Periodic Reporting for period 2 - HEARTEN (A co-operative mHEALTH environment targeting adherence and management of patients suffering from Heart Failure)
Reporting period: 2016-07-01 to 2018-03-31
Based on the final list of selected biomarkers, the corresponding biosensors were developed as part of the WP5 activities. For breath biosensors, no devices are available for the detection of volatile organic compounds for monitoring of HF patients. The project has successfully developed acetone sensors that operate at room temperature and have a detection limit of 1 ppm. This is not adequate for general purpose HF monitoring because the median exhaled acetone concentration for healthy people is 0.65 ppm, but it is suggested that the sensors can be used for HF patients with diabetes mellitus because in this case the ranges are higher.
No biosensors are available for the detection of biomarkers relevant for HF monitoring in saliva. The project has successfully developed biosensors for TNF-α, cortisol, IL-10 and NT-proBNP with the required sensitivity and selectivity. However, for IL-10 no validation has been achieved due to the lack of results from standard analytical methods, and for NT-proBNP results are only available for artificial saliva. Quantitative measurements in real saliva samples have required the use of the standard addition method. In the project for the first time the standard addition method and impedance spectroscopy have been used for quantitative measurements of biomarkers in real saliva samples. The biosensors developed operate with a high performance demonstrated by good response, recovery, stability and repeatability in the complex saliva matrix. The KMS developed in this WP is a novel tool supporting management of HF patients, consisting of 9 distinct modules. Combining different patient data (i.e. clinical, sensor –clinical and movement-, nutrition and biomarker related data) the Hearten KMS provides a novel solution for management and self-management of HF patients: it monitors patient status in an objective way (instead of the subjective NYHA class estimation), it early identifies adverse events, estimates risk for non adherence, and it estimated treatment adherence in terms of medication, nutrition and physical activity. Additionally, it supports research and hypothesis testing. In cases were a comparison to relevant approaches in literature is feasible –even if implemented differently and using different data (i.e NYHA class detection and adverse event prediction)- KMS demonstrates in general higher performance (e.g. in terms of accuracy). Using data mining/machine learning techniques and a novel selection and combination of different types of patient data the HEARTEN KMS achieves high diagnostic and prognostic accuracy. All the above are offered through a novel cloud setting. Concerning the app, an innovative, yet easy to use system was designed and developed which allows for efficient management of a heart failure patient.