L'IA può diagnosticare l'endometriosi dalle irregolarità del ciclo mestruale rilevate nei dati delle app di tracciamento del ciclo ?

L'endometriosi altera i cicli ormonali, causando spesso irregolarità nei pattern di sanguinamento. Un'app che analizza i sintomi registrati tramite intelligenza artificiale potrebbe identificare cicli atipici collegati alla malattia. Una diagnosi precoce potrebbe ridurre i ritardi, che attualmente si attestano in media tra i 7 e i 10 anni. La qualità dei dati e i bias nei resoconti degli utenti rimangono ostacoli chiave. L'approccio sfrutta pattern sanitari crowdsourced su larga scala.

Background

Endometriosis frequently disrupts menstrual cycles, producing erratic bleeding and symptom records that may differ from typical patterns. A 2023 study demonstrated that machine-learning models analyzing self-reported app data can achieve moderate accuracy in distinguishing probable endometriosis from control groups, yet they still incur high false-positive rates and lack confirmatory imaging or surgical validation—components considered essential for reliable diagnosis.

Because definitive diagnosis currently requires laparoscopic surgery or MRI, AI output based solely on menstrual irregularities is best treated as a preliminary signal rather than a conclusive verdict. Data quality issues, including user-reporting biases and incomplete logs, further complicate the approach. Present systems remain experimental and are not approved for stand-alone diagnostic use; any app-generated alert should prompt consultation with a qualified healthcare provider for appropriate testing.

— Enriched May 12, 2026 · Source: BMJ