Cardiovascular diseases (CVDs) are the leading cause of death globally. Electrocardiograms (ECGs) are crucial diagnostic tools; however, their traditional interpretations exhibit limited sensitivity and reproducibility. This systematic review discusses the recent advances in artificial intelligence (AI), including deep learning and machine learning, applied to ECG analysis for CVD detection. It examines over 100 studies from 2019 to 2025, classifying AI applications by disease type (heart failure, myocardial infarction, and atrial fibrillation), model architecture (convolutional neural networks, long short-term memory, and hybrid models), and methodological innovation (signal denoising, synthetic data generation, and explainable AI). Comparative tables and conceptual figures highlight performance metrics, dataset characteristics, and implementation challenges. Our findings indicated that AI models outperform traditional methods, especially in terms of detecting subclinical conditions and enabling real-time monitoring via wearable technologies. Nonetheless, issues such as demographic bias, lack of dataset diversity, and regulatory hurdles persist. The review concludes by offering actionable recommendations to enhance clinical translation, equity, and transparency in AI-ECG applications. These insights aim to guide interdisciplinary efforts toward the safe and effective adoption of AI in cardiovascular diagnostics.

The randomization-based feedforward neural network has raised great interest in the scientific community due to its simplicity, training speed, and accuracy comparable to traditional learning algorithms. The basic algorithm consists of randomly determining the weights and biases of the hidden layer and analytically calculating the weights of the output layer by solving a linear overdetermined system using the Moore–Penrose generalized inverse. When processing large volumes of data, randomization-based feedforward neural network models consume large amounts of memory and drastically increase training time. To efficiently solve the above problems, parallel and distributed models have recently been proposed. Previous reviews of randomization-based feedforward neural network models have mainly focused on categorizing and describing the evolution of the algorithms presented in the literature. The main contribution of this paper is to approach the topic from the perspective of the handling of large volumes of data. In this sense, we present a current and extensive review of the parallel and distributed models of randomized feedforward neural networks, focusing on extreme learning machine. In particular, we review the mathematical foundations (Moore–Penrose generalized inverse and solution of linear systems using parallel and distributed methods) and hardware and software technologies considered in current implementations.

In this research a three step method for right heart segmentation based on a gradient-based difference operator and machine learning techniques is reported. The proposed method is applied to human heart multi-slice computerized tomography (MSCT) volumes. The first step is the preprocessing, where a gradient-based difference operator is applied to exploit the functional relationship between the original input image and its edge enhanced version. In the second step, the least squares support vector machines (LSSVM) are used with a double purpose. First, an appropriate volume-of-interest is automatically established in order to isolate the structure to segment. Second, another LSSVM is trained for locating the voxels required for initializing the seed based clustering procedure. In the third step (segmentation step), the preprocessed volumes are subsequently processed with an unsupervised clustering technique based on simple linkage region growing. Dice score is used as a metric function to compare the segmentations obtained using the proposed method with respect to ground truth volumes traced by a cardiologist. The right atrium, pulmonary valve, right ventricle and venae cavae are segmented from 80 cardiac MSCT volumes. Reported metrics confirm that this method is a promising technique for right heart segmentation.

In computed tomography imaging, brain hematoma (BH) segmentation is a very challenging process due to a high variability of BH morphology, low contrast and noisy images. Because of this, BH segmentation is an open problem. In order to approach this problem, we propose an automatic technique, named nonlinear technique (NLT), based on a thresholding method, noise suppression filters, intelligent operators, a clustering strategy and a binary morphological operator. NLT performance is assessed by Jaccard's similarity index (JSI) used to compare automatic and manual BH segmentations. This assessment allows developing a tuning process for establishing the optimal parameters of each of the algorithms which constitute the proposed technique. The results indicate a good correlation, based on JSI, between the manual segmentations and the automatic ones. Finally, the BH volume is generated considering the automatic segmentation. This volume indicates whether or not the patient must undergo a surgical intervention for BH treatment.

The aim of this research is proposing a 3-D similarity enhancement technique useful for improving the segmentation of cardiac structures in Multi-Slice Computerized Tomography (MSCT) volumes. The similarity enhancement is obtained by subtracting the intensity of the current voxel and the gray levels of their adjacent voxels in two volumes resulting after preprocessing. Such volumes are: a. - a volume obtained after applying a Gaussian distribution and a morphological top-hat filter to the input and b. - a smoothed volume generated by processing the input with an average filter. Then, the similarity volume is used as input to a region growing algorithm. This algorithm is applied to extract the shape of cardiac structures, such as left and right ventricles, in MSCT volumes. Qualitative and quantitative results show the good performance of the proposed approach for discrimination of cardiac cavities.

Leishmaniasis is a complex group of diseases caused by obligate unicellular and intracellular eukaryotic protozoa of the leishmania genus. Leishmania species generate diverse syndromes ranging from skin ulcers of spontaneous resolution to fatal visceral disease. These syndromes belong to three categories: visceral leishmaniasis, cutaneous leishmaniasis and mucosal leishmaniasis. The visceral leishmaniasis is based on the reticuloendothelial system producing hepatomegaly, splenomegaly and lymphadenopathy. In the present article, a semiautomatic segmentation strategy is proposed to obtain the segmentations of the evolutionary shapes of visceral leishmaniasis called parasites, specifically of the type amastigote and promastigote. For this purpose, the optical microscopy images containing said evolutionary shapes, which are generated from a blood smear, are subjected to a process of transformation of the color intensity space into a space of intensity in gray levels that facilitate their subsequent preprocessing and adaptation. In the preprocessing stage, smoothing filters and edge detectors are used to enhance the optical microscopy images. In a complementary way, a segmentation technique that groups the pixels corresponding to each one of the parasites, presents in the considered images, is applied. The results reveal a high correspondence between the available manual segmentations and the semi-automatic segmentations which are useful for the characterization of the parasites. The obtained segmentations let us to calculate areas and perimeters associated with the parasites segmented. These results are very important in clinical context where both the area and perimeter calculated are vital for monitoring the development of visceral leishmaniasis.