ARTIFICIAL INTELLIGENCE (a)
n Efficiency: automated image analysis can significantly reduce the time spent on routine tasks,11,15,16
which can
enhance efficiency and aid laboratories in mitigating critical industry challenges, including the global shortage of pathologists17,18 growing pathologist burnout.19
and the
n Scalability: AI tools can process large volumes of cases without compromising accuracy.
n Decision support: AI algorithms aim to complement, not replace, the pathologist’s expertise. They serve as a second opinion, providing confidence in borderline cases.
(b)
These benefits are particularly crucial for resource-constrained laboratories and regions with limited access to subspecialty expertise. Yet, the potential of AI in pathology remains underutilised.
Key factors delaying AI adoption Numerous pathology AI tools have demonstrated clinical value and are certified for diagnostic use in several regions.11,20-22
Matthews et al. reported
26 products that have received regulatory approval for placement on the EEA/GB market.23
Yet, widespread adoption in (c)
clinical laboratories remains limited. While the advantages of AI are increasingly well-documented,11-16
several factors
continue to hinder implementation. A critical obstacle is the relatively low rate of full digital pathology adoption, as low as 5% in some countries,24
which constrains
the deployment of AI tools. Digital laboratories also face high integration costs and limited interoperability with existing systems.1,2
Additionally, concerns
about the reliability of deep-learning tools and uncertainties around regulatory and legal compliance contribute to institutional hesitation.25
Although costs vary depending on
Fig 1. Funnel plots showing the observed proportion of invasive breast cancer (IBC) lesions per grade per pathologist (dots) of eight laboratories relative to the mean national proportion for IBC Grades I (a), II (b) and III (c) (2013–2016). *Indicates that the distribution of Grades I–III significantly differed between pathologists within the individual laboratory (calculated by Fishers Exact test; Monte Carlo option). Source: Van Dooijeweert et al.4
prognosis, treatment planning, and eligibility for clinical trials.
The role of AI in standardising diagnostics Deep learning algorithms have demonstrated remarkable proficiency in image analysis tasks requiring high levels of precision and reproducibility. In digital pathology, AI tools can be trained to recognise specific morphological features, providing pathologists with
40
valuable support in analysing digital histopathology samples. These algorithms offer a powerful solution to tackle inter- observer variability challenges. The advantages of AI include:
n Consistency: algorithms apply the same criteria uniformly, minimising subjective interpretation. Multiple studies have demonstrated improvements in consistency in AI- assisted assessments compared to unassisted professionals.11-14
software capabilities and slide volume, Lujan et al. estimated that the initial base price of image analysis software typically starts at $20,000 (approx. £14,700), which can more than double with maintenance fees and server hardware.26 In subscription models, pricing generally ranges from $1,000 to $1,500 per month (£750 - £1,100), alongside a one-time set-up fee determined by the degree of system customisation and interoperability required.26
While these expenses (see Fig 2).
may seem significant, particularly for new solutions, AI providers often take measures to ensure lower costs, such as trials, research collaborations, and early adopter programmes. Regarding the lack of trust in AI tools, this scepticism is explained partly by the
OCTOBER 2025
WWW.PATHOLOGYINPRACTICE.COM
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56
