Applications of artificial intelligence in chronic kidney disease: a systematic review.
Jalila Andréa Sampaio Bittencourt, Aline Santana Figueredo, Naruna Aritana Costa Melo et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review examined the applications of artificial intelligence (AI) in chronic kidney disease (CKD) diagnosis and management, analyzing 26 studies. The review found that supervised and deep learning algorithms generally exhibited high performance in identifying CKD stages. However, significant methodological weaknesses were noted, particularly in controlling confounding factors and external validation.
Key findings
- Supervised and deep learning algorithms showed high accuracy, sensitivity, and specificity in identifying chronic kidney disease.
- Methodological gaps limit the generalization and incorporation of AI models into clinical practice.
- Twenty-six studies were included in the review, highlighting the potential of AI in CKD management.
Why it matters
Understanding the role of AI in chronic kidney disease can enhance early diagnosis and management strategies, potentially improving patient outcomes. However, clinicians must be aware of the limitations in current methodologies before integrating these tools into practice.
What to keep in mind
The review highlights weaknesses in controlling confounding factors and the need for external validation of AI models.