Search Results (5)

There has been controversy over whether to radiologically follow up or use local treatment for asymptomatic small-sized brain metastases from primary lung cancer. For brain tumors without local treatment, we evaluated potential factors related to the brain progression and whether systemic therapy controlled the tumor. We analyzed 96 patients with asymptomatic small-sized metastatic brain tumors from lung cancer. These underwent a radiologic follow-up every 2 or 3 months without local treatment of brain metastases. The pathologies of the tumors were adenocarcinoma (n = 74), squamous cell carcinoma (n = 11), and small cell carcinoma (n = 11). The primary lung cancer was treated with cytotoxic chemotherapy (n = 57) and targeted therapy (n = 39). Patients who received targeted therapy were divided into first generation (n = 23) and second or third generation (n = 16). The progression-free survival (PFS) of brain metastases and the overall survival (OS) of patients were analyzed depending on the age, tumor pathology, number, and location of brain metastases, the extent of other organ metastases, and chemotherapy regimens. The median PFS of brain metastases was 7.4 months (range, 1.1–48.3). Targeted therapy showed statistically significant PFS improvement compared to cytotoxic chemotherapy (p = 0.020). Especially, on univariate and multivariate analyses, the PFS in the second or third generation targeted therapy was more significantly improved compared to cytotoxic chemotherapy (hazard ratio 0.229; 95% confidence interval, 0.082–0.640; p = 0.005). The median OS of patients was 13.7 months (range, 2.0–65.0). Univariate and multivariate analyses revealed that the OS of patients was related to other organ metastases except for the brain (p = 0.010 and 0.020, respectively). Three out of 52 patients with brain recurrence showed leptomeningeal dissemination, while the recurrence patterns of brain metastases were mostly local and/or distant metastases (94.2%). Of the 52 patients who relapsed, 25 patients received local brain treatment. There was brain-related mortality in two patients (2.0%). The intracranial anti-tumor effect was superior to cytotoxic chemotherapy in the treatment of asymptomatic small-sized brain metastases with targeted therapy. Consequently, it becomes possible to determine the optimal timing for local brain treatment while conducting radiological follow-up for these tumors, which do not appear to increase brain-related mortality. Furthermore, this approach has the potential to reduce the number of cases requiring brain local treatment. Full article

Reliable indicators of cancer advancement have actively been sought recently. The detection of colorectal cancer progression markers is essential in improving diagnostic and therapeutic protocols. The aim of the study was to investigate the profile of E-cadherin expression in colorectal cancer tissue depending on the TNM staging and its correlation with several clinical and histopathological features. The study included 55 colorectal cancer patients admitted to the surgical ward for elective surgery. Tissue samples were obtained from resected specimens. Different distributions of E-cadherin expression within tumors were observed; the highest percentage of positive E-cadherin expression was found in the invasive front and in the tumor center. Additionally, the different cellular distribution of E-cadherin expression was noticed; weak membranous E-cadherin expression was the highest in the invasive front and in the budding sites, but a strong membranous pattern was most frequent in the tumor center. Various distributions of E-cadherin expression depending on cancer progression were also found; E-cadherin expression in node-positive patients was lower in the tumor center and in the tumor invasive front, whereas, in patients with distant metastases, the expression of E-Cadherin was lower in the budding sites. In patients with higher TNM stages, E-cadherin expression was lower within the tumor (in the budding sites, tumor center, and invasive front). In tumors with lymphoid follicles, E-cadherin expression was higher in all localizations within the primary tumor. E-cadherin expression in the tumor center was also lower in tumors with some higher tumor budding parameters (areas of poorly differentiated components and poorly differentiated clusters). E-cadherin expression was found to be lower at the tumor center in younger individuals, at the budding sites in men, and at the surrounding lymph nodes in rectal tumors. Low E-cadherin expression appears to be a reliable indicator of higher cancer staging and progression. When assessing the advancement of cancer, apart from the TNM classification, it is beneficial to also consider the expression of E-cadherin. High tumor budding, the poverty of lymphoid follicles, and low E-cadherin expression analyzed simultaneously may contribute to a reliable assessment of colorectal cancer staging. These three histopathological features complement each other, and their investigation, together with conventional tumor staging and grading, may be very helpful in predicting the prognosis of colorectal cancer patients and qualifying them for the best treatment. The role of E-cadherin in the diagnosis and treatment of colorectal cancer, as a part of a personalized medicine strategy, still requires comprehensive, prospective clinical evaluations to precisely target the optimal therapies for the right patients at the right time. Full article

