Ultrasound Radiomics and Peripheral Blood Indices Model for Predicting Pregnancy and Recurrence After TCRA in...

Area of Science:

• Reproductive Medicine
• Medical Imaging
• Biostatistics

Background:

• Intrauterine adhesions (IUA) pose significant concerns for pregnancy and recurrence in women of childbearing age.
• Current prevention and treatment strategies for IUA are limited, necessitating improved clinical decision-making tools.

Purpose of the Study:

• To develop a prediction model for intrauterine adhesions pregnancy and recurrence (IUA-PR).
• To guide clinical decision-making for patients with IUA regarding postoperative fertility outcomes.

Main Methods:

• Retrospective analysis of 387 IUA patients (Jan 2021-Dec 2023).
• Extraction of radiomic features from ultrasound images and integration with peripheral blood parameters.
• Utilized LASSO and BIC for feature selection; evaluated model efficacy using AUC and DCA.

Main Results:

• A model incorporating five peripheral blood inflammatory indices and six ultrasound radiomic parameters was constructed.
• The pregnancy prediction model achieved an AUC of 0.886, utilizing PLR, NLR, AISI, Rad-score, and menstrual status.
• The recurrence prediction model achieved an AUC of 0.720, using SIRI, SII, Rad-score, and menstrual status.

Conclusions:

• A novel IUA-PR prediction model was successfully developed using peripheral blood inflammatory parameters and ultrasound radiomics.
• The generalized linear regression model demonstrates superior predictive performance for postoperative pregnancy and recurrence.
• This convenient and low-cost model assists patients in making informed postoperative fertility decisions.