1 SCIENCE AND CITIZEN SCIENCE

Before science became an established and paid profession, most people conducting scientific research were nonprofessionals: Outside their formal occupation, they engaged in scientific investigations based on their intrinsic interest in certain natural phenomena or unanswered questions (Vetter, 2011). Although often referred to as ‘amateurs’ or ‘gentleman scientists’, some of these nonprofessionals were among the most knowledgeable people in their respective fields at that time. Many of them had collected large sets of data, which later proved invaluable to modern research. Such occurrences are commonly found in different branches of natural science, archeology, geology, and astronomy. However, since the professionalization of science in the 19th century (Silvertown, 2009), scientific research has gradually become the job of ‘experts’ or professional scientists, and nonprofessionals have become somewhat sidelined (Kobori et al., 2015).

Nonetheless, people outside the scientific community were still contributing to science but mostly in assisting roles such as data acquisition. Today, such engagement falls under the umbrella of Citizen Science, a longstanding practice that collectively refers to various modes of engagement and participation of the general public in scientific studies (Miller-Rushing et al., 2012). The ‘North American Bird Phenology Program’, organized by ornithologist Wells Woodbridge Cooke in 1883, is the earliest known citizen science project. The ‘Christmas Bird Count’ project, proposed by US ornithologist Frank Chapman in 1900, is the oldest citizen science project that is still organized every year by the National Audubon Society in the United States. Since then, both the number of citizen science projects and the involvement of amateur scientists have been increasing significantly (Kobori et al., 2015; Silvertown, 2009). Moreover, citizen scientists have moved beyond solely contributing to data collection and have become more involved in collaborative or independent approaches. Nowadays, they initiate ideas, create research questions, develop research protocols and methodologies, collect and analyse data, interpret findings, and share the results. Citizen science is also evolving in new domains, such as the medical and health domains (Den Broeder et al., 2018; Riggare et al., 2019). The popularity of ‘citizen science for health’ and the practice of tracking and exploring one's own health in ‘personal science’ projects have risen, leading to new and unconventional applications of science and technology (Sharon, 2017; Wolf & De Groot, 2020).

In the following, we aim to introduce the emerging field of citizen neuroscience, highlighting some exemplary projects and communities that are engaging in it. Focusing in particular on the field of sleep citizen neuroscience, we will provide a set of successful projects as well as a comprehensive overview of the open-source software and hardware solutions that allow to engage in citizen neuroscience projects. These solutions will enable the participation of individuals as well as large-scale collaborations between professional and citizen neuroscientists. We close with some reflections on the chances and challenges the field of citizen neuroscience faces.

2 FROM CITIZEN SCIENCE TO CITIZEN NEUROSCIENCE

Citizen science has a long history in different fields of the natural sciences. More recently, ‘neuroscientists’ also started enlisting the help of citizen scientists in large-scale scientific projects, initially only for data collection but increasingly also for data analysis and pattern recognition, paving the way for ‘citizen neuroscience’ (Roskams & Popovic, 2016). One of the driving factors of this development is the ‘Quantified Self’ (QS) movement, where individuals without formal training in medicine, neuroscience, or behavioural sciences are longitudinally logging and analysing physiological as well as behavioural and environmental data, usually as part of an N = 1 self-study (Swan, 2013; Wolf & De Groot, 2020). One of the authors of this paper (RtH) falls into this category of citizen neuroscientists: Trained as a bioinformatician studying the immune system and with no prior knowledge of neuroscience, he started a personal longitudinal N = 1 citizen neuroscience project in 2017 with very dense and deep phenotyping, dubbed The Quantified Scientist (Sikder et al., 2022; Ter Horst & Dresler, 2022). This ongoing project aims to study individual variations in physiology and psychology over several years using both consumer electronics and scientific equipment. Since 2018, these measurements included weekly magnetic resonance imaging (MRI) brain scans, daily sleep electroencephalography (EEG) measurements, and, until 2020, weekly polysomnography (PSG) and microbiome sampling. Additionally, using wearables, electronics, and questionnaires, different parameters are monitored on a continuous or daily basis. His simultaneous PSG and wearable EEG recordings played an essential role in the large-scale scientific validation study of the EEG wearable, accounting for about a third of the overall analysed data (Esfahani, Weber, et al., 2023). The contribution further paved the way for the early-stage development of the online and offline automatic sleep staging methods used in the open-source dream engineering toolbox (Dreamento; Esfahani, Daraie, et al., 2023), which enabled the largest multicentre lucid dream induction study with physiological data to date (Esfahani et al., 2022). His motivation for the project is to help advance scientific understanding, learn more about himself, and potentially optimize different aspects of his life. From personal experience, he can report that, although he never got any formal training in neuroscience, he gained a lot of knowledge about the fundamentals of sleep physiology and other neuroscientific topics and methods during the project.