Contextual information plays a pivotal role in the semantic segmentation of remote sensing imagery (RSI) due to the imbalanced distributions and ubiquitous intra-class variants. The emergence of the transformer intrigues the revolution of vision tasks with its impressive scalability in establishing long-range dependencies. However, the local patterns, such as inherent structures and spatial details, are broken with the tokenization of the transformer. Therefore, the ICTNet is devised to confront the deficiencies mentioned above. Principally, ICTNet inherits the encoder–decoder architecture. First of all, Swin Transformer blocks (STBs) and convolution blocks (CBs) are deployed and interlaced, accompanied by encoded feature aggregation modules (EFAs) in the encoder stage. This design allows the network to learn the local patterns and distant dependencies and their interactions simultaneously. Moreover, multiple DUpsamplings (DUPs) followed by decoded feature aggregation modules (DFAs) form the decoder of ICTNet. Specifically, the transformation and upsampling loss are shrunken while recovering features. Together with the devised encoder and decoder, the well-rounded context is captured and contributes to the inference most. Extensive experiments are conducted on the ISPRS Vaihingen, Potsdam and DeepGlobe benchmarks. Quantitative and qualitative evaluations exhibit the competitive performance of ICTNet compared to mainstream and state-of-the-art methods. Additionally, the ablation study of DFA and DUP is implemented to validate the effects. Full article

Village-level agricultural specialization in China is becoming increasingly important for rural development. However, existing knowledge of specialized agricultural villages (SAVs) based on singular assessment criteria and data describing static time points becomes insufficient in addressing multifaceted developmental questions today. We examined the long-term development patterns of SAVs in Anhui, China, with attributes from multiple angles, and explored how local factors affected SAV development across space and time using random forest regression. We found that as time elapsed, economic rationality drove specialized farmers closer to sale dependency and made SAVs more susceptible to market and economic factors, which builds upon previous findings analyzing SAVs at specific time points and consolidates the importance of market factors in the long-term development of SAVs. However, this susceptibility manifests differently in these two geographically contrasting regions north and south of Huai River. The northern SAVs received increased influences from market and economic factors, while the southern SAVs were continuously controlled by market and location factors. The dynamic spatial and temporal patterns of the two regions point to different dependencies, which emphasized local sales in the north and distant sales in the south. We propose that policies and strategies regarding SAV development accommodate these dynamics and address appropriate influencing factors accordingly. Full article

City-wide traffic flow forecasting is a significant function of the Intelligent Transport System (ITS), which plays an important role in city traffic management and public travel safety. However, this remains a very challenging task that is affected by many complex factors, such as road network distribution and external factors (e.g., weather, accidents, and holidays). In this paper, we propose a deep-learning-based multi-branch model called TFFNet (Traffic Flow Forecasting Network) to forecast the short-term traffic status (flow) throughout a city. The model uses spatiotemporal traffic flow matrices and external factors as its input and then infers and outputs the future short-term traffic status (flow) of the whole road network. For modelling the spatial correlations of the traffic flows between current and adjacent road segments, we employ a multi-layer fully convolutional framework to perform cross-correlation calculation and extract the hierarchical spatial dependencies from local to global scales. Also, we extract the temporal closeness and periodicity of traffic flow from historical observations by constructing a high-dimensional tensor comprised of traffic flow matrices from three fragments of the time axis: recent time, near history, and distant history. External factors are also considered and trained with a fully connected neural network and then fused with the output of the main component of TFFNet. The multi-branch model is automatically trained to fit complex patterns hidden in the traffic flow matrices until reaching pre-defined convergent criteria via the back-propagation method. By constructing a rational model input and network architecture, TFFNet can capture spatial and temporal dependencies simultaneously from traffic flow matrices during model training and outperforms other typical traffic flow forecasting methods in the experimental dataset. Full article