A dense residual U-net for multiple sclerosis lesions segmentation from multi-sequence 3D MR images.

Multiple Sclerosis (MS) is an autoimmune disease that causes brain and spinal cord lesions, which magnetic resonance imaging (MRI) can detect and characterize. Recently, deep learning methods have achieved remarkable results in the automated segmentation of MS lesions from MRI data. Hence, this study proposes a novel dense residual U-Net model that combines attention gate (AG), efficient channel attention (ECA), and Atrous Spatial Pyramid Pooling (ASPP) to enhance the performance of the automatic MS lesion segmentation using 3D MRI sequences. First, convolution layers in each block of the U-Net architecture are replaced by residual blocks and connected densely. Then, AGs are exploited to capture salient features passed through the skip connections. The ECA module is appended at the end of each residual block and each downsampling block of U-Net. Later, the bottleneck of U-Net is replaced with the ASSP module to extract multi-scale contextual information. Furthermore, 3D MR images of Fluid Attenuated Inversion Recovery (FLAIR), T1-weighted (T1-w), and T2-weighted (T2-w) are exploited jointly to perform better MS lesion segmentation. The proposed model is validated on the publicly available ISBI2015 and MSSEG2016 challenge datasets. This model produced an ISBI score of 92.75, a mean Dice score of 66.88%, a mean positive predictive value (PPV) of 86.50%, and a mean lesion-wise true positive rate (LTPR) of 60.64% on the ISBI2015 testing set. Also, it achieved a mean Dice score of 67.27%, a mean PPV of 65.19%, and a mean sensitivity of 74.40% on the MSSEG2016 testing set. The results show that the proposed model performs better than the results of some experts and some of the other state-of-the-art methods realized related to this particular subject. Specifically, the best Dice score and the best LTPR are obtained on the ISBI2015 testing set by using the proposed model to segment MS lesions.

A Residual-Inception U-Net (RIU-Net) Approach and Comparisons with U-Shaped CNN and Transformer Models for Building Segmentation from High-Resolution Satellite Images.

Building segmentation is crucial for applications extending from map production to urban planning. Nowadays, it is still a challenge due to CNNs' inability to model global context and Transformers' high memory need. In this study, 10 CNN and Transformer models were generated, and comparisons were realized. Alongside our proposed Residual-Inception U-Net (RIU-Net), U-Net, Residual U-Net, and Attention Residual U-Net, four CNN architectures (Inception, Inception-ResNet, Xception, and MobileNet) were implemented as encoders to U-Net-based models. Lastly, two Transformer-based approaches (Trans U-Net and Swin U-Net) were also used. Massachusetts Buildings Dataset and Inria Aerial Image Labeling Dataset were used for training and evaluation. On Inria dataset, RIU-Net achieved the highest IoU score, F1 score, and test accuracy, with 0.6736, 0.7868, and 92.23%, respectively. On Massachusetts Small dataset, Attention Residual U-Net achieved the highest IoU and F1 scores, with 0.6218 and 0.7606, and Trans U-Net reached the highest test accuracy, with 94.26%. On Massachusetts Large dataset, Residual U-Net accomplished the highest IoU and F1 scores, with 0.6165 and 0.7565, and Attention Residual U-Net attained the highest test accuracy, with 93.81%. The results showed that RIU-Net was significantly successful on Inria dataset. On Massachusetts datasets, Residual U-Net, Attention Residual U-Net, and Trans U-Net provided successful results.

New MS lesion segmentation with deep residual attention gate U-Net utilizing 2D slices of 3D MR images.

Multiple sclerosis (MS) is an autoimmune disease that causes lesions in the central nervous system of humans due to demyelinating axons. Magnetic resonance imaging (MRI) is widely used for monitoring and measuring MS lesions. Automated methods for MS lesion segmentation have usually been performed on individual MRI scans. Recently, tracking lesion activity for quantifying and monitoring MS disease progression, especially detecting new lesions, has become an important biomarker. In this study, a unique pipeline with a deep neural network that combines U-Net, attention gate, and residual learning is proposed to perform better new MS lesion segmentation using baseline and follow-up 3D FLAIR MR images. The proposed network has a similar architecture to U-Net and is formed from residual units which facilitate the training of deep networks. Networks with fewer parameters are designed with better performance through the skip connections of U-Net and residual units, which facilitate information propagation without degradation. Attention gates also learn to focus on salient features of the target structures of various sizes and shapes. The MSSEG-2 dataset was used for training and testing the proposed pipeline, and the results were compared with those of other proposed pipelines of the challenge and experts who participated in the same challenge. According to the results over the testing set, the lesion-wise F1 and dice scores were obtained as a mean of 48 and 44.30%. For the no-lesion cases, the number of tested and volume of tested lesions were obtained as a mean of 0.148 and 1.488, respectively. The proposed pipeline outperformed 22 proposed pipelines and ranked 8th in the challenge.