More generally, the study of sleep is a field that has received particular attention in recent years from both producers of wearable neurotechnologies for everyday home settings and their users. Everybody spends a considerable proportion of life asleep, and sleep has a direct, noticeable influence on one's perceived quality of life, which makes sleep and related phenomena such as dreaming popular among many citizen neuroscientists. Citizen neuroscience, when focused specifically on sleep, holds potential advantages over conventional studies due to several reasons. First, the last two decades have seen a dramatic increase in the diversity of affordable commercial health tracker technologies, ranging from smartwatches with sleep monitoring features to a growing market of wearable systems for neurophysiological sleep recording. The availability of these technologies could facilitate large-scale citizen neuroscience, overcoming the common challenge of the majority of existing studies: small sample size. Second, sleep self-experiments are relatively easy to set up for anyone, regardless of their skills or education level, and the results can be shared with or compared with other individuals and general sleep norms. Third, longitudinal daily sleep recordings, which pose challenges for conventional sleep study participation, can become feasible through citizen neuroscience engagement. Of note, there are also some cautionary points that should be considered while performing sleep citizen neuroscience studies. For instance, one should establish a protocol to define an ‘acceptable’ data standard for quality control, which will be followed to review the collected data in an unbiased manner (Roskams & Popovic, 2016). Ethical guidelines should be followed carefully (Naufel & Klein, 2020; Oberle et al., 2019; Remmers et al., 2023).

Beyond investigating one's sleep individually, groups of people have also formed that try to monitor and improve their sleep and dreams collaboratively. Such interest groups range from loose internet communities to more organized crowds of citizen neuroscientists. Examples of loosely connected citizen science groups include web communities in English 1 and German, 2 in which lay enthusiasts as well as professional researchers exchange ideas, methods, and knowledge on the phenomenon of lucid dreaming. There are also citizen science groups that meet regularly in real life to discuss scientific topics or even conduct their own scientific studies: the ‘Student Initiative Sleep and Dream’ at Osnabrück University (Germany), for example, conducts sleep experiments without any professional supervision in their own polysomnographic sleep laboratory and publishes some of their results in scientific journals (e.g., Appel et al., 2020). Another group of sleep and dream citizen scientists formed an independent nonprofit research institute: the Institute of Sleep and Dream Technologies. 3 Their goal is to “invent technologies for better sleep and better dreams”—from ideation of new sleep concepts and prototyping to conducting sleep research and putting the results into real-world applications. Some of the citizen scientists at this institute have a background in professional sleep research; however, they are not actively pursuing the topic within an academic setting anymore, whereas others bring in skills and ideas from other technological fields entirely outside academia, illustrating how classical science and citizen science can interact and complement each other.

