pictionary-based fmri paradigm to study the neural correlates of spontaneous improvisation and figural creativity - drawing tablet for mac

A novel game-
Love and creativity
A favourable fMRI paradigm was developed to assess the neural association between spontaneous improvisation and fictional creativity in healthy adults.
Participants participated in the word
PictionaryTM guessing game, Sir.
Safe drawing tablet with no clear "creative" instructions.
The main contrast between drawing a given word and drawing a control word (zigzag)
, We observed an increase in participation in Xiaoqiu, Qiu, left parietal cortex, right upper frontal lobe, left frontal lobe, and the area of the collateral/buckle band, so, activation of the debit and left prefrontal cortex negatively affects task performance.
In addition, using the parameter fMRI analysis, the subjective difficulty level of drawing words in the left-hand pre-activation was increased
Frontal cortex, senior expert-
The rated creative content in the drawings is related to the increase in the participation of bilateral small brain screens.
All in all, our data suggests
The small brain interaction behind the implicit processing of psychological representation promotes spontaneous improvisation and fictional creativity. Thirty-
Six healthy adults. 18M, 18F)
The study was initially attended.
Among them, one participant was excluded from the use of prescription anti-depression drugs;
Two participants were excluded due to excessive movement of the scanner, while the data of the three participants were incomplete.
As a result, the final data analysis is limited to 30 adult participants (
16F, average age = 28 years old. 77 years (S. D. =u20095. 54 years)
Average IQ 120 = (S. D. =u200910. 53)).
If participants were able to receive magnetic resonance imaging, they were included in the study (MRI)
The head scan is right-handed.
If the participants themselves are excluded
The current or past history of mental or neurological illness is reported, which leads them to consult medical professionals, or to have metal devices or implants in the head or body that prohibit MRI.
We send flyers by email, message board, list to recruit participants
Server and word of mouth.
Participants were recruited on and around Stanford University campus.
All the experiments were conducted in accordance with the relevant guidelines and regulations of the Stanford University Institutional Review Committee (
Human section)
All experimental programs and procedures have been approved.
Each participant in the study received written informed consent. The word-
Guessing that the fMRI task is a Pictionary-based game and using Matlab ()
And psychological toolbox version 3 ()software.
We used a block.
Block duration with 30 seconds for each of these two cases (word-
Drawing and zigzagdrawing).
In the first condition, word-
The participant is asked to draw a given word (
Mainly action or verb)
Try your best to use MR-
Safe drawing tablet, but it should be noted that others will try to guess the word later by drawing alone.
Control basic motion and visual space aspects during word-
Drawing conditions, participants are also required to make a drawing representing a control word (“zigzag”)
The second condition
Each block is separated by a fixed period, with a random duration in the range of 10-15 seconds (see ).
Each condition has a total of 10 blocks, and the total duration of the task is about 14 blocks. 5u2009minutes.
In each case, participants will see a word at the top
The top left corner of the screen.
Participants are asked to take full advantage of the given 30 seconds in each block and continue to add elements to the illustration in case they want to finish ahead of time.
Words in words-
Drawing conditions are selected from the "action words" pool in the Pictionary game.
In order to balance the difficulty of drawing different words, among the participants, the selected words are rated by a separate group of participants (Nu2009=u200910)
Draw "difficult", "medium" and "easy.
For example, drawing "cry" is rated as easy and drawing "exhaust" is rated as difficult.
In general, we chose 3 difficult words, 4 medium words and 3 easy words.
The order of words is randomly selected and consistent across all participants.
In addition, participants in the fMRI study
Rated difficulty level (
Difficult, medium or easy)
Used to draw each word in a post
Scan the questionnaire. The MR-
The security drawing tablet is designed and developed specifically for this research, using MR-
Compatible touch
Sensitive surface.
It uses KEYTEC 4-
Wire resistors attached to Teensy 2 touch the glass.
0 with custom firmware.
This device is connected via a USB port.
It uses a simple serial protocol to stream absolute locations.
The flat panel case is made of clear acrylic with a laser cutting machine.
The firmware of this tablet is open.
Source, and can be used in the Webster intelligence abbreviation scale-II (WASI-II)
Is used to measure general intelligence. The WASI-
II was designed to be managed separately in about 30 minutes.
The measurement consists of four sub-tests: vocabulary, similarity, block design, and matrix reasoning for obtaining a comprehensive IQ (FSIQ). The WASI-
The average standard of II is divided into 100 points and the standard deviation is 15 points.
Two experts from Stanford Institute of Design (authors A. R. and G. H. )
Be selected to blindly rate each drawing (from the word-
Drawing conditions)
On the scale (a)
Express and (b)creativity.
Every painting is de-
Identification and networking-
An interface-based interface was developed for the scoring staff to facilitate access to all illustrations.
The description of the "representation" scale is as follows: "How easily do you think another person can guess the words represented by the drawing ".
The rating was obtained in five months. point scale (1-5)
1 is "not representative", 2 is "small representative", 3 is "medium representative", 4 is "representative" and 5 is "very representative ".
The "creativity" rating of each painting is evaluated based on the three subscales of fluency, elaboration and originality.
These component tables are selected according to the established standardized testing of graphic ideas.
The scores of these three subscales are averaged to get the final score of creativity.
Each component table is defined as follows :(a)Fluency -
Total number of elements in the figure; (b)Elaboration -
Imagination and elaboration of details; and (c)Originality -
Statistical Frequency and unusual/uniqueness of the response.
The rating of each component table is also performed on the five component tablespoint scale (1-5).
