Utility of Tissue-of-Origin Algorithms in Cases Diagnosed as Lung Squamous Cell Carcinoma
Menée à partir de données portant sur 3 958 cas présumés de carcinome épidermoïde pulmonaire, cette étude évalue l'utilité d'une approche, utilisant l'intelligence artificielle et incluant l'évaluation de preuves orthogonales clés, pour distinguer un carcinome épidermoïde pulmonaire primitif de tumeurs métastatiques provenant d'autres tissus
Molecular profiling has enabled substantial advancements in diagnostics and clinical care, including increasingly precise tumor classification and targeted therapies. Nevertheless, there are areas of diagnostic ambiguity that remain challenging, even with all currently available approaches. Furthermore, in our complex medical system, there are numerous steps at which diagnostic discrepancies or errors can arise, ranging from preanalytic (eg, conflicting or limited clinical information) to analytic (eg, subjective interpretations of available data) to postanalytic (eg, miscommunication among treating clinicians).
JAMA Network Open , éditorial en libre accès, 2026