The present report aimed at providing information about how two independent populations, one consisting mainly of neurologically healthy students (N = 70) and one of elderly stroke survivors (N = 22), perceive the functional MRI procedure. For this, we presented a questionnaire asking for their opinion on the MRI scan directly after the scanning had finished. The results showed that the majority of healthy subjects as well as patients consider MRI scanning as a comfortable procedure, and that virtually all subjects would like to participate in a subsequent scanning. As negatives, mainly the scanning noise, the need to lie still and not move the head, and occasional feelings of dizziness were mentioned.
How convenient is a basic research scan?
Previous studies primarily assessed the tolerance but not the comfort of the MRI procedure. In particular, anxiety and the occurrence of claustrophobia during MRI scanning has been in the focus of previous research, probably because claustrophobia is the most severe problem and typically results in scan abortion [3, 12, 18–23]. This is perfectly reasonable for the clinical setting in which MRI scans serve important diagnostic purposes and for which the potential health benefits outweigh patient discomfort. However, in the basic research setting researchers and ethical review boards typically consider only moderate levels of discomfort as acceptable. Accordingly, we designed our questionnaire to assess the perception of MRI scanning not on a coarse level (scan was possible vs. had to be aborted) but on a more subtle level of perceived comfort.
Our data show that healthy subjects and patients generally found the scanning procedure comfortable. 67% of the healthy group and 63% of the patients rated the procedure positively (i.e. between 1 and 3 on a 1–7 scale (central item is 4)), and only 13% of healthy subjects and 23% of patients rated the procedure negatively (i.e. between 5 and 7). Extreme negative ratings were rare, as only 1 out of 68 healthy controls rated the procedure as 7 (very uncomfortable), and 2 rated it as 6. No patient rated the procedure as 6 or 7. This demonstrates that the large majority of participants is fine with the procedure and that only very few participants consider the MRI procedure to be very uncomfortable.
These findings are in line with the few previous studies investigating the comfort of the MRI procedure. When the rating scales given to the participants in the different studies are transformed to a universal scale ranging from 0 [uncomfortable] – 100 [comfortable] by the formula (100/(number of choices on response scale - 1) × (mean rating - 1)), we observed in the present study a mean rating of 66.2 for the healthy subjects and 71.2 for the patients. Wollman et al.  also asked how comfortable the overall experience of a research scan was and reported a transformed mean rating of 77.5 for a sample of participants older than 72 years. Sparrow et al.  asked patients after a clinical scan for the comfort of the scan. Although the scan involved injection of a contrast agent, the transformed mean rating still was 63.6, i.e. only slightly below the one observed in the present study. Dantendorfer et al.  used only a very broad scale which unfortunately cannot be transformed in a clinical study comparing a 0.5T and a 1.5T scanner. He reported that 80% (1.5T) – 88% (0.5T) of the patients found the MRI procedure easy to tolerate, 18% (1.5T) – 11% (0.5T) found it unpleasant, and only 2.2% (1.5T) – 0.7% (0.5T) found it hardly bearable. Although the rating scales are not directly comparable, we think that the findings of Dantendorfer et al.  are in general agreement with our findings. Taken together, the present finding that participants regard MRI research scans as comfortable is in line with previous studies investigating the research setting in the elderly  as well as studies investigating the clinical setting [2, 8].
A further finding supporting this conclusion regards the wish to participate again in an fMRI study. In the present study 93.4% of the healthy subjects would like to participate in an fMRI study again. In more detail, 98% of those participants who did not consider the session as too long would like to participate again, but only 62.5% of the subjects who considered the session as too long would do so. Thus, too long scanning sessions may considerably reduce the willingness to participate again. However, it should be noted that this conclusion is based on rather small absolute numbers, since only 8 subjects in total considered the scanning session as too long.
Our findings are in agreement with Wollman et al.  who reported that 100% of their (elderly) subjects would undergo the MRI procedure again (note the active wish to participate again in our study, as compared to the more passive agreement that they would do it again in Wollman et al. and MacKenzie et al.). MacKenzie et al.  reported for a clinical context a willingness to return of only 64%, with an additional 24% returning only if absolutely necessary. The reason for the lower rate is not clear, but may be found in longer scanning durations (up to 95 min in MacKenzie et al.), differences in the procedure (e.g. in MacKenzie et al. scans of many different body parts were analyzed), or differences between clinical and research scans (see next section below).
