is developed. Finally, a deep learning model, based on Mask R-CNN, is developed and trained using transfer learning to detect fire loads in images. Experimental results show that our model achieves promising accuracy, as measured by an average precision (AP) of 40.5% and AP50 of 59.2%, for instance segmentation on the dataset. A comparison with manual detection demonstrates the method’s high efficiency as it can detect fire load 1200 times faster than humans. This research contributes to the body of knowledge 1) a novel method of high accuracy and efficiency for automated fire load recognition in indoor environments based on instance segmentation; 2) training techniques for a deep learning model in a relatively small dataset of indoor images which includes complex scenes and a variety of instances; and 3) an image dataset with annotations of indoor fire loads. Although instance segmentation has been applied for several years, this is a pioneering research on using it for automated indoor fire load recognition, which paves the foundation for automatic fire load estimation and resilience assessment for the built environment.