Artificial intelligence assisted tools for prostate cancer diagnosis using whole slide biopsy images

Health Technology Wales
Record ID 32018012876
English
Authors' objectives: This report aims to identify and summarise evidence that addresses the following research question: What is the clinical and cost effectiveness of artificial intelligence (AI)-assisted review of prostate biopsy in identifying prostate cancer?
Authors' results and conclusions: AI was found to improve the sensitivity of diagnosis, without impacting negatively on specificity. There was a slight improvement in inter-observer concordance, indicating pathologists generally agreed on diagnosis and grading both with and without AI, but the AI did help to improve this. There was a reduction in case review time per-slide, per-case and for overall case turnaround time, suggesting AI can result in more efficient biopsy review. There was also a reduction in the ordering of additional tests such as immunohistochemistry, and requests for second review by pathologists. HTW performed a de-novo cost-effectiveness analysis, the results of which showed that using AI-assistance is expected to increase costs by £207 per patient and provide an additional 0.02 quality adjusted life years (QALYs) compared to pathologist review alone. This translates to an incremental cost-effectiveness ratio (ICER) of £13,278 per QALY, which is below the cost-effectiveness threshold of £20,000 per QALY, and so using AI-assistance is deemed a cost-effective strategy. Results of the probabilistic sensitivity analysis suggest that using AI-assistance has a 69% probability of being cost effective. It was indicated that most patients expected to be informed when AI would be used, but this should not be too detailed. Patients expected practitioners to retain ultimate responsibility in any final decisions but were hopeful that the introduction of AI would result in quicker and more accurate diagnoses.
Authors' recommendations: There is insufficient evidence to support the routine adoption of AI-assisted review of prostate biopsies in the detection and diagnosis of prostate cancer. The current evidence is promising and shows the potential for AI technologies to support pathologists in diagnosing prostate cancer. However, there was limited peer-reviewed evidence on the effectiveness of the AI technologies.
Authors' methods: The Evidence Appraisal Report is based on a literature search (strategy available on request) for published clinical and economic evidence on the health technology of interest. It is not a full systematic review but aims to identify the best available evidence on the health technology of interest. Researchers critically evaluate and synthesise this evidence. We include the following clinical evidence in order of priority: systematic reviews; randomised trials; non-randomised trials. We only include evidence for “lower priority” evidence where outcomes are not reported by a “higher priority” source. We also search for economic evaluations or original research that can form the basis of an assessment of costs/cost comparison. We carry out various levels of economic evaluation, according to the evidence that is available to inform this.
Authors' identified further research: Further research is recommended to understand the clinical and cost-effectiveness of the AI systems, as well as their impact on real-world clinical practice. In particular, HTW would support the analysis of data from the ongoing pilot project using an AI system for prostate cancer diagnosis in NHS Wales.
Details
Project Status: Completed
Year Published: 2024
English language abstract: An English language summary is available
Publication Type: Rapid Review
Country: Wales, United Kingdom
MeSH Terms
  • Prostatic Neoplasms
  • Biopsy
  • Pathologists
  • Artificial Intelligence
  • Machine Learning
  • Diagnosis, Computer-Assisted
  • Image Interpretation, Computer-Assisted
  • Cost-Effectiveness Analysis
Keywords
  • Prostate cancer
  • Biopsy
  • Pathology
  • Artificial intelligence
  • Diagnosis
Contact
Organisation Name: Health Technology Wales
Contact Address: Floor 3, 2 Capital Quarter, Tyndall Street, Cardiff, CF10 4BZ
Contact Name: Susan Myles, PhD
Contact Email: healthtechnology@wales.nhs.uk
This is a bibliographic record of a published health technology assessment from a member of INAHTA or other HTA producer. No evaluation of the quality of this assessment has been made for the HTA database.