GIS based evaluation of social determinants of children's health in Turkey: Case study of Istanbul.

Basic elements considered as social determinants of the health varies in political, socio-economic, structural and intermediary contexts. While socio-economic and political contexts are directly related with the social, economic, public and health policies in country scale. The structural context additionally includes socio-economic dimensions such as income, education, occupation, social class, gender and race/ethnicity. In addition to these basic determinants, the public health, and especially the children health is also affected by the intermediary determinants, which are material circumstances including physical conditions of the working, housing and neighborhood environments and consumption potential (i.e. healthy foods, proper clothing etc.). Existing experiences provided that, the children who grow up on low socioeconomic conditions or on inappropriate environmental conditions including the residential structures tend to become more often ill than the children living in better environmental and socio-economic conditions. This situation reveals the importance of the city planning in terms of providing better conditions for children's health. This study aims to evaluate the social determinants of children's health by the use of Geographic Information System (GIS) technology. For this purpose, a variety of social determinants in terms of political (quality and quantity of health services), structural (education and social class) and intermediary (physical environment, housing, and neighborhood) contexts were examined in Bakirköy and Esenler districts, which are located European side of Istanbul. For this purpose, 2017 dated official dataset including census information and the statistics on the quality and the quantity of the education and health services in two districts were used for examining the political and structural determinants. The spatial characteristics of the physical environment and housing conditions in the study area were constructed from cadastral maps and development plans by use of GIS tools. As a last step, children's health data that consists of pediatric patient visits and diagnosis reports from 12 hospitals in Bakirköy and Esenler districts were also examined for understanding the potential relationships between the social determinants and existing health conditions. Results of this research revealed that the Bakirköy district has better conditions in terms of all health determinants when compared with Esenler district. Therefore, the health status of children living in Bakirköy is expected to be better than those living in the Esenler, which coincides with the evaluation of official children health data.

GIS based spatial pattern analysis: Children with Hepatitis A in Turkey.

This study aimed to provide an insight into the geographic distribution of Hepatitis A incidence considering their temporal distribution, spatial patterns, hot spots and clusters identification in three different age-group (0-4, 5-9 and 10-14) in Turkey. Province based tabular data, including monthly numbers of Hepatitis A cases in children, and the populations from 2001 to 2011 were used as the basic input of the study. Time series maps were created using Geographic Information Systems (GIS) to introduce the temporal changes in the morbidity rates of Hepatitis A. The spatial variation of Hepatitis A was measured using Moran's I at the global level and the local indicators of spatial associations (LISAs) Moran's I and Getis-Ord Gi *(d) in order to identify influential locations through clusters and hot spots detection of Hepatitis A cases. The morbidity rates in children under the age of 5 were found significantly lower than the other age-groups, whereas the age-group 5-9 revealed the highest morbidity rates in the study area. The morbidity of Hepatitis A was detected very high for the years 2001, and 2005-2007. The identification of the highly vulnerable provinces was conducted using local Moran's I and local Getis-Ord Gi *(d). The majority of clusters and hot spots were detected to be agglomerated in the Eastern Mediterranean and South-Eastern Anatolian Regions and Ceyhan, Asi and Southeast part of Firat-Dicle river basins in Turkey.

Comparison of low cost 3D structured light scanners for face modeling.

This study aims to compare three different structured light scanner systems to generate accurate 3D human face models. Among these systems, the most dense and expensive one was denoted as the reference and the other two that were low cost and low resolution were compared according to the reference system. One female face and one male face were scanned with three light scanner systems. Point-cloud filtering, mesh generation, and hole-filling steps were carried out using a trial version of commercial software; moreover, the data evaluation process was realized using CloudCompare open-source software. Various filtering and mesh smoothing levels were applied on reference data to compare with other low-cost systems. Thus, the optimum reduction level of reference data was evaluated to continue further processes. The outcome of the presented study shows that low-cost structured light scanners have a great potential for 3D object modeling, including the human face. A considerable cheap structured light system has been used due to its capacity to obtain spatial and morphological information in the case study of 3D human face modeling. This study also discusses the benefits and accuracy of low-cost structured light systems.

Monitoring the drastic growth of ship breaking yards in Sitakunda: a threat to the coastal environment of Bangladesh.

