Table 7 Diagnostic and predictive models built using imaging.
From: Clinical use of artificial intelligence in endometriosis: a scoping review
Authors [ref.] | Stage of endometriosis | Type of endometriosis | Sample size | Inputs used | AI methods used | Method accuracy |
|---|---|---|---|---|---|---|
Maicus et al.61 | NR | Endometriosis with POD obliteration | 749 sliding sign transvaginal ultrasound videos | Presence of sliding sign on transvaginal U/S | Resnet (2 + 1)D | SE = 89% SP = 90% |
Guerriero et al.59 | NR | Rectosigmoid endometriosis | 106 patients with U/S diagnosis of rectosigmoid endometriosis | Age; presence of U/S signs of uterine adenomyosis; presence of an endometrioma; adhesions of the ovary to the uterus; presence of “kissing ovaries”; absence of sliding sign | K-nearest Neighbor | SE = 66% SP = 71% |
Guerriero et al.59 | NR | Rectosigmoid endometriosis | 106 patients with U/S diagnosis of rectosigmoid endometriosis | Age; presence of U/S signs of uterine adenomyosis; presence of an endometrioma; adhesions of the ovary to the uterus; presence of “kissing ovaries”; absence of sliding sign | Naive Bayes | SE = 72% SP = 77% |
Guerriero et al.59 | NR | Rectosigmoid endometriosis | 106 patients with U/S diagnosis of rectosigmoid endometriosis | Age; presence of U/S signs of uterine adenomyosis; presence of an endometrioma; adhesions of the ovary to the uterus; presence of “kissing ovaries”; absence of sliding sign | Neural Networks | SE = 72% SP = 73% |
Guerriero et al.59 | NR | Rectosigmoid endometriosis | 106 patients with U/S diagnosis of rectosigmoid endometriosis | Age; presence of U/S signs of uterine adenomyosis; presence of an endometrioma; adhesions of the ovary to the uterus; presence of “kissing ovaries”; absence of sliding sign | Support Vector Machine | SE = 84% SP = 71% |
Guerriero et al.59 | NR | Rectosigmoid endometriosis | 106 patients with U/S diagnosis of rectosigmoid endometriosis | Age; presence of U/S signs of uterine adenomyosis; presence of an endometrioma; adhesions of the ovary to the uterus; presence of “kissing ovaries”; absence of sliding sign | Decision Tree | SE = 66% SP = 77% |
Guerriero et al.59 | NR | Rectosigmoid endometriosis | 106 patients with U/S diagnosis of rectosigmoid endometriosis | Age; presence of U/S signs of uterine adenomyosis; presence of an endometrioma; adhesions of the ovary to the uterus; presence of “kissing ovaries”; absence of sliding sign | Random Forest | SE = 66% SP = 72% |
Guerriero et al.59 | NR | Rectosigmoid endometriosis | 106 patients with U/S diagnosis of rectosigmoid endometriosis | Age; presence of U/S signs of uterine adenomyosis; presence of an endometrioma; adhesions of the ovary to the uterus; presence of “kissing ovaries”; absence of sliding sign | Logistic Regression | SE = 72% SP = 73% |
Reid et al.60 | NR | NR | 189 women (100 training set, 89 test set) with suspected endometriosis | POD 1 model: posterior compartment deep endometriosis, right ovarian fixation, negative “sliding sign”; POD 2 model: unilateral ovarian fixation, unilateral endometrioma, negative “sliding sign” | Logistic Regression | POD 1: SE = 88% SP = 97% POD 2: SE = 88% SP = 99% |
