Answers are grounded in this paper and cite their sources.
Expert-level radiologists had obviously large variances in performance for differentiating benign from malignant solid renal masses. Radiologic-radiomic ML can be a potential way to improve interreader concordance and performance.
No takes yet. Share an insight, caveat, or question.
Sun et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: