This document discusses the Chronious project, which uses soft computing techniques on smart devices to monitor chronic diseases. Specifically, it monitors elderly patients over 60 affected by chronic obstructive pulmonary disease (COPD) and chronic kidney disease (CKD). The system combines remote sensors, a smart device, and a central decision support system using an ontology, literature search, and rule-based inference engine. It discusses the communication framework, algorithms run on a PDA device, and classification systems used, including rule-based and supervised learning approaches. Issues addressed include resource consumption on the PDA and difficulties updating training algorithms.
18. What I’m here to present ? Soft computing techniques + Smart devices + Chronic diseases in
19. Why Chronious COPD? COPD: by 2020 3.5 million deaths in the world (at least) USA 2000: 10 million adults reported COPD = 8 million physician office and hospital outpatients visit + 1.5 million emergency visit + 726,000 hospitalizations + 119,000 deaths
22. Why Chronious COPD? Another cause of COPD in development countries
23. Free advice? Want to reduce risk of contracting COPD? DON’T SMOKE! DON’T DRIVE CARS! DON’T DO BARBECUE!
24. Why Chronious CKD? In the US, 9.6% of non-institutionalized adults are estimated to have CKD Reducing the mortality rates associated to the CKD could save 10% of the loss extimated in 8 bilion USD only in development countries
25. Why Chronious CKD? CKD is difficult to treat because of comorbidities The renal functionality becoming worse at every exacerbation We choose to go to the moon in this decade and do the other things, not because they are easy, but because they are hard (J.F.Kennedy)
40. Controlling the stress index Bayesian network Use 9 attributes for evaluating a stress index that can be used to understand if the patient condirion can worse.
41. Problems with the PDA Heavy resource consumption -> bottleneck during preprocessing phase in case of alerting situation Difficulties on updating the training algorithms