Sessions

CANCELLED

Patients' remote monitoring platforms can be enhanced with intelligent recommendations and gamification functionalities to support their adherence to care plans. The current paper aims to present a methodology for creating personalized recommendations, which can be used to improve patient remote monitoring and care platforms. The current pilot system design is aimed to support patients by providing recommendations for Sleep, Physical Activity, BMI, Blood sugar, Mental Health, Heart Health, and Chronic Obstructive Pulmonary Disease aspects. The users, through the application, can select the types of recommendations they are interested in. Thus, personalized recommendations based on data obtained by the patients’ records anticipated to be a valuable and a safe approach for patient coaching. The paper discusses the main technical details and provides some initial results. 

Traditional sample size estimation for sufficient model performance is not applicable for machine learning, especially in the field of electrocardiograms (ECGs). This presentation outlines a sample size estimation strategy for binary classification problems on ECGs. The post-hoc sample size estimations are based on a benchmark across different architectures and different classification targets. The results indicate trends for required sample sizes for given tasks and architectures, which can be used as orientation for future ECG studies or feasibility aspects.

Electroencephalography (EEG) has recently gained popularity in user authentication systems since it is unique and less impacted by fraudulent interceptions. Although EEG is known to be sensitive to emotions, understanding the stability of brain responses to EEG-based authentication systems is challenging. In this study, we compared the effect of different emotion stimuli for the application in the EEG-based biometrics system (EBS). Initially, we pre-processed audio-visual evoked EEG potentials from the ‘A Database for Emotion Analysis using Physiological Signals’ (DEAP) dataset. A total of 21 time-domain and 33 frequency-domain features were extracted from the considered EEG signals in response to Low valence Low arousal (LVLA) and High valence low arousal (HVLA) stimuli. These features were fed as input to an XGBoost classifier to evaluate the performance and identify the significant features. The model performance was validated using leave-one-out cross-validation. The pipeline achieved high performance with multiclass accuracy of 80.97% and a binary-class accuracy of 99.41% with LVLA stimuli. In addition, it also achieved recall, precision and F-measure scores of 80.97%, 81.58% and 80.95%, respectively. For both the cases of LVLA and LVHA, skewness was the stand-out feature. We conclude that boring stimuli (negative experience) that fall under the LVLA category can elicit a more unique neuronal response than its counterpart the LVHA (positive experience). Thus, the proposed pipeline involving LVLA stimuli could be a potential authentication technique in security applications.

We suggest a process for finding elements of interest from FHIR profiles to support the setup of data quality assessments for data stored in FHIR servers.