As a newcomer in this field, I am currently engaged in quality control assessments for bold reports. Within the rating widget, I have the option to select different checkboxes that correspond to specific aspects of the images. I would greatly appreciate your assistance in clarifying the meaning behind each of the QC checks. Additionally, it would be extremely helpful if someone could provide me with examples showcasing excellent brain images as opposed to poor ones. I have encountered difficulties in locating such illustrations.
There was recently a Research Topic project called “Demonstrating Quality Control (QC) Procedures in fMRI”:
This included an open invitation to software developers and researchers across the community to perform QC on FMRI data with whatever tools they wanted, and to describe their results in detail, providing examples of both good and bad (and questionable) data. The setup and goal of the project were to address just such a question, to help all researchers (and esp. newcomers) to see didactic examples of data of all quality, and descriptions of why it might (or might not) be included for further analysis.
As noted in the Project’s overview Editorial, This Research Topic, “Demonstrating quality control (QC) procedures in fMRI1,” focused on promoting quality control descriptions and discussions within the FMRI community. We invited anyone in the field to participate and perform their QC protocol of choice on sets of task-based and resting state FMRI data, describing their steps and criteria in detail. Ten teams participated, utilizing processing and QC methods that are available from a wide variety of software packages. The resulting set of articles represents a didactic resource for the field moving forward, as a reference for teaching and describing QC procedures.
The examined data collection came from real, unaltered, and publicly available datasets from widely used distributions.
There are many articles to browse through and you will likely find valuable comments and suggestions in each. I will note that the one our group contributed to has many images of raw, intermediate and processed datasets, with the varied degrees of suitability for analysis commented on in them:
This may be useful, though it uses different software than your question asked about. This one made use of MRIQC specifically in its processing: