Summary of two questionnaires designed to understand the research climate for Bioimage Analysts in the UK between 2016-2019

Ethical considerations

Due to the non-sensitive nature of the content, in addition to the lack of identifying information being included in this publication, no ethical approval was sought for this study. None of the data presented here, or in the accompanying Zenodo repository, can be used to identify survey participants, and the data is not accessible through the SurveyMonkey platform used to host the questionnaires (https://www.surveymonkey.com/mp/policy/privacy-policy/). Completion of the questionnaire after reading the information sheet was taken as consent to participate in the survey.

Questionnaire 1

Many researchers who identify as being Bioimage Analysts have developed the necessary expertise for themselves having responded to a need in the community. As such they often come from a diverse range of roles and are classified according to their employers in a variety of different ways. When we asked scientists who perform bioimage analysis, "What is your current job/position title?" we got a very broad distribution of answers.

Q1.1: What is your current job/position title? Out of the 99 people that answered the first questionnaire only a maximum of three people gave the same answer to the above question and very few people gave the answer ‘Image Analyst‘ and no-one gave the answer ‘Bioimage Analyst‘. The above ‘Word Cloud Image‘ (Figure 1) gives an idea of the range of nomenclature for individuals in this discipline. Going by their job title it suggests that bioimage analysis is not the main focus of their work.

Figure 1. A "Word Cloud" representing the answers from Q1.1.

Participants were able to submit one or more words describing their job position/title. Each word was treated independently.

Q1.2: Do you feel that your current job/position title accurately describes the work that you do? In Figure 2 we show the outcome from asking: "Do you feel that your current job/position title accurately describes the work that you do?" To this question 50.5% of individuals said ’YES’ to the above question whereas around 43.4% said ’Not completely’ and 6.1% said ’No’, not at all. This suggests that there is a large component of individuals for whom their job title and description doesn’t quite describe the focus of their work. This doesn’t prove that they should be named ’Image Analyst’ or similar but it does suggest individuals are assuming or taking on roles which are not reflected by their job-title. As we see later, there is evidence to suggest these individuals are mostly performing image analysis.

Figure 2. Pie chart representing answers from Q1.2.

Participants were able to select only one answer.

Q1.3: Which of the following categories best describes your working relationship with image analysis? We posed the question, "Which of the following categories best describes your working relationship with image analysis?". The responses to this question were very interesting (Figure 3). In total, 35.4% saw themselves as ’Image Analysis Users’. These are individuals who treat image analysis as a tool that they use in their research. Although potentially skilled, they do not create or develop analysis methods, algorithms or pipelines. Many of these researchers are likely laboratory bioscientists interested in using image analysis tools for testing biological hypotheses, without an active role in developing algorithms and pipelines. The remaining 64.6% of individuals put themselves in a more specialist role, mostly either as a Bioimage Analyst, Specialist Image Analyst but a few used the term Software Engineers. This essentially means that 64.6% of the individuals who filled out this questionnaire have a professional level of image analysis expertise and are employed in the life sciences. Along with the first two questions, it suggests that these individuals in their roles are not defined according to the actual job they are performing, but have found themselves in these roles as means to do what they want, or what is needed (i.e. image analysis). In our follow-up questionnaire (section 2, Figure 7 and Figure 8) we asked the same question but provided more focused answers and analysis.

Figure 3. Pie chart representing answers from Q1.3.

Participants were able to select only one answer.

Q1.4: Often we perform different roles at different times. how much of your time, on average, do you commit to the following activities? For this question subjects were asked how they organise their day-to-day research between the activities of a Bioimage Analyst, Software Engineer, Image Analysis User and other research activities. We weighted people’s answers depending on whether they stated that they spent all of their time (weight = 8), most of their time (6), half of their time (4), just some (1) or none of their time (0) performing the specified activities. Broadly speaking most individuals spent most of their time performing some kind of image analysis, whilst compacting all other activities into a smaller proportion of their time (Figure 4). This is very interesting as it shows that most of the individuals who filled out this questionnaire are indeed Image Analysts and that other forms of research and experimentation are secondary to their data analysis.

Figure 4. Pie chart representing answers from Q1.4.

Participants were able to select multiple answers.

