Algorithms and systems for home telemonitoring in biomedical applications

During the past decades, the interest of the healthcare community shifted from the simple treatment of the diseases towards the prevention and maintenance of a healthy lifestyle. This approach is associated to a reduced cost for the Health Systems, having to face the constantly increased expenditures due to the reduced mortality for chronical diseases and to the progressive population ageing. Nevertheless, the high costs related to hospitalization of patients for monitoring procedures that could be better performed at home hamper the full implementation of this approach in a traditional way. Information and Communication Technology can provide a solution to implement a care model closer to the patient, crossing the physical boundaries of the hospitals and thus allowing to reach also those patients that, for a geographical or social condition, could not access the health services as other luckier subjects. This is the case of telemonitoring systems, whose aim is that of providing monitoring services for some health-related parameters at a distance, by means of custom-designed electronic devices. In this thesis, the specific issues associated to two telemonitoring applications are presented, along with the proposed solutions and the achieved results. The first telemonitoring application considered is the fetal electrocardiography. Non-invasive fetal electrocardiography is the recording of the fetal heart electrical activity using electrodes placed on the maternal abdomen. It can provide important diagnostic parameters, such as the beat-to-beat heart rate variability, whose recurring analysis would be useful in assessing and monitoring fetal health during pregnancy. Long term electrocardiographic monitoring is sustained by the absence of any collateral effects for both the mother and the fetus. This application has been tackled from several perspectives, mainly acquisition and processing. From the acquisition viewpoint a study on different skin treatments, disposable commercial electrodes and textile electrodes has been performed with the aim of improving the signal acquisition quality, while simplifying the measurement setup. From the processing viewpoint, different algorithms have been developed to allow extracting the fetal ECG heart rate, starting from an on-line ICA algorithm or exploiting a subtractive approach to work on recordings acquired with a reduced number of electrodes. The latter, took part to the international "Physionet/Computing in Cardiology Challenge" in 2013 entering into the top ten best-performing open-source algorithms. The improved version of this algorithm is also presented, which would mark the 5th and 4th position in the final ranking related to the fetal heart rate and fetal RR interval measurements performance, reserved to the open-source challenge entries, taking into account both official and unofficial entrants. The research in this field has been carried out in collaboration with the Pediatric Cardiology Unit of the Hospital G. Brotzu in Cagliari, for the acquisition of non-invasive fetal ECG signals from pregnant voluntary patients. The second telemonitoring application considered is the telerehabilitation of the hand. The execution of rehabilitation exercises has been proven to be effective in recovering hand functionality in a wide variety of invalidating diseases, but the lack of standardization and continuous medical control cause the patients neglecting this therapeutic procedures. Telemonitoring the rehabilitation sessions would allow the physician to closely follow the patients' progresses and compliance to the prescribed adapted exercises. This application leads to the development of a sensorized telerehabilitation system for the execution and objective monitoring of therapeutic exercises at the patients' home and of the telemedicine infrastructure that give the physician the opportunity to monitor patients' progresses through parameters summarizing the patients' performance. The proposed non-CE marked medical device, patent pending, underwent a clinical trial, reviewed and approved by the Italian Public Health Department, involving 20 patients with Rheumatoid Arthritis and 20 with Systemic Sclerosis randomly assigned to the experimental or the control arm, enrolled for 12 weeks in a home rehabilitation program. The trial, carried out with the collaboration of the Rheumatology Department of the Policlinico Universitario of Cagliari, revealed promising results in terms of hand functionality recovering, highlighting greater improvements for the patients enrolled in the experimental arm, that use the proposed telerehabilitation system, with respect to those of the control arm, which perform similar rehabilitation exercises using common objects.