It is important that if any drawing is not clear (e. g.
, Due to the unintentional lines drawn by multiple
Touch on tablet)
, The scoring staff marked these drawings as "confusing" and not included in the analysis.
Overall, less than 4% of the 300 drawings were excluded.
Two scorers were trained in a small sample Drawing (36 drawings)
And their Inter
The reliability index of all drawings (
In-class correlation coefficient (ICC))was 0.
80 representative and 0.
Creative scale 884.
Finally, as mentioned earlier, participants are also self-involved
Rate the difficulty of each drawing (
Difficult, medium or easy)
In terms of difficulty in later drawing
Scan the questionnaire.
Participants are imaging on 3 Tesla scanners (
GE MR750, Milwaukee, WI)
Cognitive and Neurobiological Imaging Center at Stanford University (CNI)using a 32-
Channel radio frequency receiving head coil (
Xinxing Medical Co. , Ltd.
Ma Wilmington).
By filling the foam between the temple and the inner surface of the receiver coil to stabilize the participant's head to minimize movement during scanning, a plethysmograph is placed on one finger in the left hand to monitor the peripheral pulse.
To limit the extra movement of the hand, the mat is placed under the tablet and under the arm of the participant.
A total of 435-
Collected brain volume on 42 axesOblique slice (2. 9u2009mm thick)
Provisions parallel to the board (AC-PC)
Line, use T2 *-
Weighted gradient echo pulse sequence sensitive to oxygen leveldependence (BOLD)
Comparison with the following acquisition parameters: Echo time (TE)
= u2009 30 u2009 ms repeat cycle (TR)
= 2000 ms, flip angle = 77 °, fov = 23.
2 cm, acquisition matrix = 80 u2009 × 80, approximate body size = 2. 9u2009×u20092. 9u2009×u20092. 9u2009mm.
To reduce blur and signal loss due to field input
Highly automated homogenizer
Before obtaining a functional MRI scan, a sequential filling method based on gradient echo capture was used. A high-resolution T1-weighted three-
The two-dimensional BRAVO pulse sequence acquisition for co-is obtained.
Register with the following parameters: Echo time (TE)=u20092.
8 u2009 ms, repeat time (TR)=u20097.
2 MS, flip angle = 12 °, fov = 23 ° cm, slice thickness = 0.
9mm, 190 slices of vector surface;
Matrix 256 × 256 =;
The resolution obtained = resolution 0. 9u2009×u20090. 9u2009×u20090. 9u2009mm.
The image reconstructed house is used as a matrix of 256 u2009 × 256 u2009 × 190.
The use of functional magnetic resonance imaging data processing (
Expert analysis tool for functional magnetic resonanceVersion 6.
00, some FSL (
Software library of FMRIB, www. fmrib. ox. ac. uk/fsl).
The following pre-
Statistical processing steps are applied: using MCFLIRT, non-
Remove the brain using BET, use the Gaussian kernel of FWHM 5mm for spatial smoothing, grand-
Filtering by single multiplication factor and high-pass time (Gaussian-weighted least-
Square straight line fitting with sigma = 50. 0s)
, And probability independent component analysis implemented in MELODIC (
Multiple exploratory Linear Decomposition of independent components)Version 3. 10, part of FSL.
After preprocessing, register the function data to the high of each individualresolution T1-
Weighted image then registered to MNI152 standard-
Using FMRIB's linear image registration tool for spatial similarity linear registration (FLIRT).
For each participant, independent components are classified as "artifacts" or non-
Artifact using in-house emi-
Automatic rtifact expression ool (SMART;
Similar to a tool for making EEG data).
SMART classifies each component as an artifact using the following rules :(a)
When the time series of a component is highly correlated (ru2009>0. 4)
Has only the motion profile, but has nothing to do with the task design; or (b)
When a component has most of its energy (>70%)
In the high frequency range.
Once sorted, SMART generates an HTML-based web-
Quality inspection tools, the operator can easily cover the automatic classification of SMART.
Incorporate quality inspection steps to ensure that the classification of components as artifacts is done conservatively; e. g.
, If the time process of the component shows transient correlation with the task design, the components are retained as components that may contain bold signals.
After this quality check, SMART uses the fsl_regfilt utility (
FSL available)
Return the manual fact component to recreate 4-
Data set for generalized linear model analysis.
In addition, the fsl_motion_outliers script is used (
FSL available)
In addition to the six motion parameters, it was eliminated (from MCFLIRT).
Register to high-
Use FLIRT for resolution structures and standard spatial images. Time-
A series of statistical analysis is performed using a film with local self-correlation correction. Group-
The feat of using the analytic hierarchy process (
Expert analysis tool for functional magnetic resonance. Z (
T/F)
The statistical image is threshold using the cluster determined by Z-transform> 2. 3 and a (corrected)
The significance of cluster = cluster 0 P threshold. 05.
Featquery tool (Provided by FSL)
Percentage changes in parameter estimates used to extract function definitions (
Active cluster)
Areas of interest.
MRIcron is used to visualize neural imaging results on structured brain images.
The Talaraich client is used to search for the Brodmann region activated by the peak cluster.
For the modulation analysis of fMRI parameters, in the participants ()
Horizontal, we subtract the mean (
In 10 drawings made by participants)
Expert creative ratings from each of the creative ratings received.
Similarly, we reduced the average of the subjective difficulty level from each of the 10 subjective difficulty levels provided by his/her drawings.
This demeaning procedure (or normalizing)
Completed before running individual level multiple regression to reduce the impact of individual differences in groups-level analysis.