The vast coastal and marine resources that occur along the southern edge of Bangladesh make it one of the most productive areas of the world. However, due to growing anthropogenic impacts, this area is under considerable environmental pressure from both physical and chemical stress factors. Ship breaking, or the dismantling and demolition of out-of-service ocean-going vessels, has become increasingly common in many coastal areas. To investigate the extent of ship breaking activities in Bangladesh along the Sitakunda coast, various spatial and non-spatial data were obtained, including remote sensing imagery, statistical records and published reports. Impacts to coastal and marine life were documented. Available data show that ship breaking activities cause significant physical disturbance and release toxic materials into the environment, resulting in adverse effects to numerous marine taxonomic groups such as fish, mammals, birds, reptiles, plants, phytoplankton, zooplankton and benthic invertebrates. Landsat imagery illustrates that the negatively impacted coastal area has grown 308.7 % from 367 ha in 1989 to 1,133 ha in 2010. Physicochemical and biological properties of coastal soil and water indicate substantially elevated pollution that poses a risk of local, regional and even global contamination through sea water and atmospheric transport. While damage to the coastal environment of Bangladesh is a recognized hazard that must be addressed, the economic benefits of ship breaking through job creation and fulfilling the domestic demand for recycled steel must be considered. Rather than an outright ban on beach breaking of ships, the enterprise must be recognized as a true and influential industry that should be held responsible for developing an economically viable and environmentally proactive growth strategy. Evolution of the industry toward a sustainable system can be aided through reasonable and enforceable legislative and judicial action that takes a balanced approach, but does not diminish the value of coastal conservation.

Water quality assessment and meta model development in Melen watershed - Turkey.

Istanbul, being one of the highly populated metropolitan areas of the world, has been facing water scarcity since the past decade. Water transfer from Melen Watershed was considered as the most feasible option to supply water to Istanbul due to its high water potential and relatively less degraded water quality. This study consists of two parts. In the first part, water quality data covering 26 parameters from 5 monitoring stations were analyzed and assessed due to the requirements of the "Quality Required of Surface Water Intended for the Abstraction of Drinking Water" regulation. In the second part, a one-dimensional stream water quality model with simple water quality kinetics was developed. It formed a basic design for more advanced water quality models for the watershed. The reason for assessing the water quality data and developing a model was to provide information for decision making on preliminary actions to prevent any further deterioration of existing water quality. According to the water quality assessment at the water abstraction point, Melen River has relatively poor water quality with regard to NH(4)(+), BOD(5), faecal streptococcus, manganese and phenol parameters, and is unsuitable for drinking water abstraction in terms of COD, PO(4)(3-), total coliform, total suspended solids, mercury and total chromium parameters. The results derived from the model were found to be consistent with the water quality assessment. It also showed that relatively high inorganic nitrogen and phosphorus concentrations along the streams are related to diffuse nutrient loads that should be managed together with municipal and industrial wastewaters.

Implementation of soil survey and analyses for promoting watershed modelling applications.

Soil is an important component of a watershed. Understanding soils and their interactions with the other components are, thus, considered to be critical and essential for conservation of resources and management of the watershed. Development of soil sampling and analysis programs are crucial for these purposes. Site-specific soil data are needed to identify current soil characteristics, as well as to validate datasets gathered for watershed-scale modelling of non-point sources (NPS) of pollutants arising from various land-use activities, hydrodynamics and water quality. The Koycegiz Lake-Dalyan Lagoon watershed, located in the southwest of Turkey along the Mediterranean Sea Coast, was selected as the study area for watershed modelling purposes. Development of soil sampling plans, their practical optimization, soil analyses and interpretation are presented in this article. The soil analyses conducted include physical, chemical and specific soil characteristics. Within the framework of this study, soil fertility parameters are presented and evaluated. Such an approach used is recommended for especially developing countries where up-to-date data sets are not fully available and/or centrally publicized.

Use of remotely sensed data for analysis of land-use change in a highly urbanized district of mega city, Istanbul.

The study forms an example on monitoring and understanding urban dynamics by using remotely sensed data. The selected region is a rapidly urbanizing district of the mega city Istanbul, Gaziosmanpasa, whose population has almost doubled between years 1990 and 2000. The significance of this district besides its urban sprawl is that 61% of its land lies within the boundaries of an important drinking water reservoir watershed of the mega city, the Alibeykoy Reservoir. The land-use/cover changes that has occurred in the years of 1987 and 2001 are analyzed by utilizing a variety of data sources including satellite images (Landsat TM image of September 1987 and Landsat ETM+ image of May 2001), aerial photographs, orthophoto maps, standard 1:25000 scale topographic maps, and various thematic maps together with ground survey. Land-use changes are analyzed on the basis of protection zones of the reservoir watershed and the conversion of bare land and forests to settlements are clearly observed despite the national regulation on watershed protection. The decline of forests within the protection zones was from 69% to 63.6% whereas the increase in settlements was from 0.8% to 3.9%. The associated impact of establishing new residential sites with insufficient infrastructure is then linked with the water quality of the reservoir that has already reached to Class III characteristics regarding the recently revised national legislation stating that any class exceeding Class II cannot be used as a drinking water supply that in turn, had consequences on regulating the water services such as upgrading the existing water treatment plant. The paper aims to help the managers, decision-makers and urban planners by informing them of the past and current land-use/cover changes, to influence the cessation of illegal urbanization through suitable decision-making and environmental policy that adhere to sustainable resource use.