Claudio Crema

A smartphone-enhanced pill-dispenser providing patient identification and in-take recognition

The wider and wider availability of powerful, low-cost mobile devices (e.g., smartphones or tablets) is deeply changing healthcare, so that the mHealth term has been coined. The announcement of healthcare projects by market big players as Apple and Samsung confirms this trend. In particular, the opportunity to collect reliable patient data automatically allows to enhance patient/user self-management and helps in better delivering therapies. In this paper, authors propose an innovative architecture for a smart pill-dispenser enhanced by a smartdevice that furnishes the capability of automatically identifying the user, other than logging medicine in-take activities. A real-world prototype, based on an emulated pill-dispenser connected via an NFC link to different smartdevices, has been purposely realized. Experimental tests confirm the architecture feasibility. Low-cost requirements are satisfied and a user-friendly interface has been implemented.

Efficient R-peak detection algorithm for real-time analysis of ECG in portable devices

Detection of R-peaks in an electrocardiogram (ECG) acquisition is the primary goal of any algorithm for the automatic processing of ECG signals. Several methods have been proposed in the past to accomplish this task, but usually they are designed for the use in clinical situations, where strict real-time requirements are not always needed and availability of devices with high computational resources is not a problem. The recent and broad success of personal and portable devices for health monitoring has opened a new scenario, in which new algorithms need to be developed in order to satisfy constraints due to limited computational resources and the need of having embedded and real-time data processing. This work describes an efficient algorithm for the R-peak detection, designed for low-cost and low-performing portable devices. Performance of the proposed approach, evaluated with a set of well-known ECG waveform, is comparable with traditional methods found in literature. Finally, the successful implementation of the algorithm in a smartphone-based low-cost ECG acquisition system has validated the feasibility of the proposed approach.

The WearPhone: Changing smartphones into multichannel vital signs monitors

Traditional healthcare solutions oblige patients to stay in hospital; bio-signal monitoring and treatment dosing require the presence of specialized personnel. In turn, this scenario implies high-cost, especially for chronic diseases or fragile people surveillance. Wearable sensor-based healthcare solutions may be an effective approach for lowering costs, since they offer good performance and can be used directly at home, e.g., for permanent self-monitoring. In addition, their communication capabilities make easier to exchange information with remotely located professionals. Wearables are smart objects, consisting of sensors, processing units and communication modules, which rely on a smartphone for data collection, display and remote communication. In this paper, authors modify this paradigm moving the computational capability out of the wearable into the smartphone; in this way, raw vital signal stream can be obtained, overcoming possible limitations of the wearable itself. The aim of the work is to verify multichannel capability of the proposed approach, which paves the way to multi-parametric bio-signal analysis. An experimental setup has been arranged using a real-world smartphone and tests have been carried out to evaluate the performance. In particular, it has been verified that three ECG-like signals can be simultaneously collected leveraging on simple multiplexing in the frequency domain.

IMU-based solution for automatic detection and classification of exercises in the fitness scenario

Causal relationship between physical activity and prevention of several diseases has been known for some time. Recently, attempts to quantify dose-response relationship between physical activity and health show that automatic tracking and quantification of the exercise efforts not only help in motivating people but improve health conditions as well. However, no commercial devices are available for weight training and calisthenics. This work tries to overcome this limit, exploiting machine learning technique (particularly Linear Discriminant Analysis, LDA) for analyzing data coming from wearable inertial measurement units, (IMUs) and classifying/counting such exercises. Computational requirements are compatible with embedded implementation and reported results confirm the feasibility of the proposed approach, offering an average accuracy in the detection of exercises on the order of 85%.

Virtual Respiratory Rate Sensors: An Example of A Smartphone-Based Integrated and Multiparametric mHealth Gateway

In the last few years, several wearables appeared in the market, for fitness and healthcare applications. Such smart devices have been proposed as a possible solution for lowering the costs of healthcare, leading to the mHealth revolution. In the typical scenario, each wearable, embedding sensors, processing units and communication modules, adopts a smartphone for data collection, data displaying, and remote communication. In this paper, authors modify this paradigm simplifying the wearables (e.g., relying only on simple analog front ends and communication interfaces) and exploiting the (relatively large) computational capability of the smartphone, not only for implementing gateway features but also for processing raw biosignals as well. Several experiments verify the feasibility of the proposed approach and demonstrate that “local” biosensor virtualization is possible, expanding possibilities of mHealth. In particular, tests have been carried out to evaluate the performance of hearth rate computation and respiratory rate virtual sensor, starting from a single-lead electrocardiogram signal.

Kalman-Based Approach to Bladder Volume Estimation for People with Neurogenic Dysfunction of the Urinary Bladder

People with neurogenic dysfunction of urinary bladder often require daily catheterism because of their impairment. This issue is particularly critical for those that have not the urinary stimulus, because they have not the ability to understand when the bladder is full or not. From user’s point of view, the absence of a urinary conscious stimulus can cause refluxes, damaging patient’s health and his psychological status. For such necessities, most patients require professional nursing, increasing the work of the staff and the overall medical costs. Furthermore, catheterism itself applied every day for a long period can cause infection in the urinary tract. The authors propose a non-invasive bladder monitoring system based on real-time bioimpedance measurement. A Klaman filter was developed in order to estimate the bladder volume due to the intrinsic-uncertainly of the model itself and to remove the artifacts due to patient’s movements by using accelerometer by monitoring it’s activity. Theoretical analysis, in-system measurements and experimentations prove the effectiveness of the proposed solution.