*3.4. Subjective Evaluation by the Listening Test Method*

Various emotions can be induced in humans by various auditory and/or visual stimuli. A set of six basic emotions is conventionally used by psychologists, although these discrete emotions can be projected into a two-dimensional plane with continuous values of pleasure and arousal axes [18]. The pleasure can change from negative (anger, sadness, fear, disgust) through neutral to positive (surprise, joy). The arousal can acquire values from low (passive emotions like apathetic sleepiness or boredom) to high (active emotion like frantic excitement) [19].

Our subjective method of evaluation based on the listening test was focused on analysis of influence of vibration and noise generated by the gradient system of the MRI device on the human psyche. The aim of the test was to assess how listeners perceive different noises from emotional point of view. The server realization of the listening test "Evaluation of pleasure and arousal of sound" was operated at the internet site http://www.lef.um.savba.sk/Scripts/itstposl2.dll. This automatic application runs on the server PC as an MS ISAPI/NSAPI DLL script and communicates with the user within the framework of the HTTP protocol by means of the HTML pages—see an example of a screenshot in Figure 11. This system, designed originally for assessment of speech signal quality [20], serves at present the purpose of the MRI noise and vibration evaluation. Twenty-seven evaluators (8 women and 19 men) within the age range of 20 to 57 years participated in the listening test experiment open from 26 February to 20 March 2019. For compatibility with other emotional databases, like the International Affective Digitized Sounds [21], the listeners were instructed to use two parameters: Pleasantness (pleasure) and intensity (arousal), within the evaluation range of 1 to 9. The entire test consisted of 10 sets with 4 sounds each, so 40 different noises were evaluated. The listeners were allowed to play the audio stimuli as many times as they wished; low acoustic noise conditions and headphones were advised. The obtained evaluation results of the recorded MRI noises were presented by bar-graph comparisons. Figure 12 shows the emotional influence of the SE sequence with different slice orientations and the GE sequence with different TR times. Figure 13 gives the impact of the SE sequence for different masses in the scanning area, and finally, the effect of different scan sequences.

**Figure 11.** Screen shot of the user communication HTML page used to perform the listening test evaluation.

**Figure 12.** Bar-graph comparisons of evaluated pleasure and arousal parameters for MRI noises of (**a**) the Hi-Res SE sequence with different slice orientations, and (**b**) the Hi-Res GE sequence with different TR times.

**Figure 13.** Bar-graph comparisons of evaluated pleasure and arousal parameters for MRI noises of the Hi-Res SE sequence for (**a**) different objects placed in the MRI scanning area and (**b**) different sequence types {Hi-Res SE-HE, Hi-Res SE-HF, Hi-Res GE-T2, SS-3Dbal, 3D-CE} using a water phantom; sagittal slice orientation in all cases.
