Early prediction of the highest workload in incremental cardiopulmonary tests
ACM Transactions on Intelligent Systems and Technology (TIST) - Survey papers, special sections on the semantic adaptive social web, intelligent systems for health informatics, regular papers
| Hi-index | 0.00 |
Cardiopulmonary exercise testing is an objective method to evaluate both the cardiac and pulmonary functions. It is used in different application domains, ranging from the clinical domain to sport sciences, to assess possible cardiac failures as well as athete performance. The highest workload reached in the test is a key information to evaluate the individual's physiological characteristics, to plan rehabilitation and/or training sessions. However, these tests are physically very demanding and may expose the tested individual to cardiopulmonary overload. This paper presents a new approach that allows an early prediction of the highest workload that will be reached in the cardiopulmonary test. The test can be prematurely stopped, avoiding its entire execution. The proposed approach relies on a new index, the CardioPulmonary Efficiency Index, which describes the cardiopulmonary response of an individual by summarizing the physiological signals monitored during the test. A k-Nearest Neighbor based classifier analyzes the index trend during the test, together with the characteristics of the individual, and predicts the highest workload. Preliminary experiments, performed on a real dataset provided by the CSA Sport Training Center, showed that the proposed approach is able to effectively predict the highest workload with a limited error since the first steps of the test.