The power of Artificial Intelligence (AI) combined with the surgeons’ expertise leads to breakthroughs in surgical care, bringing new hope to patients. Utilizing deep learning-based computer vision techniques in surgical procedures will enhance the healthcare industry. Laparoscopic surgery holds excellent potential for computer vision due to the abundance of real-time laparoscopic recordings captured by digital cameras containing significant unexplored information. Furthermore, with computing power resources becoming increasingly accessible and Machine Learning methods expanding across various industries, the potential for AI in healthcare is vast. There are several objectives of AI’s contribution to laparoscopic surgery; one is an image guidance system to identify anatomical structures in real-time. However, few studies are concerned with intraoperative anatomy recognition in laparoscopic surgery. This study provides a comprehensive review of the current state-of-the-art semantic segmentation techniques, which can guide surgeons during laparoscopic procedures by identifying specific anatomical structures for dissection or avoiding hazardous areas. This review aims to enhance research in AI for surgery to guide innovations towards more successful experiments that can be applied in real-world clinical settings. This AI contribution could revolutionize the field of laparoscopic surgery and improve patient outcomes.

This research underscores the importance of enhancing the early detection of diabetic retinopathy and glaucoma, two prominent culprits behind vision loss. Typically, retinal diseases lurk without symptoms until they inflict severe vision impairment, underscoring the critical need for early identification. The research is centered on the potential of leveraging fundus images, which offer invaluable insights by analyzing various attributes of retinal blood vessels, such as their length, width, tortuosity, and branching patterns. The conventional practice of manually segmenting retinal vessels by medical professionals is both intricate and time-consuming, demanding specialized expertise. This approach, reliant on pathologists, grapples with limitations related to scalability and accessibility. To surmount these challenges, the research introduces an automated solution employing computer vision. It conducts an evaluation of diverse retinal vessel segmentation and classification methods, including machine learning, filtering-based, and model-based techniques. Robust performance assessments, involving metrics like the true positive rate, true negative rate, and accuracy, facilitate a comprehensive comparison of these methodologies. The ultimate goal of this research is to create more efficient and accessible diagnostic tools, consequently enhancing the early detection of eye diseases through automated retinal vessel segmentation and classification. This endeavor combines the capabilities of computer vision and deep learning to pioneer new benchmarks in the realm of biomedical imaging, thereby addressing the pressing issues surrounding eye disease diagnosis.

To save patients’ lives, it is important to go for an early diagnosis of intracranial hemorrhage (ICH). For diagnosing ICH, the widely used method is non-contrast computed tomography (NCCT). It has fast acquisition and availability in medical emergency facilities. To predict hematoma progression and mortality, it is important to estimate the volume of intracranial hemorrhage. Radiologists can manually delineate the ICH region to estimate the hematoma volume. This process takes time and undergoes inter-rater variability. In this research paper, we develop and discuss a fine segmentation model and a coarse model for intracranial hemorrhage segmentations. Basically, two different models are discussed for intracranial hemorrhage segmentation. We trained a 2DDensNet in the first model for coarse segmentation and cascaded the coarse segmentation mask output in the fine segmentation model along with input training samples. A nnUNet model is trained in the second fine stage and will use the segmentation labels of the coarse model with true labels for intracranial hemorrhage segmentation. An optimal performance for intracranial hemorrhage segmentation solution is obtained.

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%.

In view of the fact that the convolution neural network segmentation method lacks to capture the global dependency of infected areas in COVID-19 images, which is not conducive to the complete segmentation of scattered lesion areas, this paper proposes a COVID-19 lesion segmentation method UniUNet based on UniFormer with its strong ability to capture global dependency. Firstly, a U-shaped encoder-decoder structure based on UniFormer is designed, which can enhance the cooperation ability of local and global relations. Secondly, Swin spatial pyramid pooling module is introduced to compensate the influence of spatial resolution reduction in the encoder process and generate multi-scale representation. Multi-scale attention gate is introduced at the skip connection to suppress redundant features and enhance important features. Experiment results show that, compared with the other four methods, the proposed model achieves better results in Dice, loU and Recall on COVID-19-CT-Seg and CC-CCIII dataset, and achieves a more complete segmentation of the lesion area.

The present work shows an application of the Chan-Vese algorithm for the semi-automatic segmentation of anatomical structures of interest (lungs and lung tumor) in 4DCT images of the thorax, as well as their three-dimensional reconstruction. The segmentation and reconstruction were performed on 10 CT images, which make up an inspiration-expiration cycle. The maximum displacement was calculated for the case of the lung tumor using the reconstructions of the onset of inspiration, the onset of expiration, and the voxel information. The proposed method achieves appropriate segmentation of the studied structures regardless of their size and shape. The three-dimensional reconstruction allows us to visualize the dynamics of the structures of interest throughout the respiratory cycle. In the future, it is expected to have more evidence of the good performance of the proposed method and to have the feedback of the clinical expert, since the knowledge of the characteristics of anatomical structures, such as their dimension and spatial position, helps in the planning of Radiotherapy (RT) treatments, optimizing the radiation dose to cancer cells and minimizing it in healthy organs. Therefore, the information found in this work may be of interest for the planning of RT treatments.