Periodic Reporting for period 1
Reporting period: 2015-11-01 to 2017-10-31

Summary of the context and overall objectives of the project

Heart failure (HF, inability of the heart to pump blood) is a disabling, deadly and costly condition. Worldwide, HF is recognized as an escalating public health problem, likely to reach ‘epidemic’ proportions in developed countries. Healthcare cost of HF treatment in developed countries consumes 1–2% of the total health care budget. HF is the single most expensive diagnosis in the US health care system, and total costs for HF care were estimated at $37 billion in 2009.
HF is as ‘malignant’ as many common types of cancer: 50% of patients die within 5 years once diagnosed for HF. However, HF is often difficult to diagnose in the primary care setting, and many patients remain undiagnosed and untreated. In most developed countries, there are routine cancer screening programs (e.g. mammography). However, there is a general lack of screening programs for the early detection of HF.
Diagnosis of HF is made based on symptoms and clinical signs. Unfortunately, the most frequent signs and symptoms of the HF (fatigue, breathlessness, tiredness, and palpitation) are nonspecific and occur in majority of common disease (viral infections, anaemia, depression, etc.). In the early stages of the diseases, symptoms occur only during physical effort. ECG changes are either absent or nonspecific. The most useful diagnostic test for HF is the echocardiogram (ECHO). However, duration of the ECHO test is 20-30 minutes, it needs highly trained personnel, and has prohibitively high cost. Thus, ECHO is not suitable for use as a screening test, and is not available in primary care (family practitioners, GP’s). As such, ECHO is not generally available for a large number of potential patients with HF. Thus, the diagnosis of heart failure is usually made at a very late stage of the disease.
Early detection of HF would require a diagnostic screening test that would be simple, fast, operator-independent and low cost technology (ECG is an example of such good screening test in cardiology, however, it has poor specificity in detecting HF).
European start-up company Diasens develops a new device and method which will combine two conventional diagnostic technologies, ECG (electrocardiograph) and phonocardiograph (PHONO, electronic stethoscope), with novel fibre Bragg grating (FG) optical sensors used for detection of the movement of thorax surface caused by the movement of the heart apex and with CSC laser or FG sensors for measurement the blood pressure pulse waveforms in one of the major arteries.
The integration of four different diagnostic signals into a single device combined with the sophisticated data analyses methods will enable extraction of the hidden diagnostic markers for early detection of HF and other heart conditions (LVH-Left Ventricular Hypertrophy, valvular disease, etc.) that are not efficiently detected by the conventional ECG.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

In the first reporting period the RISE CARDIALLY consortium have achieved the following results:

We developed a new technique and the horizontal visibility graph to the experimentally measured laser output intensity. This reveal the presence of temporal correlations during the transition between two states of laser operation. Both methods allow us to unveil coherent structures with well defined time-scales and strong correlations both, in the timing of the laser pulses and in their peak intensities.

In two years 260 patients with out-of-hospital cardiac arrest treated by the emergency medical services in Brescia, Italy, were considered. Advanced cardiac life support management, ECG and all relevant demographic information were documented and recorded according to the Utstein guidelines. A detailed and complementary evaluation has been performed by two experienced cardiologists from Medical school, University of Belgrade, Serbia, organized and supervised by MD Vladan Vukcevic. This way we derived a final data set for developing and testing developed statistical methods for diagnostic data analysis.

ECG tracings have been characterized in details through various amplitude and frequency measures and nonlinear measures. We have considered 26 features: Root Mean Square, Segment average, Mean Amplitude, Wave Amplitude, Maximum Amplitude, Minimum Amplitude, Peak to Trough, Amplitude range, Mean Slope, Median Slope, Signal Integral, Amplitude Spectrum Area, Power Spectral Analysis, Energy, Centroid Frequency, Centroid Power, Dominant Frequency, Edge Frequency, Max Power, Spectral Flatness Measure, Median stepping increment, Wavelet energy in the 1-3 Hz, 3-10 Hz, and 10-32 Hz bands. To evaluate the outcome of the defibrillation, we have investigated different durations (from 1 to 9s) of the ECG signal prior to the shock delivery. For every segment all the features were calculated.

During the first two years, efficient numerical algorithms were developed, for qualitative and quantitative studies and prediction of universal dynamical behaviour and complex pattern formation in ECG tracings. We developed a unified approach, based on the scalar and vector coupled Ginzburg-Landau equations, which unifies the dynamics of nonlinear waves of fiber lasers and of nonlinear excited waves in the human heart.

The multi-channel non-invasive diagnostic data acquisition prototype development was completed. The device has modular structure with five modules: light source control module, fibre optical sensor data acquisition module, phonocardio data acquisition module, ECG module, accelerometer sensors and photopletismographic sensors module. The preparation of the documentation necessary for submission for approval of the device usage in clinical trials and submission to obtain approval from the Ethical committee is progressing.

The work on the algorithms development of for extraction of hidden diagnostic markers for detecting atrial fibrillation and atrial flutter was continued. The algorithms based on Nonlinear Fourier Transform and Convolutional Neural Networks were developed and tested.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

With these features the developed device and method will have potential to become a screening diagnostic methods available in the primary health care units covering not only diagnoses achievable by the conventional ECG and PHONO methods but also early detection of heart failure (HF), LVH (Left Ventricular hypertrophy), valvular diseases that are not detectable by conventional screening methods. Introduction of this device into the primary health care system will produce extremely high impact saving millions lives and significantly reducing HF treatment costs.