Vehicle detection stands out as a rapidly developing technology today and is further strengthened by deep learning algorithms. This technology is critical in traffic management, automated driving systems, security, urban planning, environmental impacts, transportation, and emergency response applications. Vehicle detection, which is used in many application areas such as monitoring traffic flow, assessing density, increasing security, and vehicle detection in automatic driving systems, makes an effective contribution to a wide range of areas, from urban planning to security measures. Moreover, the integration of this technology represents an important step for the development of smart cities and sustainable urban life. Deep learning models, especially algorithms such as You Only Look Once version 5 (YOLOv5) and You Only Look Once version 8 (YOLOv8), show effective vehicle detection results with satellite image data. According to the comparisons, the precision and recall values of the YOLOv5 model are 1.63% and 2.49% higher, respectively, than the YOLOv8 model. The reason for this difference is that the YOLOv8 model makes more sensitive vehicle detection than the YOLOv5. In the comparison based on the F1 score, the F1 score of YOLOv5 was measured as 0.958, while the F1 score of YOLOv8 was measured as 0.938. Ignoring sensitivity amounts, the increase in F1 score of YOLOv8 compared to YOLOv5 was found to be 0.06%.

Monitoring marine biodiversity is a challenge in some vulnerable and difficult-to-access habitats, such as underwater caves. Underwater caves are a great focus of biodiversity, concentrating a large number of species in their environment. However, most of the sessile species that live on the rocky walls are very vulnerable, and they are often threatened by different pressures. The use of these spaces as a destination for recreational divers can cause different impacts on the benthic habitat. In this work, we propose a methodology based on video recordings of cave walls and image analysis with deep learning algorithms to estimate the spatial density of structuring species in a study area. We propose a combination of automatic frame overlap detection, estimation of the actual extent of surface cover, and semantic segmentation of the main 10 species of corals and sponges to obtain species density maps. These maps can be the data source for monitoring biodiversity over time. In this paper, we analyzed the performance of three different semantic segmentation algorithms and backbones for this task and found that the Mask R-CNN model with the Xception101 backbone achieves the best accuracy, with an average segmentation accuracy of 82%.

The human brain has been described as a complex system. Its study by means of neurophysiological signals has revealed the presence of linear and nonlinear interactions. In this context, entropy metrics have been used to uncover brain behavior in the presence and absence of neurological disturbances. Entropy mapping is of great interest for the study of progressive neurodegenerative diseases such as Alzheimer’s disease. The aim of this study was to characterize the dynamics of brain oscillations in such disease by means of entropy and amplitude of low frequency oscillations from Bold signals of the default network and the executive control network in Alzheimer’s patients and healthy individuals, using a database extracted from the Open Access Imaging Studies series. The results revealed higher discriminative power of entropy by permutations compared to low-frequency fluctuation amplitude and fractional amplitude of low-frequency fluctuations. Increased entropy by permutations was obtained in regions of the default network and the executive control network in patients. The posterior cingulate cortex and the precuneus showed differential characteristics when assessing entropy by permutations in both groups. There were no findings when correlating metrics with clinical scales. The results demonstrated that entropy by permutations allows characterizing brain function in Alzheimer’s patients, and also reveals information about nonlinear interactions complementary to the characteristics obtained by calculating the amplitude of low frequency oscillations.

The integration of medical images is the process of registering and fusing them to obtain a greater amount of diagnostic information. In this work an analysis is performed for the integration of images obtained through computed axial tomography and magnetic resonance imaging, for which a tool was developed in the Matlab program, where the registration is implemented through equivalent features; in addition, the pairs of images are compared by several fusion rules, with a view to identify the best algorithm in which the resulting fused image contains the most information from the original representations.

In the present research work, we investigated the use of the image intensifier in the extraction of radiopaque foreign bodies in traumatology. First of all, it is necessary to clarify that this method constitutes an essential component of practically generalized use, in which low current level radiation is used, that is, fluoroscopic radiation, so that it can be applied for a considerably longer time than that of the longest radiographic exposure. This tool works with a tube intended for this purpose, which is known as fluoroscopy. The radiations from the tube pass through the patient and reach the serigraph, on which the image intensifier or fluoroscopic screen is mounted. In the latter case, this is where the chain ends, since it is on this screen that the image is formed and where the physician directly observes the region to be studied. It is also necessary to define that a foreign body is any element foreign to the body that enters it, either through the skin or through any natural orifice such as the eyes, nose, throat, preventing its normal functioning. It was possible to obtain as a result that the advantages of fluoroscopic navigation are the reduction of surgical time and the amount of irradiation, which goes from about 140 seconds without navigation to only 8 seconds, which is a substantial difference. Among the conclusions, it was possible to highlight that in the case of a radiopaque object, it is essential to have an image intensifier for localization of the foreign body during surgery; while in the case of a radiolucent foreign body, it is more advisable to locate it through the clinic, since these tend to form granulomas.

The cost of diagnostic errors has been high in the developed world economics according to a number of recent studies and continues to rise. Up till now, a common process of performing image diagnostics for a growing number of conditions has been examination by a single human specialist (i.e., single-channel recognition and classification decision system). Such a system has natural limitations of unmitigated error that can be detected only much later in the treatment cycle, as well as resource intensity and poor ability to scale to the rising demand. At the same time Machine Intelligence (ML, AI) systems, specifically those including deep neural network and large visual domain models have made significant progress in the field of general image recognition, in many instances achieving the level of an average human and in a growing number of cases, a human specialist in the effectiveness of image recognition tasks. The objectives of the AI in Medicine (AIM) program were set to leverage the opportunities and advantages of the rapidly evolving Artificial Intelligence technology to achieve real and measurable gains in public healthcare, in quality, access, public confidence and cost efficiency. The proposal for a collaborative AI-human image diagnostics system falls directly into the scope of this program.