Conference Agenda

This systematic literature review (SLR) aims to analyze medical image processing techniques based on artificial intelligence (AI) and machine learning to improve diagnostic accuracy and efficiency in medical institutions. Based on a search of the SCOPUS database, 2757 articles were identified, from which, after applying inclusion and exclusion criteria, 53 relevant studies were selected. The results show that machine learning techniques overcome the limitations of traditional methods by increasing diagnostic accuracy and reducing processing times. Major advances include the use of deep learning algorithms and automation in image analysis, although challenges remain, such as reliance on trained operators, technical constraints in certain institutions, and the need for broader clinical validation. In conclusion, SLR highlights the transformative potential of AI in medical diagnostics, with the potential to optimize clinical processes and outcomes. However, to maximize its effectiveness, it is crucial to overcome the identified technological and implementation barriers, thus ensuring its sustainable integration into medical practice.

The article presents an innovative approach for diabetes prediction by applying a phantom deep learning model, designed to optimize accuracy and efficiency in early diagnosis. This model uses a robust clinical data set, integrating advanced artificial intelligence techniques with a lightweight and efficient architecture, allowing computational costs to be significantly reduced without sacrificing predictive performance. Key features of the model include its ability to handle imbalanced data sets, common in medical environments, and its adaptability to various clinical contexts, making it especially versatile. Experimental results demonstrate that the phantom deep learning model greatly outperforms conventional methods such as artificial neural networks (ANNs) in terms of accuracy and efficiency. Specifically, the phantom model achieved an accuracy of 79.24% without overfitting in identifying patients at risk of diabetes, compared to the 88.57% obtained by the conventional ANN model with overfitting. In addition, the loss of the phantom model has a greater generalization capacity and lower error in predictions. These results highlight the superiority of the phantom model in terms of accuracy and stability. Its scalability and low resource requirements make it a viable option for implementation in public and private health systems; The phantom deep learning model represents a significant advance in the application of artificial intelligence in the healthcare field.

The inspection of surface-mount device (SMD) components on printed circuit boards (PCBs) is crucial to ensuring quality in electronic manufacturing. Conventional methods often lack precision and speed, resulting in defects and higher costs. This study compared two advanced neural networks, Roboflow 3.0 and YOLO v11, to address these challenges. Using a dataset of 1,300 images that included capacitors, resistors, and transistors under varying conditions, the models were trained and evaluated based on metrics such as mAP50, mAP50:95, precision, and recall. The results showed that Roboflow 3.0 achieved superior performance with a mAP50 of 95.6% and a mAP50:95 of 64.9%, along with consistent improvements in precision and recall during incremental training. In contrast, YOLO v11 demonstrated stability but achieved lower metrics, with a mAP50 of 90.2% and a mAP50:95 of 61.3%. While both models offered robust detection capabilities, Roboflow 3.0 excelled in adapting to diverse variations in lighting and geometry. This study highlights the potential of convolutional neural networks to transform PCB inspection in quality environments, offering greater precision and efficiency, thereby reducing human errors and associated costs while optimizing production processes.

Breast cancer is one of the most significant health problems worldwide, and its early detection is crucial to improve clinical outcomes for patients. In this context, machine learning models have become valuable tools to predict the presence of this disease with greater precision. This study performs a comparative analysis of three machine learning models: logistic regression, random forest, and support vector machines (SVM), using the Wisconsin Breast Cancer Diagnosis dataset. This data set includes features derived from fine-needle aspiration images of breast masses, with 357 benign and 212 malignant cases. The results of the study reveal that the random forest model outperforms the other two in terms of predictive accuracy. This model, which uses the top 5 predictors ("concave point mean", "area mean", "radius mean", "perimeter mean", and "concavity mean"), achieves an accuracy of about 94.15% and a cross-validation score of about 95.61% on the test data set. These findings highlight the effectiveness of random forest in identifying complex patterns in data, making it a promising tool for breast cancer prediction. In conclusion, this study demonstrates the potential of machine learning models, particularly random forest, in improving early detection of breast cancer. These advances could have a significant impact on clinical practice, facilitating more accurate and timely diagnoses, which in turn could improve patient outcomes and reduce mortality associated with this disease.

This paper addresses the development of a convolu-
tional neural network (CNN) model capable of detecting diseases
in cocoa fruits, specifically black pod and moniliasis, using
images collected in the field and processed with the Roboflow
platform. These diseases represent a significant challenge for
farmers due to the economic losses they generate, highlighting the
importance of early and accurate detection. 2,000 representative
images were collected, adjusted with saturation variations (+25%
and -25%) and capture distances (10 cm and 30 cm), which
were used to train specialized neural networks. The developed
models achieved outstanding metrics, exceeding 95% accuracy
and recall in the detection of both diseases in the mixed network.
Among the designed networks, the network focused on black cob
showed a performance higher than 96%, while the network for
moniliasis obtained slightly lower, but satisfactory results of 91%,
highlighting the relevance of representative data and iterative
training to optimize the model performance. The conclusions
highlighted the effectiveness of the model developed, the relevance
of the quality and diversity of the data collected, and the
positive impact that this technology can have on agricultural
management.