Key takeaways
- Chest X-ray AI can flag findings and help prioritise worklists; WHO recommends computer-aided detection for tuberculosis screening in people aged 15 and over.
- Lung CT AI supports detection and measurement of nodules, which matters for screening programmes.
- Performance depends on the product, the indication and the local population, so local testing is important.
- AI supports the radiologist; reporting responsibility stays with the clinician.
Why radiology is a leading area for AI
Imaging produces large volumes of standardised digital data, and demand is growing faster than the number of radiologists in many health systems. Tools that highlight findings or sort the worklist can help urgent studies reach a radiologist sooner.
Chest X-ray AI
The chest radiograph is the most common imaging test. AI products in this area are typically designed to detect findings such as nodules, consolidation, pleural effusion or pneumothorax, and to flag studies for priority reading. Some are designed for tuberculosis screening: since 2021, the World Health Organization has recommended computer-aided detection as an option to interpret chest radiographs for TB screening and triage in people aged 15 years and older.
Lung CT AI
Low-dose CT screening for lung cancer reduced lung-cancer mortality in large trials, including the US National Lung Screening Trial and the Dutch-Belgian NELSON trial. Screening generates many scans with small nodules that must be found, measured and followed over time. AI tools support nodule detection, volume measurement and comparison with prior scans, and can help apply structured reporting categories.
What to check before adopting a tool
- Indication and regulatory status: exactly what the product is cleared or certified to do, and in which countries.
- Evidence: published validation, ideally on populations and equipment similar to yours.
- Local testing: a pilot or retrospective test on your own data before clinical use.
- Integration: how results reach PACS, RIS and the reporting workflow.
- Monitoring: how performance is tracked after go-live, since changes in equipment or patient mix can affect results.
- Data governance: where images are processed and stored, and compliance with local data rules.
Keeping the clinician in charge
Most radiology AI tools are decision support. They are most useful when radiologists understand what the tool is designed to detect, where it performs less well and how its output appears in their workflow.
Related: Medical AI solutions
Talk to our team
We can help you choose the right solution for your facility.
References
- World Health Organization. WHO consolidated guidelines on tuberculosis. Module 2: screening. Systematic screening for tuberculosis disease. Geneva: WHO; 2021.
- National Lung Screening Trial Research Team; Aberle DR, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365:395–409.
- de Koning HJ, van der Aalst CM, de Jong PA, et al. Reduced lung-cancer mortality with volume CT screening in a randomized trial. N Engl J Med. 2020;382:503–513.
This article summarises published guidance and research for general information and is not medical or engineering advice. Device selection and use should follow the manufacturer's instructions, local regulations and your facility's policies.