In addition to pure monitoring of brain function, emerging biohacking/neurohacking communities aim to explore the technological enhancement of their brains, bodies, and behaviour. These perceived enhancements involve various modifications, ranging from the gain of additional sensing or control functions by implementing magnets or radio frequency identification (RFID) tag in the arm to creating a neural interface by surgically inserting a multielectrode array into the arm to ‘telepathically’ control a robotic arm (Yetisen, 2018). Less invasive neurotechnologies are increasingly being used not only for passive monitoring but also for active optimization of daily behavioural performance brain function (Dresler et al., 2019). As part of the above mentioned The Quantified Scientist project, RtH tests consumer devices for their accuracy and reports the results on a YouTube channel. 4 A significant part of the >200k subscribers of this channel are generally interested in tracking and improving different parts of their lives using wearable and neural devices, and some of these could thus be considered citizen neuroscientists. Interestingly, a survey among subscribers revealed that all individuals that used neurotechnological devices themselves considered the topic of self-enhancement as a significant matter (Ter Horst & Dresler, 2022), confirming that many citizen neuroscientists aim to modulate and improve aspects of their lives using neurotechnology.

3 OPEN SOFTWARE FOR THE CITIZEN NEUROSCIENCE OF SLEEP

Open software is often preferable in research, and when it comes to citizen science, its importance is even greater. Benefits of open-source software for citizen neuroscience include the adaptability to the research goals, no additional cost for scaling due to licensing, and the possibility to involve the citizens in the development and improvement of the software.

Sleep staging is typically the first step for sleep data analyses. Given the large-scale data produced in citizen neuroscience studies, employing valid automatic sleep scoring methods should be prioritized over human scoring of the data. While several autoscoring algorithms have been introduced in the literature, the majority require a specific EEG montage (e.g., Hsu et al., 2013; Supratak et al., 2017; Supratak & Guo, 2020; Tsinalis et al., 2016), and the ones compatible with different EEG montages (Perslev et al., 2021; Vallat & Walker, 2021) may not be used ‘out of the box’ for wearables. This is because wearables typically employ different electronics and nonconventional EEG derivations (Esfahani, Weber, et al., 2023; Esfahani, Daraie, et al., 2023). Therefore, it is essential to further develop autoscoring models exclusively for wearable systems data or to retrain some of the robust existing models with wearable data through transfer-learning.

Table 2 summarizes open-source software that is developed for sleep EEG analysis and may be employed by citizen neuroscientists. This list encompasses the tools including graphical user interfaces (GUI) that can be used for (1) real-time data monitoring, recording, post-processing, analysis, and stimulation such as COsleep (RRID:SCR_017053) and Dreamento (Esfahani, Daraie, et al., 2023), (2) data visualization, post-processing, and analysis including FASST (Schrouff et al., 2009), Luna and moonlight, 6 Sleep (Combrisson et al., 2017), SleepTrip (RRID:SCR_017318), SpiSOP, 7 Visbrain (Combrisson et al., 2019), SmartHypnos (Chriskos et al., 2019), Wonambi, 8 and YASA (Vallat & Walker, 2021), and (3) tools that have been explicitly developed for sleep events detection (e.g., spindles or K-Complexes) comprising Spinky (Lajnef et al., 2017), SWA (Mensen et al., 2016), and Tensorpac (Combrisson et al., 2020).

Similar to hardware, the software selection for citizen neuroscience projects should align with the study objectives. By listing the required features to achieve the study objectives, one may refer to the details outlined in Table 2 to make the most appropriate choice.

In the following, we further elaborate on two examples, selected based on their comprehensive features ranging from data monitoring and recording to potential stimulation capabilities and post-processing of the results. The first example, COsleep, can be used for citizen neuroscience projects employing complex EEG systems such as OpenBCI (see Table 1) and is configurable as a full-PSG with open standards, whereas the second example (Dreamento) is developed to support minimal sensing systems such as EEG wearables.

3.1 Open software for open hardware: COsleep

We developed COsleep 9 (RRID:SCR_017053, Figure 1a): an open-source python software suite to record full-PSG and drive stimulation using auditory closed- and open-loop protocols with relatively inexpensive open-source OpenBCI hardware (see Table 1; all Cyton boards v1 or later, 125 Hz or 250 Hz, 8 or 16 channels). COsleep is flexible to the experimental setup and helps to set up an OpenBCI Cyton board automatically to record and display PSG-signals according to a user-defined channel montage. A variety of setups can be chosen in COsleep for (1) stimulation studies, (2) real-time recording without stimulation, or (3) just recording on the micro-SD card of the OpenBCI Cyton board without real-time monitoring/stimulation.

