Ricerc@Sapienza

1. Wearable Sensors (Collaboration with STMicroelectronics)

Parkinson’s Disease Patients affected by Parkinson’s Disease suffer from severe impairments essentially in the motion sphere. Freezing of Gait (FOG) is probably the most disabling gait disorder. FOG is unpredictable, with leg trembling and the shuffling-like types of clinical manifestations, trunk oscillations and imbalance associated to an increased fall probability. It is essential monitoring this disorder during the daily life in order to assess the disease stage, optimize the pharmacological therapy and prevent falls. In this framework, wireless body area network (WBAN) is playing a role to monitor different medical signs of patients with the support of sensors. This concept has revolutionized the way in which patients’ data is collected. We are engineering different configurations of wearable systems for the automatic, continuous and ubiquitous analysis of FOG episodes. The targets of this research are optimizing the system hardware and software for detecting the FOG episodes, distinguishing the clinical manifestations and analyzing the muscle activity during and outside FOG. To this aim, wearable boards need to be engineered, with gyroscopes and surface electromyography (sEMG) integrated together with an ultra-low-power microprocessor, a low-energy Bluetooth unit and a power supply with a battery. Sensors record simultaneously movements and action potentials of antagonist leg muscles, while dedicated algorithms allow the timely detection of the FOG episode and, for the first time, the automatic distinction of the FOG phenotypes, which enable associating a fall risk to the subtype. Some functionalities of the system operate in real-time with a direct feedback to the patient, some other off-line for the electronic agenda. In the latter case, data are transmitted wireless to an external workstation. The hardware and software need to be optimized with respect to power consumption related to data sensing, pre-processing, transmission. An international patent is pending on a particular configuration of the system, using only two devices in a master-slave hierarchy. It allows real time operation with a consistent reduction of the power consumption and longer device autonomy.

Polygraphic monitoring Wearable technologies with extremely low power consumption would open to a variety of applications in health and wellness strongly desired today.  First of all, low power devices would enable to acquire synchronously more than one type of biopotentials (i.e., polygraphic recording) for hours and days, being compact and comfortable at the same time, and this would be a breakthrough for many applications in medicine. Furthermore, at the present moment, the current processing power and battery life of wearables might constrain the upcoming use of sophisticated machine learning algorithms, which, on the contrary, forecast invaluable advantages unexplored to date.

We focused on the realization of a monolithic multi-sensor platform in MEMS technology featuring commercial charge transfer sensor (Qvar) originally intended for presence tracking, to acquire an ECG trace and a single-channel EEG along with EOG. Not only charge transfer provides an incredibly low power, unexplored solution to biopotentials acquisition but it also enables adding this type of functionality to any existing MEMS. This means embedding even more functionalities to MEMS boards at near-zero additional power consumption (< 20 mA per channel). For these reasons, those electrostatic sensors open to additional possibilities for wearable sensing and strengthen the role of MEMS technology in medical wearable applications for long time synchronous acquisition of ECG, EEG, EOG, inertial signals.

Sleep quality REM behavior disorders (RBD) are disturbs considered prodromal of cognitive impairment in neurodegenerative disease as Parkinson’s Disease and Alzheimer disease. For this reason, neurologists strongly desire a non-invasive tool for continuous monitoring of REM sleep quality in domestic environment. On the contrary, today, REM sleep classification is usually performed by cumbersome and obtrusive polysomnography, with the patient hospitalized. As an alternative approach, we proposed a wearable system for a preliminary pervasive screening to be performed at home. Starting from three well known features of the REM sleep (random eye movements, complete atony of the masseter/mentalis muscle, heart rate reduction) we focused on the possibility to realize synchronous acquisition of those three biopotentials (ECG, electrooculogram -EOG-, surface electromyography -sEMG-) for a whole night in the best patient condition, i.e. at home. In particular, the surface electromyography is in charge of recording the masseter/mental muscle atony, the (artifact of the) EEG on frontal lobe is in charge of detecting the random eye movements, the ECG is in charge of recording decrease of heart rate (in alternative also a PPG on the frontal can be used for this scope). Taking advantage of the low power wearable system developed in collaboration with STMicroelectronics, we verified that the electrostatic sensors Qvar is also able to monitor atony of those muscle and tested the whole system validating the approach.

