Is research one of your New Year's resolutions?

I am reflecting on my own plans for 2024, and research is definitely on my ‘resolution list’, is it on yours? Perhaps you are thinking of enrolling in a higher research degree? Or you have an idea for a project you've been wanting to get off the ground? Maybe you are sitting on a database that is just waiting to be transformed into a paper that has the potential to explain, inform or, indeed, change practice? If any of these resonate with you, then I encourage you to begin your next, or indeed first, research project.

Every research project starts with a research question, but the methods used vary depending on that question. The study design and approach will be defined by the question, and it is important to remember that there is no one way to undertake ultrasound research. There are many different study approaches: from audits and educational research to outcomes research.

All research starts with data. In this issue of AJUM, Moore et al.¹ take you through the steps needed to mine data from an imaging archive. Most of us have a digital database of images, measurements and reports, but extracting those data in a consistent way, especially when there are free text fields, is not as straightforward as one would think.

When little is known about a condition, research often takes the form of clinical audit of cases or case series. We have three distinct examples of this in this AJUM issue: Hosokawa et al.² compare the detection of submandibular sialoliths by ultrasound with magnetic resonance imaging (MRI) and computed tomography (CT), while Boman et al.³ present a small case series demonstrating a potential role for ultrasound to detect small vessel inflammation as a pre-curser to rheumatic vasculitis. Lastly, Walsh and Lees⁴ present a clinical audit of over 3000 consecutive routine mid-semester obstetric scans and report the proportion of major and minor abnormalities. There are two important points to note within this study, which speak to the veracity of the data: this is a single operator clinic such that the ultrasound approach was likely consistent over the time period, and they included consecutive patients. Clinical research should be undertaken using consecutive patients, which ensures all eligible patients are recruited. When non-consecutive, or convenience, sampling is used, I often wonder why they were excluded. Because of image quality? Because their results were outliers—not aligned to the researchers' bias?

It is often said that the strongest form of evidence, after the randomised clinical trial, is meta-analysis or systematic review because all the available evidence is presented in a summative and unbiased manner. One such review by Giri et al.⁵ considered the data related to contrast-enhanced ultrasound for differentiation of benign and malignant portal vein thrombosis in patients with hepatocellular carcinoma and concluded that ultrasound could be an alternative modality to CT or MRI. It is a recurring theme of ultrasound research that comparison is made with CT or MRI, as gold standards, and these types of comparisons are very useful for advancing the independent efficacy of ultrasound.

Lastly, there are two scoping reviews in this issue: Deslandes et al.⁶ undertook a scoping review investigating the role of artificial intelligence in sonographer education related to gynaecological ultrasound, while Joyce et al.⁷ looked at the diagnostic accuracy of point-of-care ultrasound performed for acute cholecystitis in the emergency department. Scoping reviews are an excellent way to find out what has already been done in your area of research and can support the novelty, or not, of your own research idea. Once you have a novel research question, you need collaborators. Research collaborators bring different ideas and share the workload and may be local, interstate or across international borders. All you need is a shared passion for ultrasound imaging and a great research question. So the question remains, will you begin or continue on your research pathway in 2024?