PHA for Hypertension


Hypertension is one of the most prevalent diseases of the modern world. According to the World Health Organization (WHO), hypertension affects more than 40% of the adults over the age of 25. In the early stages hypertension is mostly asymptomatic – this makes early detection difficult. When symptoms start showing, it is usually after there has been a significant damage to the heart, arteries and other organs. This makes early detection of hypertension is of paramount importance. Standard methods of diagnosis includes checking for increase in blood pressure. However, most people usually have their blood pressure measured only during clinical visits – which happens once or twice a year. Doctors advocate the use of ambulatory blood pressure measurement devices for continuous monitoring of blood pressure at home for early diagnosis. However, continuous out-of-clinic monitoring of blood pressure is cumbersome – the cuff-based devices depend on compression of the arm for measurement making measurement during sleep and activity difficult. Therefore there is a need for investigating the use of other physiological signals for hypertension detection. In our research we investigate whether physiological signals which can be measured using off-the-shelf wearable devices can be used for the detection of hypertension.

Progression of hypertension has been shown to produce functional and structural changes to the cardiac and vascular system, and early markers for these changes can often be detected even before elevated blood pressure is detected. Indicators for sympathetic activation, such as Heart Rate Variability (HRV) have been shown to be correlated with incidence of hypertension. In our research we investigate how various physiological signals such as heart rate variability, galvanic skin response, skin temperature, and blood volume pulse and their combinations can be used for the continuous prediction of stress. We also demonstrate that a combination of these physiological signals can achieve a high degree of precision in detecting hypertension.

Once detected, management of hypertension must consist of a combination of both pharmacological as well as non-pharmacological approaches. Lifestyle modifications such as increase in physical activity and exercise, reduction of salt, following a healthy diet and decrease in alcohol consumption and smoking have been shown to be beneficial for patients. In this research work we develop intelligent personal healthcare agent technologies which along with wearable devices can continuously monitor the physiological signals of subjects – while using intelligent interaction techniques to obtain regular structured and unstructured annotations and responses for stress, workload, activity levels, food and beverage consumption, and mental state.

Such an agent learns about the lifestyle of the user, and is continuously able to monitor the various covert (physiological) and overt (speech and language interactions, activity, behaviour, habits) signals of the user. Through this the personal agent can increase a user’s awareness and adherence towards a healthy lifestyle thus helping manage the condition. Through the early detection of lifestyle factors which might contribute to the development of hypertension, such an agent can also help in the prevention of the incidence of the disease. Using state-of-the-art signal processing and machine learning techniques we demonstrate that such an agent can combine the covert and overt signal streams to achieve a high degree of precision in detecting stress and hypertension.


Danieli M., Ghosh A., Berra E., Fulcheri C., Rabbia F., Testa E., Veglio F., Riccardi G., “Automatically classifying essential arterial hypertension from physiological and daily life stress responses.” ESH 2016 – The 26th  European Meeting on Hypertension and Cardiovascular Protection, Paris, France, June 10 -13 2016.

Danieli M., Ghosh A., Berra E., Rabbia F., Testa E., Veglio F., Riccardi G., Comprendere l’ipertensione arteriosa essenziale a partire da costrutti psicologici e segnali fisiologici.” XXXII Congresso Nazionale della Societa Italiana dell’Ipertensione Arteriosa 2015

Ghosh A., Torres J.M.M., Danieli M., Riccardi G., Detection of Essential Hypertension with Physiological Signals from Wearable Devices”, Engineering in Medicine and Biology Society (EMBC), 2015 37th Annual International Conference of the IEEE.

Ghosh A., Danieli M., Riccardi G., Annotation and Prediction of Stress and Workload from Physiological and Inertial Signals.” Engineering in Medicine and Biology Society (EMBC), 2015 37th Annual International Conference of the IEEE