Douwe, Matthew, and Xinyi assessed the quality of AI research in Soft-tissue and bone tumours (STBT) using the Checklist for AI in Medical Imaging (CLAIM) and FUTURE-AI guidelines, which are designed to promote responsible AI development and clinical translation, and their considerable contribution were outlined in the paper, “AI in radiological imaging of soft-tissue and bone tumours: a systematic review evaluating against CLAIM and FUTURE-AI guidelines”.
In this work, they (Douwe with the co-authors, Matthew Marzetti (shared first co-author) and in particular, Xinyi (shared first co-author) , Wiro, Stefan, and Martijn from the BIGR) present their systematic review on AI in radiological imaging of STBT. This paper includes the following key insights:
- While AI shows promise in diagnosing and predicting outcomes for these rare and challenging tumours, our findings indicate that most studies remain at the proof-of-concept stage, with significant room for improvement in areas like study design, reproducibility, and real-world evaluation.
- Identified gaps in explainability, clinical integration, and bias evaluation.
- Provided recommendations to enhance future AI research for better clinical applicability.
This work has now been published in eBioMedicine – The Lancet Discovery Science.
Douwe said “Looking forward to seeing how these insights contribute to advancing AI research and clinical practice in STBT. Let me know your thoughts or if you’d like to discuss further!”
Read a related post here and the full open-access article here.