Take-Home Messages

• To achieve increased resolution and specificity in microwave imaging reconstruction a soft prior regularization can be used from pre-microwave imaging MRI performed as part of standard clinical care for breast cancer.
• The targeted application is breast cancer imaging for neoadjuvant chemotherapy monitoring.
• Utilizing a limited set of skin fiducial markers numerical algorithms have been developed to warp the original breast MR images into a geometry closely resembling that for when the breast is pendant in the microwave system.
• Leveraging magnetic resonance images acquired prior to neoadjuvant chemotherapy a soft prior regularization of the microwave imaging can be performed to increase resolution and specificity thus potentially improving the ability of the oncologist to predict chemotherapy effectiveness.

Take-Home Messages

• We have used our microwave-imaging system in order to successfully reconstruct the Supelec-breast phantom and evaluate its usability.
• Our system is capable of recover images of the studied phantoms, although the interior of the Supelec phantom is challenging due to the high plastic content.
• The studied phantom was developed to be a tool within the microwave-imaging community.
• These are the first published reconstructed images of the Supelec phantom and shows that the phantom is a useful tool when developing a microwave-imaging system.
• It is also shown that our system is capable of reconstructing an image of the phantom without utilizing anya prioriinformation of the phantom, which is unique in the microwave-imaging community.

Take-Home Messages

• A novel approach able to deeply optimize clinical treatments of magnetic hyperthermia with nanoparticles.
• Innovative determination of the complex magnetic permeability of magnetic colloidal fluid by simply using invitro SAR measurements.
• A rapid, alternative and broadband approach for determining the electromagnetic properties of magneticnanoparticles.
• Possibility to test the efficiency of nanoparticles in a tissue-like environment: optimized and realistictreatment planning for magnetic hyperthermia.

Take-Home Messages

• We assess the viability of having a dual-band skin-attached repeater antenna for in- and off-body communication for the transmission of medical data in the context of Body Area Networks. It is the first time a single-port antenna with both ISM (2.4 GHz) and UWB (4-10.6 GHz) operation bands is proposed for BAN applications.
• The study demonstrates the implementation of a skin-adhesive antenna and the feasibility of both in- and off-body communication links with a passive implant and an external base station, respectively.
• The proposed repeater antenna can be used in continuous monitoring applications, in order to detect health conditions in advance.
• We study for the first time the possibility of having a skin-attached repeater antenna with capability of establishing an in-body communication link with multiple implants and further relay the collected information through a UWB off-body link, thus increasing the off-body communication range and also diminishing the transmission time.
• The present paper includes the complete study of the in- and off-body links not only in the time-domain, in which we transmit a QPSK-modulated ECG signal, but also in the frequency-domain, where we assess the power transfer and link budget.

Take-Home Messages

• REM Behavior Sleep Disorder is a serious, sometimes dangerous condition that causes people to “act out” their dreams; thus, monitoring and diagnosis of such disease are significant.
• This paper demonstrates the design and implementation of an innovative system for monitoring patients suffering from REM sleep behavior disorder.
• Continuous monitoring of RBD patients is crucial to prevent injury; the proposed identify RBD episodes with accuracy higher than 90%.
• The work leverages the finer-grained information to monitor patients suffering from RBD and fully exploits the amplitude and phase information.

Take-Home Messages

• We propose to combine radar and wearable sensors (accelerometer, gyroscope, and magnetometer) data to simultaneously improve human activities classification rates and minimise false alarms when detecting critical events such as falls.
• Feature selection and fusion methods, at both feature level and decision level, are tested on experimental data collected simultaneously using radar and wearable sensors.
• Sequential forward selection (SFS) significantly helps reduce the dimension of the feature subset and improve the classification accuracy.
• For the feature level fusion, we were able to increase the classification accuracy by 12.2% and 3.3%, compared to the use of radar and wearable sensors on their own.
• Decision level fusion also improves to 94.8% and 96.7% classification accuracy with Fuzzy logic-based decision making and LOGP soft fusion. A novel voting system was also proposed by combining two SVM and KNN classifiers trained by radar and inertial sensor independently. This is demonstrated to achieve 97.8% classification accuracy and 100% fall detection specificity for the ‘fall event’ class.
• Future work will explore the robustness of this multi-sensory approach with additional data collection involving more subjects and more sensors types and configurations.