Advanced search
Start date
Betweenand

Scoliosis classification using baropodometric data and machine learning techniques

Abstract

Scoliosis Classification Using Baropodometric Data and Machine Learning Techniques.ABSTRACT Although procedures for identifying idiopathic scoliosis are well established, there are many questions to be considered when it is necessary to make its diagnosis. For example, its dependence on X-rays is a factor, in many cases, due to costs, radiation exposure, availability, etc. Research carried out recently at LIEB (Laboratory of Instrumentation and Biomedical Engineering, UNESP) showed that the baropodometer is an instrument that can be used to classify patients with and without scoliosis. The main objective of this project is to develop procedures to classify idiopathic scoliosis using data from the plantar pressure distribution, stabiliometric, and biometric data of volunteers, and Machine Learning techniques. The data will be obtained with baropodometer with the patient not only in the standing position, but also in the gait initiation. The baropodometer will be validated with a force platform, the reference instrument for measuring the individual's center of pressure and in postural analysis. Volunteers will be classified into three groups, according to the Cobb angle: 1) Cobb<10o; 2) 10odCobb<20o and Cobbe20o. With the performance obtained with different techniques of machine learning, it will be possible to establish the feasibility of using baropodometric systems as idiopathic scoliosis identification and classification systems. As a result, the overall accuracy of scoliosis classification is expected to be greater than 75%. Due to the innovative nature of the research, considerable scientific contributions are expected. One of them will be the implementation of a database with baropodometric, stabiliometric and biometric data of 300 volunteers, which will be totally new and very useful for researchers and health professionals. (AU)

Articles published in Agência FAPESP Newsletter about the research grant:
Articles published in other media outlets (0 total):
More itemsLess items
VEICULO: TITULO (DATA)
VEICULO: TITULO (DATA)