MRI in the clinical and in the basic research setting
Clinical and research scans differ in some important aspects so that the results gained from one may not hold for the other. First, MRI research settings may impose greater discomfort than clinical scans. For instance, MRI research employs different imaging parameters. Critically, the most frequently employed sequence in research, i.e. echoplanar imaging (EPI), is characterized by a very high volume hammering gradient noise. A further difference is that clinical scans usually require the patient to lie passively in the MRI scanner, while the experimental tasks in research scans can be quite demanding (mentally as well as physically as for instance in the case of the patients who were required to press a handle with the arm affected by a stroke). Interestingly, these differences do not seem to affect the perceived comfort, since the present study and Wollman et al.  found comparable ratings of comfort as compared to clinical studies [2, 8].
Furthermore, research scans tend to take longer than clinical scans. While clinical scans take typically less than 30 minutes, research scans frequently take 45 – 75 minutes, with durations up to 120 minutes. It appears plausible to assume that longer scanning duration increases the experienced discomfort, at least beyond some point. However, only few healthy subjects (8 of 66; i.e. 12.1%) considered the scanning session as too long. Notably, this was neither correlated to the actual study duration nor related to the comfort ratings. This suggests that healthy participants seem to be fine with scanning sessions of at least up to 1 h in total.
On the other hand, one may hypothesize that clinical scans are perceived as less comfortable than research scans. For example, anxiety levels in clinical scans are high, with up to 37% of patients reporting moderate to high levels of anxiety [7, 10, 11], up to 6.5% of aborted scans [24, 25], and up to 14.2% of patients needing sedation to tolerate MRI .
These scans are prescribed and usually conducted for diagnostic purposes. Patients may therefore undergo this procedure despite their anxiety. The prospect of the diagnosis may further aggravate anxiety levels in general and make participants more susceptible to the feeling of anxiety in the scanner setting [5, 6, 11]. In contrast to the clinical environment, participation in MRI scanning is voluntary, and one would therefore assume that persons who worry about the scanning experience would simply not volunteer to participate in those studies . In other words, for research-based scanning a self-selection bias automatically leads to a cohort of volunteers who are likely to be comfortable with the scanning procedure. However, the present evidence does not directly support this conclusion. All previous studies investigating the comfort of the MRI procedure showed that participants and patients generally perceive the procedure as comfortable, irrespective of age or the use of contrast agents. Whether there are truly no differences between clinical and research scans or whether larger samples and improved questionnaires are required to unveil them is a question for future research.
Although side effects such as seeing stars or tingling sensations have been describe for MRI, to present knowledge these side effects are not harmful. Accordingly, they are usually not considered in the clinical context. However, they may become relevant in the research setting, as they may constitute a factor of discomfort, for instance because participants may take them as indication of some harmful condition. We assessed such side effects by the question "Do you think you felt something strange caused by the MRI scanner?".
A surprisingly high number of 34% of the healthy subjects indeed noted something strange. However, a closer inspection of the comments revealed that 7 subjects just reported tiredness, which is well explained by the fact of lying still for 30–60 min and either doing nothing or working on a highly repetitive task. Five subjects reported dizziness during the scan, and two further subjects reported feeling slightly discoordinated and disoriented. Comments potentially relating to effects which would occur in the same context without the MRI scanner as well encompass notes such as (at the end of the scan) "seeing stars against the [bright] white screen", or "feeling wobbly when standing up after the end of the MRI scan" (i.e., after at least 40 min of lying motionless). Two subjects reported feeling slightly nauseated after a diffusion tensor imaging (DTI) scan, during which the scanner bed shakes noticeably. One participant reported a feeling of panic after some time, caused by the enclosed space in the MRI bore and the noise of the anatomical MPRAGE scan (however, the scan did not need to be aborted). Other comments, which cannot easily be categorized or explained comprised "heart rate fell into step with vibration of the scanner", "strange – during the last stage I saw 'things' flying about, but I couldn't focus on them", "Also I felt like I could feel something applied in the back of my head when the scan(s) started", "The sensation of magnets 'pulling' my head in various directions. Also my mind was making songs out of the noises", "isolated – like in a bubble, so all other perception strange", "it's hypnotic; felt like the bed was sinking downwards", "I felt a little bit hot (in my head too), but that's probably because I was tense". These latter comments may at least partially be caused by a heightened awareness or self-focused attention.
Because participation was voluntary and participants have been warned about claustrophobia, we expected a lower rate of claustrophobia than the previously reported incidence of about 2% (e.g. [3, 12]). Of the 70 healthy subjects, only one reported a feeling of claustrophobia (without the need to cancel the scan), which is an incidence of 1.4%. However, whether this 1.4% is significantly lower than the previously reported 2% cannot be answered reliably by the present data. Instead, we suggest that for reliable inferences with such low incidence rates much larger samples are required. Thus, future research is needed to test the hypothesis that a self-selection bias in the research setting results in a lower incidence of claustrophobia.
