Based on current evidence, UK National Screening Committee does not recommend using Artificial Intelligence

Figure is taken from [Free21a]. It shows some potential roles of AI in screening pathway.

The current review “Use of artificial intelligence for image analysis in breast cancer screening” looked at the evidence on:

  • how good AI is at finding cancers in breast cancer screening
  • what benefits and harms AI has for the women who are screened or for thescreening programme and the health professionals involved.

Based on the current evidence, the UK NSC does not recommend using AI in the NHS breast cancer screening programme. 

[Free21a]

However, this is only a “Rapid Review”, that was published a couple of days ago. It is for open consultation (the committee is inviting comments).
The recommendation is justified as follows:

– the use of AI systems would change the current screening programme therefore it is important to assess how accurate AI is in breast screening clinical practice before changing it

– the performance of AI systems varies in different settings but there are no good quality studies in the UK

– it is unclear how good AI is at finding different types of breast cancer or at finding breast cancers in different groups of women (for example different ethnic groups)

– AI might reduce the workload of staff, the number of cancers missed at screening, and the number of women called back for further tests when they do not have cancer, however, the quality of evidence is very low.

[Free21a]

Thanks to Dagmar Monett, who published this information on LinkedIn [Mone21a].

More:
[Free21a] Freeman, K., et al. Use of artificial intelligence for image analysis in breast cancer screening. UK NSC external review Draft report v5.0 10/05/2021, May 2021.

[Mone21a] https://www.linkedin.com/posts/monettdiaz_use-of-ai-in-breast-cancer-screening-rapid-activity-6800875501889036288-ISy3