COsleep has tools in helping to (a) create, mix, and manage stimulation sounds, (b) accurately determine the hearing threshold, and (c) obfuscate the stimulation/sham condition for blinded experimentation. COsleep comes with slow-wave/sleep spindle detection and targeting algorithms to stimulate at slow-wave downstates and upstates or sleep spindles. These features allow COsleep to reproduce most published auditory stimulation research in sleep like targeted memory reactivation (TMR; Hu et al., 2020; Rudoy et al., 2009), or slow-wave/sleep spindle triggered stimulation (e.g., Ngo et al., 2013; Wang et al., 2022) with relatively higher sample size. In summary, as the importance of transferring appropriate technologies to investigate the clinical applications of auditory stimulation during deep sleep has been recently emphasized (e.g., Esfahani, Farboud, et al., 2023), COsleep enables citizen scientists to conduct full-PSG experiments in a naturalistic home environment, allowing to gain more insights into the topic.

4 OPEN SOFTWARE FOR EEG WEARABLES: DREAMENTO

Objective modulation of sleep is an intriguing topic for citizen neuroscience, especially when it comes to the topics that most citizens are engaged in, for example, sleep staging, sleep quality assessment, and (lucid) dreaming. Recently, we developed an open-source dream engineering toolbox (currently only) for the ZMax EEG headband, dubbed Dreamento 10 (Figure 1b), capable of recording, monitoring, modulating (using sensory stimuli such as visual, auditory, and tactile), and analysing (e.g., automatic sleep scoring, power spectrum, and time-frequency representation) sleep in real-time as well as postprocessing the resulting data in an interactive GUI (Esfahani, Daraie, et al., 2023). These features give Dreamento the capability to not only be a practical tool for data collection but also be a user-friendly solution to analyse sleep data by citizen neuroscientists.

While the performance of Dreamento as an all-in-one package for (lucid) dream engineering is currently being validated in a multicentre study (Esfahani et al., 2022), it has potential applications for the upcoming citizen neuroscience topics. As a case in point, citizen neuroscientists may use Dreamento to present light, audio, and vibration stimuli during their REM sleep as a valid method to induce lucid dreams (Stumbrys et al., 2012) or use the text-to-speech feature of the toolbox to establish a medium for communication while dreaming (Konkoly et al., 2021).

5 CITIZEN NEUROSCIENCE: OUTLOOK

The advances in wearable neurotechnologies have likely facilitated the increasing number of citizen neuroscience projects. Still, citizens need a certain level of expertise to contribute to or conduct citizen neuroscience projects. They may experience difficulties in understanding the project's goals, following data collection protocols or complex procedures to integrate data, or adhere to privacy and ethical guidelines. Therefore, some studies have questioned the quality and reliability of data collected by citizen scientists and scrutinized their role in scientific research (Aceves-Bueno et al., 2017; Meschini et al., 2021; Resnik, Elliott, & Miller, 2015; Riesch & Potter, 2014). For example, citizen science projects are suggested to be at higher risk of data fabrication, result falsification, tempering, participation bias, plagiarism, or partial data presentation or analysis (Gilbert & Denison, 2003; Haklay, 2016). Furthermore, the involvement between the citizen neuroscientists and nonprofit, private, or political organizations (or personal relationships with the study subjects) may raise a conflict of interest (COI) in the study (Resnik, 2007; Resnik, Elliott, & Miller, 2015). In addition, possibilities for exploitation were mentioned, which occurs due to lack of consent, inequity, mistreatment, or if people are harmed somehow (Guerrini et al., 2018; Riesch & Potter, 2014). Although many countries follow well-established policies to avoid issues related to research misconduct (Resnik, Rasmussen, & Kissling, 2015), they have not yet been appropriately transited into citizen science projects. To overcome these challenges, we suggest that the active involvement of professional researchers can mitigate the issues by providing citizen neuroscientists with the essential training, tools and resources, and assistance for practical works such as data collection, analysis, and integration. Citizen neuroscientists can be informed about the rules and regulations of data sharing, transparency, and ethical research transparently to ensure that these guidelines are strictly followed in every stage of the project (Remmers et al., 2023). Furthermore, the researchers can equip citizen scientists with the necessary devices for data collection and infrastructure for big data management and sharing. For successful knowledge discovery from citizen neuroscience projects, both groups must work hand-in-hand, co-create, and progress methodically towards achieving the goals of the project (Roskams & Popovic, 2016).