Another activity related to sleep focusses on Sleeping Apnea (SA), which is a big issue in medical regarding the neurological and otolaryngology sphere, with severe repercussions in the patient daily health and well-being. SA is a sleeping disorder, which can have two origins, i.e. obstructive or central, and requires urgent treatment. It causes abnormal pauses in the breathing during sleep, which puts a patient very close to death. In SA, insufficient air flows through the nose and mouth into the lungs, regardless of breathing effort continues, due to which blood oxygen reduces. The very short drops in the blood-oxygen levels can decrease the ability to concentrate, think, learn and understand; moreover, it also causes morning headaches. SA is traditionally monitored with Polysomnography (PSG). PSG test is performed in a sleeping hospital laboratory or sleeping center. The patient is likely to stay overnight with electrodes and other monitors placed on the chest, face, scalp, and fingers. SA can only be monitored during sleep, more than 22 wires are attached to the patients, which is very uncomfortable for them. For the detection of SA, patient does not have to stay in the hospital like in PSG test. We are engineering a light and compact integrated device supporting photoplethismography (PPG) monitoring SpO₂ in blood, through reflection analysis of green, red and infrared lights. The device also integrates inertial sensors for distinguishing the central from the obstructive type of apnea, and an electrocardiogram sensor. Finally, low-power computing and data transmission units, a battery with the power supply circuitry are present. The device will be positioned on the nose or, better, on the neck, with very low impact on the sleeping patient. It will collect data over the night and continuously transmit them to a tablet/PC positioned on a bedside table. It is not possible to wake-up the patient in the case of detection of a SA episode, so the algorithms only operates off-line. The algorithms will need to detect all the episodes and their exact duration, distinguish the central from the obstructive apnea type. This will enable doctors to make the correct diagnosis and implement the proper measures.

Hypoglicemia Diabetes is a metabolic disorder characterized by high blood glucose that affects, today, approximately half a billion persons worldwide, with an estimated year cost of around 10% of the global health expenditure. Insulin administration in general cannot fix precisely the physiological value, and easily leads to dangerous glucose oscillations. Nocturnal hypoglycemia is particularly dangerous because its symptoms may be blurred by sleep, bringing to degradation of physical and mental abilities, heart arrhythmia, coma and even death. In spite of the gravity of those symptoms, the physiology of hypoglycemia is not known. We proposed an innovative wearable device recording simultaneously the ECG and the EEG alpha waves for monitoring vital signs. First, we identified the minimum set of parameters that can be extracted from ECG and EEG traces in time (amplitude) and in frequency (power spectral density-PSD-) domains that alter hundreds of seconds before the onset of a hypoglycemia episode. Then, we designed a compact, comfortable, easy-to-use, wearable device able to record the two biopotentials synchronously. The hard system consists essentially of two integrated boards (each mounting an electrostatic sensor, a microcontroller, a battery, a microSD memory) fixed on the chest: one sensor acquires ECG with two electrodes, the other sensor acquires the EEG with two electrodes on the occipital sites. It is wearable and comfortable, with just one chest band and four electrodes; the battery life is very long (days) thanks to the sensor very low power consumption; data are stored locally on a microSD. Preliminary verification of correct operation has been performed by comparison with a gold standard. Test on a patient was also performed in domestic environment. Results are encouraging

2. Neuromorphic circuits with NAND memory technology (Collaboration with Micron Semiconductors)

This activity focusses on the analysis and simulation of neuromorphic circuits based on artificial synapses made of solid-state memories in 3D 3bit per cell V-NAND Flash technology. An international patent is pending.

3. Reliability of SuperFlash ESF3 memories (Collaboration with Infineon)

The research aims to look for new strategies of operating ESF3 memory cells in order to overcome current problems in programming, erase, reading without degrading reliability. The approach is at a simulation and experimental level, using the state-of-art node of TSMC-Infineon technology.

4. Performance and reliability of CMOS-image sensors (Collaboration with LFoundry)

This activity focusses on the performance and reliability degradation of Back-Side-Illuminated (BSI) respect to Front-Side-Illuminated (FSI) configuration of CMOS image sensors. An additional distribution of oxide traps in BSI has been found to be originated during the plasma etch step for the Through Silicon Via opening on the wafer back-side. Influence on the performance and reliability is under investigation, together with possible technological solutions at the process level.