Effect of gender
In healthy subjects, females rated the procedure significantly less comfortable than males. The reason for this difference is not clear, but may be found in the trend that women more often suffer from claustrophobia or panic attacks in the MRI environment [3, 11, 12, 26]; and that they show increased levels of state anxiety before as well as after the MRI scan [6, 8].
We observed the reversed pattern in patients: females found the procedure significantly more comfortable than males. This could be explained by the findings of Wollman et al.  who showed that, in the elderly, males find some aspects of the procedure less comfortable than females (such as lying flat, positioning, and not moving). However, further research is needed to decide whether this is a genuine gender by age interaction or whether these differences are due to some other differences in the imaging procedure.
Effect of Age
Age did not affect any questionnaire item, which suggests that the perceived comfort is not affected by the age in the range of the investigated population (17 – 69 years). This is in line with Wollman et al.  who showed that two populations of elderly participants (means 76 and 92 years, respectively) did not differ significantly in their estimate of the overall comfort.
With respect to claustrophobia, it seems to be unclear whether age has an effect. While the findings of Eshed et al.  who showed that MRI-related claustrophobia is evenly distributed across all age groups between 20 and 80 years is in agreement with our findings, Sarji et al.  mentions non-significant age effects in their study of more than 3000 patients. In the so far largest study, Dewey et al.  found that in particular the age group 40 – 65 years is characterized by a higher incidence of claustrophobia (2.6%), while younger and older patients showed comparably lower rates of claustrophobia (1.3% and 1.5%, respectively). First, one should note that the incidence of claustrophobia is not necessarily related to the experienced comfort of the MRI scanning. Second, in our study age is confounded with the sample, i.e. the healthy subjects were younger (mean 26 years) than the patients (mean 54). Both groups did not only differ in age, but as well in other factors such as study duration which may result in opposite effects on the comfort as expected by the results of Dewey et al. . Therefore, the discrepancy between the present results and Dewey et al.  may be due to the fact that claustrophobia and perceived comfort are largely unrelated, or to other differences in the study design.
Effect of a DTI scan
The inclusion of a DTI scan had no effect. This is interesting, as the DTI sequence is characterized not only by a very different sounding gradient noise, but as well by a considerably shaking scanner bed. Thus, it seems that our findings are somewhat robust to changes in the employed MRI protocol.
Approximately one third of the subjects and patients underwent an MRI scan before. Unfortunately, we do not have the data about the exact proportion, and we do not have the information which particular participants were scanned before, so that we cannot test for the effect of prior scanning on the perceived convenience. In the clinical context, MacKenzie et al.  assessed the previous imaging experience (MRI and computer tomography) and showed that on average previous experience did not affect the state anxiety before the MRI scan (compare also ). However, state anxiety before the MRI scan was lower if the previous experience was pleasant and higher when it was unpleasant. Assuming that in the basic research context participants only return if the experience has not been too unpleasant, one may speculate that these participants will most likely show on average more positive ratings in the present questionnaire as well. Thus, it may be that the perceived convenience is lower if a sample being scanned for the first time would be investigated. However, despite this potential bias we think that the present data are not unrepresentative of the everyday research fMRI scanning. The main reason for this is that most sites test participants repeatedly in different studies, and thus usually have a sample which consists of participants with prior scanning experience (of course, proportions may vary). Thus, we conclude that although the present data cannot disentangle the effect of prior scanning on perceived convenience, they are representative for the typical setting of basic fMRI research.
To our knowledge, no standardized questionnaire exists to assess general comfort during MRI scanning. Accordingly, we used a self-developed questionnaire for this purpose, as has been done by others [2, 4, 8]. However, to assess aspects related to comfort it probably is beneficial to use standardized questionnaires, such as the Claustrophobia Questionnaire (CLQ)  or the Spielberger State-Trait Anxiety Inventory.
An interesting option for future studies would be to use a pre- and a post-questionnaire, especially if it is known whether participants had prior experience with (functional) MRI. With such a procedure it would be possible to disentangle expectations about the MRI procedure from actual experience. In addition, it would be of interest whether different prior experiences (e.g. anatomical MRI of other body parts than head, anatomical MRI of head, functional MRI) result in the same (in)congruence between expectation and experience as assessed by a pre- and post-questionnaire. Such inferences cannot be made in the current study, since we assessed only the actual experience of the MRI scan, but not the prior expectations.
It should be noted that the present results are specific to our setting and MRI procedure and that there are a number of factors which may affect the scanning experience [12, 22, 28]. For instance, different variants of EPI sequences as well as different MRI machines can change the noise levels and characteristics. Furthermore, different scanner models may have different designs regarding depth and diameter of the bore, which may affect the feeling of claustrophobia. In the same way, different headcoils (e.g. narrow or closed ones), wearing goggles, or fixing the arms with straps or vacuum cushions may affect the comfort in general and the feeling of claustrophobia in particular.