Artificial Intelligence (AI) is defined as the ability of machines to perform tasks like humans. Its beginning date back to the 1940s and, thanks to the improvement of computational power and the creation of new tools, it presented great development mainly after the 2000s. One of the sub-divisions of AI is machine learning (ML), which is the ability of a computer to learn from predefined examples, without being explicitly programmed. Its great utility is to process large amounts of information, making decisions and/or predictions. Based on the pre-classification of the database, ML can be supervised, semi-supervised or unsupervised. The algorithms generally perform classification, segmentation and regression tasks. The main types of algorithms used in health care are linear classification, non-linear classification, decision tree and artificial neural networks. Deep Learning, one of the ML fields, and which uses artificial neural networks, enables autonomous machine learning and is useful in healthcare for processing unstructured images. Big Data refers to a large amount of variable and complex information, the processing of which has only been possible with the advances in computer technology in recent years. In Healthcare, AI is already present in research, clinical practice and population health. In various medical specialties AI has already been incorporated into the routine such as pathology, radiology, ophthalmology, dermatology, cardiology etc., and it is believed that in the future all medical specialties will use it. Several existing examples demonstrate that AI algorithms can improve professional learning, assist in decision making, improve flow, assist in health promotion, serve as a tool for inclusion, and optimize resources. In the pandemic of COVID 19, AI has proven useful both in terms of population and in finding treatment for a previously unknown disease. As any new technology, AI can bring many benefits in healthcare, however, it is necessary to discuss ethical and legal aspects and also to ensure wide access, especially in low- and middle-income countries.

Radiomics is designed to extract and organize a large number of features from digital medical images and perform statistical calculations that can correlate the findings with a specific diagnosis, genetic characteristic, staging, response to treatment or outcome. Radiomic processes produce a huge amount of data (features) that can be evaluated individually or together and that require complex technology for its storage, organization and interpretation. Thus, a complex and essential field for the development of radiomics is the ability to deal with big data, which encompasses the data systematic collection and organization, in addition to its correct interpretation. This process begins with the acquisition of images with high quality and reproducibility. These images are standardized, evaluated and segmented. The segmented volume is then subjected to artificial intelligence (AI) models that will extract the features and correlate with the patient's data. Finally, statistical calculations are performed, associating a specific feature with the diagnosis. At the beginning of the radiomic expansion, the features were previously selected and the model should statistically relate them to a certain outcome, a process known as supervised machine learning. Deep learning, a field of machine learning, is based on the model's ability to identify characteristics that may not be perceptible to human evaluation and statistically correlate with a given outcome. Radiogenomics is a field of radiomics that correlates the image data with the genetic alterations of a lesion. In the era of precision medicine, the genetic characterization of lesions has proved to be an efficient tool in therapeutic planning, survival prediction and, in the choice of target therapies. Oncology directly benefits from the development of radiomics and radiogenomics. Almost all oncologic patients undergo imaging scans. Thus these data can be used for early tumor detection, diagnosis and treatment follow-up.

In this chapter, the authors resume the main articles published in the last years about Neuroradiology and Artificial Intelligence (AI), highlighting the main concepts from each one and the knowledge that was built over the years. Neuroradiology, from its beginning, is a field of radiology that has always been developing alongside technological innovation. It has embraced new imaging techniques such as computed tomography and magnetic resonance, which were used to develop AI algorithms since 2010. Nowadays, there are many AI models already available in the market, making the everyday work of neuroradiologists more productive and bringing new information (especially quantitative data) to the reports. With the discussion of these topics in the next pages, the authors hope to give a glimpse of what are the main applications for AI in neuroradiology, the main task that they could aid, what is already available in the market, and which are the main challenges and opportunities in the near future.

INTRODUCTION:Aortic valve bioprostheses ring fracture in valve-in-valve procedures has shown low complication rates and presents as an option in the treatment of patients at high risk for conventional surgery, avoiding high transvalvular gradients, which are associated with increased mortality. Some prostheses available in the market cannot be fractured. In an ex vivo test, the possibility of ring fracture of aortic valve bioprostheses produced in Brazil when submitted to radial force application using a high-pressure non-compliant balloon was evaluated.METHODS:One unit of each aortic valve bioprosthesis model, sizes 19 and 21 mm, produced by Brazilian companies (Braile Biomédica, Cardioprótese, and Labcor), was used. In the experiment, a non-compliant high-pressure balloon (Atlas®-Gold), 1 mm larger than the external diameter of the prosthesis, was positioned inside the valve annulus and inflated gradually aiming to fracture the prosthesis. Fracture pressures and photographic and radiological images of the prostheses before and after test were recorded.RESULTS:All prostheses were fractured. In the models with metal ring, the fracture pressures were between 23 and 25 atm. In the other prostheses, the rupture occurred between 10 and 13 atm. No deformations in the structure were observed, which could potentially damage the aortic root.CONCLUSION:All the Brazilian prostheses evaluated were fractured, although the presence of a metal ring in the prosthesis framework increases the pressure required for fracture. The information obtained helps in the planning of valve-in-valve procedures in patients with aortic valve bioprostheses.

