Artificial Intelligence (AI) has become an important tool for computer-aided diagnosis of medical imaging. This review aims to provide an overview for clinicians, explaining the relevant aspects of artificial intelligence and machine learning (ML) and presenting up-to-date applications of AI and ML techniques to medical imaging methods such as angiography, magnetic resonance, and echocardiography. For each imaging method, we present the acquisition process, the types of diagnostic test interpretation, and the challenges related to them, as well as how AI/ML techniques, have improved the process of decision making. A summary of selected works applying AI/ML techniques to medical imaging is organized into a table, which highlights the scope of the study, the dataset used, the details of each approach as well as the measured results, including objectives and criteria. The overall benefits of AI in medical imaging are extracted based on the diverse applications and high evaluation scores. In the end, cardiologists should have an advanced understanding of using AI to integrate clinical data and making the final decision in diagnosis.
Background/Objectives: Sepsis and septic shock are critical conditions associated with high mortality rates and substantial impacts on healthcare systems. Accurate and rapid diagnosis is essential for the management of these conditions. The objective of this study is to assess the accuracy of contemporary and traditional methods for diagnosing sepsis and to determine whether improvements have been made concerning the integration of novel diagnostic approaches, to facilitate a prompt diagnosis, taking into account the rapid progression of complications associated with this disease. For this purpose, studies published between 2014 and 2024 were examined to highlight the benefits and limitations of each approach. Methods: A systematic literature review was conducted, including randomized clinical trials, observational studies, and retrospective studies assessing both conventional diagnostic methods (blood cultures and clinical scoring systems) and modern methods (rapid molecular tests, specific biomarkers, and machine learning algorithms). The studies included were selected based on strict design and methodology criteria to ensure a rigorous comparative evaluation of the interventions and technologies used in diagnosing and monitoring patients with sepsis. Results: A total of 23,822 patients were reviewed across the studies included in this systematic analysis. Modern methods, such as continuous monitoring through integrated biosensors and the use of molecular panels for pathogen detection, demonstrated high potential for the early and accurate diagnosis of sepsis. The reviewed studies suggest that these methods can significantly reduce diagnostic time and improve the ability to stratify mortality risk compared to conventional methods. Conclusions: Integrating modern diagnostic technologies, such as rapid pathogen identification tests and specific biomarkers, may complement traditional methods and bring significant benefits in the management of sepsis.
