| Title |
|---|
| What is the role of human decisions in a world of artificial intelligence: an economic evaluation of human-AI collaboration in diabetic retinopathy screening |
| Authors |
|---|
| Yueye WANG, Xiaotong HAN, Mingguang HE, Lei ZHANG |
| Presenting |
|---|
| Yueye WANG |
| PURPOSE: |
|---|
| Artificial intelligence (AI) is emerging in the medical field. While emphasis is growing on AI's potential to replace human decision-making, there is limited exploration into the collaborative role of humans with AI and how this collaboration can optimise population health outcomes. |
| METHODS: |
|---|
| This study investigates optimal strategies for human-AI collaboration in primary care decision-making from a health-economic perspective in the context of diabetic retinopathy screening in China. We analyze eight AI-based screening strategies in a total of 270 human-AI collaboration scenarios in a case study of a real-world diabetic retinopathy screening programme. |
| RESULTS: |
|---|
| We find that annual ‘copilot human-AI’ screening in all ages, where both AI and human expertise are leveraged, is the most cost-effective strategy. Implementing the 'copilot' strategy would save US$1.32 million while gaining 14.73 QALYs over a life expectancy, bringing health benefits equivalent to a net monetary benefit of US$1.88 million per 100,000 population compared to the status quo. This strategy effectively combines the advantages of both AI and human graders, maximizing the detection of positive cases and enabling timely intervention. |
| CONCLUSIONS: |
|---|
| Our findings demonstrate the indispensable role of human involvement in maximising clinical benefits and cost-effectiveness in AI-driven medical applications. This study highlights the need to optimize human-AI collaboration strategies for the best AI implementation into healthcare systems. |