OBJECTIVE:Engage the UNIFESP Cardiovascular Surgery residents in coronary anastomosis, assess their skills and certify results, using the Arroyo Anastomosis Simulator and UNIFESP surgical models.METHODS:First to 6th year residents attended a weekly program of technical training in coronary anastomosis, using 4 simulation models: 1. Arroyo simulator; 2. Dummy with a plastic heart; 3. Dummy with a bovine heart; and 4. Dummy with a beating pig heart. The assessment test was comprised of 10 items, using a scale from 1 to 5 points in each of them, creating a global score of 50 points maximum.RESULTS:The technical performance of the candidate showed improvement in all items, especially manual skill and technical progress, critical sense of the work performed, confidence in the procedure and reduction of the time needed to perform the anastomosis after 12 weeks practice. In response to the multiplicity of factors that currently influence the cardiovascular surgeon training, there have been combined efforts to reform the practices of surgical medical training.CONCLUSION:1 - The four models of simulators offer a considerable contribution to the field of cardiovascular surgery, improving the skill and dexterity of the surgeon in training. 2 - Residents have shown interest in training and cooperate in the development of innovative procedures for surgical medical training in the art.

The sequential implantation of a transcatheter heart valve (THV) within a similar device, also known as valve-in-valve-in-valve, will be an important concept in the future, considering patients with elevated surgical risk and failed THVs within surgical aortic valves (SAV). However, this procedure

Several studies show that portions of intramyocardial coronary arteries are spared of arteriosclerosis, involving morphological, embryological, biochemical and pathophysiological aspects. Endothelial function is significantly affected in the segment of transition, as estimated by the vasoactive response to Ach. These findings suggest that myocardial bridge can provide protection against arteriosclerosis by counteracting the negative effects of endothelial dysfunction. The intramyocardial portion's protection phenomenon deserves further scientific research on all research fronts. Improved morphological, biomechanical and especially physiological and embryological knowledge may be the key to a future window of opportunity for chronic arterial disease therapy and prevention. In addition, this review discusses possible therapeutic approaches for symptomatic coronary ischemia caused by myocardial bridges.

Several studies show that portions of intramyocardial coronary arteries are spared of arteriosclerosis, involving morphological, embryological, biochemical and pathophysiological aspects. Endothelial function is significantly affected in the segment of transition, as estimated by the vasoactive response to Ach. These findings suggest that myocardial bridge can provide protection against arteriosclerosis by counteracting the negative effects of endothelial dysfunction. The intramyocardial portion's protection phenomenon deserves further scientific research on all research fronts. Improved morphological, biomechanical and especially physiological and embryological knowledge may be the key to a future window of opportunity for chronic arterial disease therapy and prevention. In addition, this review discusses possible therapeutic approaches for symptomatic coronary ischemia caused by myocardial bridgesDiversos estudos demonstram que as porções intramiocárdicas das artérias coronárias são poupadas da arteriosclerose, envolvendo aspectos morfológicos, embriológicos, biomecânicos e aspectos fisiopatológicos. A função endotelial é significativamente afetada no segmento de transição, tal como estimado pela resposta vasoativa para acetilcolina (Ach). Esses achados sugerem que ponte miocárdica pode fornecer proteção contra a arteriosclerose, por contrariar os efeitos negativos da disfunção endotelial. O fenômeno dessa proteção da porção intramiocárdica merece maior investigação científica em todas as frentes de pesquisa. Maiores conhecimentos sobre os aspectos morfológicos, biomecânicos e, principalmente, fisiológicos e embriológicos podem ser a chave para uma futura janela de oportunidades de terapia e prevenção da doença arterial crônica. Nessa revisão, discutem-se, também, possíveis abordagens terapêuticas para fenômenos coronarianos isquêmicos causados por pontes miocárdicas

A presente revisão tem por objetivo ressaltar alguns aspectos pouco discutidos da circulação extracorpórea (CEC), levando-se em consideração fisiologia, fisiopatologia e algumas novas tecnologias de perfusão. Assim, alguns aspectos, até certo ponto filosóficos, motivaram a elaboração dessa revisão: a) Preservar e atualizar os conhecimentos do cirurgião sobre a CEC, pelo simples fato de manter a sua liderança pedagógica sobre a sua equipe; b) Questionar se pacientes idosos e diabéticos pelas suas características individuais, assim como adotado para crianças, talvez merecessem protocolos mais apropriados; c) Questionar a reação inflamatória sistêmica causada pela exposição do sangue à superfície não endotelizada do circuito de CEC diante da importância crescente do contato do sangue com a ferida cirúrgica; d) Em relação ao tratamento da síndrome vasoplégica, o azul de metileno continua sendo a melhor opção terapêutica, embora, muitas vezes não seja eficiente pela existência de uma "janela terapêutica" embasada na dinâmica da ação da guanilato ciclase (saturação e síntese "de novo") e; finalmente, e) Razão da escolha do título, ressaltando que, em seus moldes atuais, a CEC seria conseqüência do empirismo, arte, ou da ciência? A mensagem final vem com a convicção de que tanto o empirismo, a arte e a ciência são muito fortes em se tratando da CEC.

The aim of the present review is to highlight some less discussed aspects of the cardiopulmonary bypass (CPB), taking into consideration the physiology, physiopathology, and some new technologies of perfusion. Thus, some points, to a certain extent philosophical, have motivated this revision: a) To preserve and update the surgeon knowledge regarding CPB, even to keep his/her pedagogical leadership on his/her surgical team; b) To question if elderly and diabetic patients, as a result of their individual characteristics deserve more appropriate protocols similar to those adopted for children; c) One third aspect would be the questioning of the systemic inflammatory reaction caused by the blood exposure to CPB non-endothelized circuit surface, in face of the increasing importance of blood contact with the surgical wound; d) In relation to the treatment of the vasoplegic syndrome, methylene blue continues being the best therapeutical option, even so, many times are not efficient on account of a highly probable existence of a "therapeutical window" based on the guanylate cyclase dynamics of action (saturation and synthesis "de novo") and; finally, e) The reason of the title, highlighting that based on its current patterns, would the CPB be an outcome of empiricism, art, or science? The bottom line of this article carries the certainty of that as much as the empiricism, art, and science are highly related to CPB.