Each year the PhysioNet in cooperation with computers in cardiology hosts Challenges and invites participants to address significant unsolved problems in clinical and scientific filed of research [1]. Vast number of teams all over the world are attracted by the subject of challenge and compete in an intense and friendly environment. The main task in this challenge is coming up with automated algorithms embedding the expertise in signal processing, machine learning and clinical knowledge.

The PhysioNet funded since 2007 with the sponsorship of National Center for Research Resources (NIH), proposes a collection of large databases of physiological signals and open-source softwares providing algorithms for analysing the bio-signals.

After the closure of challenge, the participants and scientific community have the opportunity to share ideas and discuss in annual Computers in Cardiology (CINC) conference. This meeting, gathering researchers from diverse nations with distinctive expertise opens the room for lively professional discussions. What is significant is that diverse approaches and methodologies for solving a common problem are exposed in the specific scientific sessions of the challenge. Having the chance to have efficient debates triggers strong collaborations for further finding.

This year’s challenge focuses on classifying sleep arousals using physiological data gathered from 1985 subjects recorded in a sleep laboratory at the Massachusetts General Hospital. In clinical practice detecting arousals (defined as sudden sleep state transitions to lighter sleep stages or wakefulness) from long lasting PSG recordings is a tough and time-consuming task for sleep experts and medical doctors. Hence, automatic computerised algorithms have been powerful to facilitate medical technologists for this purpose.

The physiological signals during sleep are acquired using a device called polysomnography (PSG). Using this embedded multi-channel data acquisition system, various necessary and desired signals of patients are recorded during the sleep. The dataset provided by the challenge includes channels from Electroencephalogram (EEG), Electrocardiogram (ECG), Electromyogram (EMG), Electrooculogram (EOG) and oxygen saturation (SaO2).

Connection between Challenge and AffecTech

A growing body of research has proven that there is a strong and complex interplay between sleep and emotion regulation. The quality of sleep is closely associated with the health related measures of quality of life, physical or mental health and safety [2]. Negative emotions can arise as a consequence of inadequate or bad quality sleep [3]. Therefore, investigating the sleep architecture provide us a great insight of dynamic relationship between sleep and emotional functioning.This fact indicates the strong link between this year’s challenge and the main aim of the AffecTech project.

This year, early stage researchers of the AffecTech team, Andrea Patane and Shadi Ghiasi, having correlated and diverse expertise in research competed in this challenge. The previous experience of Shadi in the last PhysioNet challenges having an expertise in biomedical signal processing embedded with the knowledge in pattern recognition and machine learning from Andrea resulted in a good accomplishment in the first phase of challenge (ranked 4th).

This challenge is ongoing and there are still prospective plans to achieve. They both believe that competing in such a global environment is a very unique, rewarding and fun experience. It provides a good opportunity to learn and employ the knowledge and the experience gained from it in the framework of AffecTech project to meet the main goal of improving mental health.

References

1) https://physionet.org/challenge/2018/
2) Palmer, C.A. and Alfano, C.A., 2017. Sleep and emotion regulation: an organizing, integrative review. Sleep medicine reviews, 31, pp.6-16.
3) Lee, M., Choh, A.C., Demerath, E.W., Knutson, K.L., Duren, D.L., Sherwood, R.J., Sun, S.S., Chumlea, W.C., Towne, B., Siervogel, R.M. and Czerwinski, S.A., 2009. Sleep disturbance in relation to health-related quality of life in adults: the Fels Longitudinal Study. JNHA-The Journal of Nutrition, Health and Aging, 13(6), pp.576-583.

AffecTech

AffecTech is an international research network advancing personal health technologies to help people with affective health disorders (depression, anxiety and bipolar disorder). AffecTech is a project funded by the European Union under the Horizon 2020 Marie Skłodowska-Curie Innovative Training Networks. AffecTech has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 722022.