Small sample size, lack of generalizability, confinement to the lab settings rather than naturalistic home environments, and restricted amount of recordings per participant are among the major reasons that contributed to several controversial findings in different fields of neuroscience, such as sleep. Given the potential of citizen science, if the challenges are carefully considered, the contributions of citizen neuroscientists are indispensable to advancing the field. Along with the proper tools and skillset to achieve best research practices, they must possess a certain motivation and interest in doing science to make their participation meaningful and worthwhile. They contribute to projects due to their desire to learn, understand, and improve their own sleep patterns as well as the sleep of others, establish networks, or obtain a good reputation in the (scientific) community (Asingizwe et al., 2020). The collaboration between citizen neuroscientists and professionals can contribute to the motivation and commitment of citizens to a project (Hobbs & White, 2012). Creating a welcoming and collaborative environment and rewarding citizen neuroscientists for their contributions are some crucial things that researchers can do to promote citizen science. Rewards could include personalized sleep reports or detailed dream analyses, certificates for participation, acknowledgments or co-authorship in publications, or even monetary rewards (Cappa et al., 2018; Hunter & Hsu, 2015).

Recently, several research groups started to perform home studies with wearable EEG systems (e.g., Debellemaniere et al., 2018; Esfahani, Weber, et al., 2023; Healthy Brain Study consortium et al., 2021; Lustenberger et al., 2022; Talamini et al., in preparation; Ter Avest et al., 2023; Xi et al., 2023). However, the majority of home studies, including the aforementioned ones, still recruit participants using conventional methods. We believe that it is time for a bold translational effort to make citizen neuroscientific studies more common—topics that meet the interest of the wider public, such as sleep monitoring or enhancement and (lucid) dreaming, seem a particularly well-suited starting point for such an endeavour.

Citizen science projects benefit both researchers and citizen scientists, which is why this practice has been growing worldwide (Vohland et al., 2021). It allows researchers to generate large longitudinal datasets and get direct input from the public. From the citizen scientists' point of view, participating in these projects can have numerous benefits (Den Broeder et al., 2018). From a societal point of view, these projects engage people in scientific studies, which have a lot of benefits (Pocock et al., 2019). Such engagement promotes science and technology among populations outside academia and familiarizes them with scientific thinking. By training citizen scientists, these projects help to create the next generation of scientists, researchers, thinkers, leaders, and conservationists, who will be knowledgeable and receptive to science and resistant to disinformation, misconception, and false narratives. All things considered, citizen (neuro)science projects are mutually beneficial and have significance for the long-term prosperity of science.

AUTHOR CONTRIBUTIONS

Mahdad Jafarzadeh Esfahani: Conceptualization; methodology; software; visualization; writing—original draft; writing—review and editing. Niloy Sikder: Conceptualization; writing—original draft; writing—review and editing. Rob ter Horst: Conceptualization; methodology; writing—original draft; writing—review and editing. Amir Hossein Daraie: Conceptualization; methodology; software; writing—original draft; writing—review and editing. Kristoffer Appel: Conceptualization; writing—original draft; writing—review and editing. Frederik D. Weber: Conceptualization; methodology; software; supervision; writing—original draft; writing—review and editing. Kirsten E. Bevelander: Conceptualization; writing—original draft; writing—review and editing. Martin Dresler: Conceptualization; funding acquisition; project administration; supervision; writing—review and editing.