Table 6 Diagnostic and predictive models built using mixed variables.
From: Clinical use of artificial intelligence in endometriosis: a scoping review
AI methods used | Authors [ref.] | Stage of endometriosis | Type of endometriosis | Sample size | Inputs used | Evaluation Metric |
|---|---|---|---|---|---|---|
Logistic Regression | Guo et al.58 | All stages of endometriosis and stage 3/4 endometriosis | NR | 1016 infertile patients | for any-stage endometriosis nomogram: BMI, Cycle length, parity, palpable nodularity, endometrioma diagnosed on TVS, tubal pathology; for stage 3–4 endometriosis nomogram: pain, palpable nodularity, endometrioma diagnosed on TVS | SE = NR SP = NR |
Logistic Regression | Chattot et al.57 | Not specified | NR | 119 patients (47 endometriosis with rectosigmoid involvement, 72 endometriosis without rectosigmoid involvement) | Palpation of a posterior nodule on digital examination, UBESS score of 3 on ultrasonography, rectosigmoid involvement in endometriosis infiltration on MRI, presence of blood in the stools during menstruation | SE = NR SP = NR |
Logistic Regression | Nnoaham et al.27 | Stage 3 and 4 endometriosis | NR | 1396 symptomatic women | Ultrasound evidence, menstrual dyschezia, ethnicity, history of benign ovarian cysts | SE = 82.6% SP = 75.8% |