Q1.5: If you are interested in pursuing image analysis as potential career, do you feel as though there are sufficient options for career progression in your chosen field of academia? Figure 5 shows the distribution to answers from the following question: "If you are interested in pursuing image analysis as potential career, do you feel as though there are sufficient options for career progression in your chosen field of academia?" The distribution of answers to this question is very clear. Of those individuals interested in pursuing image analysis 63.4% of them said that they did not have sufficient career options to develop in this direction. This essentially means that although these individuals are ful-filling an important and high-demand area of bioscience research, there is not a sufficient framework in place to support their specific career development. The previous questions prove that many individuals are performing image analysis as the dominant aspect of their day-to-day research, even if it isn’t their obvious job description.

Figure 5. Pie chart representing answers from Q1.5.

Participants were able to select only one answer.

We reasoned that the issues affecting the bioimage community highlighted in the first questionnaire relate ultimately to the organization of funding in the UK. In our follow-up questionnaire (questionnaire 2), we asked a number of questions relating to the funding of scientists that perform some kind of bioimage analysis in their work. Through this characterization (84 respondents) we have been able to clarify the funding sources which are funding bioimage analysis and are now in a position to make suggestions to the community as a whole.

Questionnaire 2

Q2.3: What best describes your position? (Please choose the closest role). Q2.3 (Figure 6) was designed to find the closest job description which scientists will classify themselves into. Previously we showed the diversity of titles that people will assign to themselves in (Q1.1), but here we ask specifically which title best describes their position. At the moment, the category that is most dominant among those doing bioimage analysis is Facility Manager (31%), followed by post-docs in a research group (19%). This is to be as expected as scientists, from within a facility, will need to assist researchers by providing image analysis expertise, whereas post-docs will be performing analysis to pursue their own research or developing new techniques for the community. If we compare the ratio of facility-based staff (44.1%) to research group staff (41.5%) who are doing some form of bioimage analysis, the ratio is about 50:50. Therefore a simplified view of this is that from those that filled out questionnaire approximately 50% of scientists are on an academic career track and the other 50% can be considered research technical professionals.

Figure 6. Pie chart summarising answers from Q2.3.

What best describes your position? Participants were able to select only one answer.

Q2.4: Which of the following categories best describes your working relationship with image analysis? We asked the participants to describe their relation with image analysis (Figure 7, Figure 8) by checking the categories they most associated with the role. We took the answers and pooled those individuals whom associated mostly with facilities and compared it to those who predominantly associated with research groups as determined from Q2.3. Figure 7 and Figure 8 show that there are some clear differences between the two cohorts.

Figure 7. Answers from Q2.4 pooled amongst facility associated staff (e.g. Facility Manager, Research assistant in a facility, Research Software Engineer in a facility, Post-doc in a facility).

Participants were able to check multiple options.

Figure 8. Answers from Q2.4 pooled amongst research group associated staff (e.g. Principal Investigator, Research Assistant in a research group, Research Software Engineer in a research group, PhD, Post-doc in a Research Group).

Participants were able to check multiple options.

The main differences between the work relating to bioimage analysis between a facility and a research group is that much more training is being performed in the facility, whereas the ’specialist image analyst’ and also the ’Image Analysis Users’ are more abundant in research groups. The proportion of individuals performing training in research groups (5.6%) is low compared to those performing training in a facility (30.6%). Conversely however there are more ‘Image Analysis Users’ in research groups (facilities: 11.8%, research groups: 23.5%) and also more ‘Specialist image analysts’ (facilities: 15.3%, research groups: 23.5%). This shows that research groups are more interested in applying algorithms through software (as an Image Analysis User) and/or developing specialist algorithms and approaches. Interestingly however the amount of bioimage analysis is consistent amongst the groups (facilities: 25.9%, research groups: 23.5%) showing that connecting pipelines of algorithms is a common goal, whether it be to support users or to develop novel pipelines of analysis.

Q2.5: Where does your funding come from? (0–100)% The academic positions (i.e. PhD positions, post-doc, PI positions) attract mainly funding from research grants, fellowships, with some PIs being partially funded by lectureships (Figure 9). Facility managers and research assistant positions are funded predominantly by cost-recovery and core-funding. This is fairly typical in the life sciences where facility and technical staff are mainly funded by core-funding or funding generated within an institute whereas researchers are attracting funding mainly from research grants

Figure 9. Radial plots displaying answers to question 2.5.

Participants were asked to gauge the approximate funding contributions (0–100%) of their position from the following sources: CF (Core Facility), RG (Research Grants), UL (University Lectureship), FG (Fellowship Grant), CR (External and Internal Cost Recovery), OF (Other funding). Center, 0% funding, Outside Circle, 100%, each graduation represents 20%. Values in each category were averaged over each class of participant.

