Proposals for Osseus Improvement
Osseus; Osteoporosis; Diagnosis; Machine learning; Radiofrequency.
Osteoporosis is a systemic osteometabolic disease that attacks an increasing portion of the Brazilian population, which generates significant costs to the Brazilian Unified Health System (SUS) and that with previous diagnosis preventive measures can be implemented, which may avoid fractures and reduce SUS costs. In order to help in this prevention and consequent cost reduction, it was proposed the development of Osseus, an instrument that combines techniques and concepts from various areas such as software engineering, electrical, electronics, computing and biomedical. In addition, the equipment was proposed to be developed with low cost, to be easily accessible to the population and noninvasive, developed in the Laboratory of Health Innovations and Technologies (LAIS). The latest version of Osseus, 3.0, presented some problems that prevent it from being implemented on an industrial scale. These problems are related to some operating instabilities observed in the radio frequency (RF) part. Thus, this work proposes the implementation of improvements in Osseus, aiming to eliminate these instabilities and propose the use of new computational intelligence algorithms to be used in the equipment. Among the proposed improvements is the change of the operating frequency from 2.45 GHz to 5.8 GHz, aiming at the miniaturization of the antennas and the reduction of the equipment dimensions, as well as providing a better operation of the RF part. In addition, a study will be done on new machine learning tools consisting of classification algorithms such as KNN (K-Nearest Neighbors) and XGBOOST (Extreme Gradient Boosting). The results obtained with this new proposal of Osseus will be compared with the results obtained with Osseus 3.0. This is expected to improve the equipment and provide a low cost device, helping LAIS to fulfill its mission, which is to make science an instrument of love for others.