Table 4.
Diagnostic and predictive models built using clinical variables and symptoms.
| AI methods used | Authors [ref.] | Stage of endometriosis | Type of endometriosis | Sample size | Inputs used | Method accuracy |
|---|---|---|---|---|---|---|
| Logistic Regression | Bendifallah et al.50 | Not specified | Ovarian, superficial or deep endometriosis | Training set (1126 patients), validation set (100 patients) | Mother/daughter history of endometriosis, history of surgery for endometriosis, age, BMI, dysmenorrhea/VAS of dysmenorrhea, abdominal pain outside menstruation, pain suggesting of sciatica, pain during sexual intercourse, lower back pain outside menstruation, painful defecation, urinary pain during menstruation, right shoulder pain near or during menstruation, blood in the stools during menstruation, blood in urine during menstruation, absenteeism duration in the last 6 months, number of non-hormonal pain treatments used |
SE = 95% SP = 81% |
| Vesale et al.45 | Not specified | Deep endometriosis with colorectal involvement | Training set (789 patients), validation set (333 patients) | Age, type of colorectal management, colpectomy and parametrectomy |
SE = NR SP = NR |
|
| Benoit et al.46 | All four stages of endometriosisa | Not specified | 297 patients who underwent ART after surgery for endometriosis-associated infertility | Age, duration of infertility, number of ICSI-IVF cycles, ovarian reserve, rAFS score |
SE = NR SP = NR |
|
| Lafay Pillet et al.47 | Not specified | Deep endometriosis in patients with ovarian endometrioma | 326 patients (164 with DE lesions associated with endometrioma, 162 patients with no associated DE lesions) | VAS of gastrointestinal symptoms ≥5 or of deep dyspareunia >5, duration of pain greater than 24 months, severe dysmenorrhea (defined as the prescription of the OCP for the treatment of a primary dysmenorrhea or the worsening of a secondary dysmenorrhea), primary or secondary infertility |
SE = 51% SP = 94% |
|
| Ballester et al.48 | Not specified | Deep endometriosis | training set: 94 patients who underwent ICSI-IVF, validation set: 48 consecutive patients | Patient’s age, presence of DIE, AMH serum level >1 ng/ml, number of ICS-IVF cycles |
SE = 66.7% SP = 95.7% |
|
| Chapron et al.49 | Not specified | Posterior deep endometriosis | 134 patients (51 with posterior DE, 83 with other disorders) | Painful defecation during menses, VAS for dyspareunia > or =8, previous surgery for endometriosis, pain other than non-cyclic |
SE = 68.6% SP = 77.1% |
|
| Decision Tree | Bendifallah et al.50 | Not specified | Ovarian, superficial or deep endometriosis | Training set (1126 patients), validation set (100 patients) | See above. |
SE = 91% SP = 66% |
| Wang et al.51 | Not specified | Ovarian endometriomas | 178 case records | Patients’ basic information (age, number of pregnancies, number of births, number of miscarriages, past histories, menstruation periods, regularity of menstruations, periods of menstrual flow, severity of dysmenorrhea, urges to defecate, dyspareunia, whether other pains exist and other concomitant histories); clinical test values (endometrioma counts, sizes of endometriomas, follicle counts, CA125 blood values, sizes of uteruses, level of ovarian adhesions and contents of endometriomas); treatment-related information (medication prior to surgery, medication following surgery, route of drug administration, surgical method, surgical routine, UGA method, UGA site, UGA with irrigation and medication used) |
SE = NR SP = NR |
|
| Random Forest | Bendifallah et al.50 | Not specified | Ovarian, superficial or deep endometriosis | Training set (1126 patients), validation set (100 patients) | See above. |
SE = 92% SP = 92% |
| eXtreme Gradient Boosting | Bendifallah et al.50 | Not specified | Ovarian, superficial or deep endometriosis | Training set (1126 patients), validation set (100 patients) | See above. |
SE = 93% SP = 92% |
| Voting Classifier (soft/hard) | Bendifallah et al.50 | Not specified | Ovarian, superficial or deep endometriosis | Training set (1126 patients), validation set (100 patients) | See above. |
Voting Classifier Soft SE = 93% SP = 88% Voting Classifier Hard SE = 91% SP = 92% |
NR not reported, DE deep endometriosis, ICSI-IVF intracytoplasmic sperm injection in vitro fertilization, rAFS revised American Fertility Society, OCP oral contraceptive pill, VAS visual analogic scale, BMI body mass index, CA-125 cancer antigen 125, UGA ultrasound-guided aspiration, SE sensitivity, SP specificity.
aMinimal, mild, moderate and severe stages of endometriosis were included.