Q2.6: Which UK Research Councils or Charities fund your research? From the answers generated from Q2.6 it is clear the main funders for bioimage analysis at the moment are the MRC, BBSRC¹, EPSRC, Wellcome and CRUK (Figure 10). This result is fairly consistent with the idea that the research grants are indirectly funding bioimage analysis research through conventional funding routes.

Figure 10. Bar chart visualising answers to Q2.6.

Participants were able to check multiple options.

Q2.7: How much longer will your current funding last? (If you have multiple funding sources, choose the one that runs out first). The more academic posts (post-docs, PhDs, PI) tend to have a more narrow funding horizon, whereas the facility roles and research assistant roles can (not always) benefit from longer funding (Figure 11). This is shown by the apparent skew of distributions toward the left-hand side in the categories of Post-doc in research group, PhD in research group and PI. This highlights the known career dynamic that influences research and facility positions. Facility positions often mean your academic career is limited, but can often mean you have access to better job security, a factor which influences many researchers whom take up these posts.

Figure 11. Bar charts representing funding horizon for different job types, answers from Q2.6.

Categories represent: <6m (< 6 months), 1y (1 year), 2y (2 years), 3y (3 years), 4y (4 years), 5+ (5 years and more), UR (Until retirement).

Q2.8: Have you found that your funding/university have included adequate resources for you to develop professionally (i.e. have been allowed and encouraged to apply for your own funding, attend conferences, network)? The answers from Q2.8 suggest that the majority (54.8%) of researchers in 2019 have sufficient money and opportunities to develop professionally (Figure 12). In total, 33.3% believe more opportunities are needed, but that there are some opportunities available for them. This is a fairly good representation of how the existing funding opportunities support career development in the life sciences through conference attendance, networking and opportunities to apply for certain kinds of grants.

Figure 12. Bar chart representing answers to Q2.8.

Data was pooled from across job descriptions for this question. Participants were able to select multiple answers.

Q2.9: Statement: Due to the sometimes mixed and ad-hoc nature of funding sources it is possible that one’s source of funding has been detrimental to your efforts to support your local community as a whole (i.e. you have not been able to assist people with general problems because of commitments to a particular funding source). Question. Do you agree with the above statement? We asked Q2.9 to ascertain whether participants in their current position felt as though they have enough flexibility to address the needs of their institute. There replies were broad to this question with a range of answers representing a unique set of circumstances (Figure 13). In general, most participants felt that they could assist the local community to some level of capacity with only 4.8% saying that they couldn’t address the needs of the local community, due to tight funding restrictions. This shows that once funding has been obtained then there is sufficient flexibility within the funding’s remit to support the needs of the local community.

Figure 13. Bar chart representing answers to Q2.9.

Participants were able to select only one answer. Data was pooled from across job descriptions for this question.

Q2.10: Statement: Individual researchers, in terms of their career, are feeling as though they must choose in a binary sense between pursuing independent research or a service/support role. Question: Would you be interested in funding which financially assisted you to support your local community (i.e. training and assisting in other people’s projects), but also allowed some time to pursue your own independent research interests (similar to a fellowship)? The answer to this question was very clear (Figure 14); a job role whereby you can have your own ongoing research, but also support the community directly, is seen as ideal. Most Bioimage Analysts on an academic track realise that for good uptake of their approaches they need to work directly with users and support their ongoing work. Conversely, if novel techniques are developed these should be allowed to form the genesis of academic progression, even if the work was done in support of another researcher’s research project. This mixed model of research and support would be a good definition for establishing and representing Bioimage Analysts as a discrete entity, a system which allows them to have an identity in both the academic and technical settings. In addition to the set answers, we allowed the participants to also input their own answers. These are included in the Extended data³ and echo the results from the main question.

Figure 14. Pie chart representing the answers to Q2.10.

Participants were able to select only one answer. Data was pooled from across job descriptions for this question.

Underlying data

Zenodo: Underlying data from two questionnaires used to understand the research climate of Bioimage Analysts in the UK between 2016–2019, http://doi.org/10.5281/zenodo.4605416³.

This project contains the following underlying data:

Extended data

Zenodo: Underlying data from two questionnaires used to understand the research climate of Bioimage Analysts in the UK between 2016–2019, http://doi.org/10.5281/zenodo.4605416³.

This project contains the following extended data:

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).