A combined model of T2-weighted imaging radiomics and clinical imaging features was effective for predicting the risk of reintervention prior to high-intensity focused ultrasound ablation.
Uterine fibroids are described as benign gynecological tumors that are treated by high-intensity focused ultrasound (HIFU) ablation, among other interventions. There is a lack of radiomic studies in the medical literature exploring the risk of reintervention following HIFU.
A recent study published in the International Journal of Hyperthermia predicted the reintervention risk of uterine fibroids after HIFU by utilizing T2-weighted imaging (T2WI) radiomics and clinical imaging features before performing the ablation.
A total of 226 patients were part of the reintervention follow-up survey in the training and validation cohorts. The median duration of follow-up was 33 months (interquartile range: 25–41 months). Of the total patients, four developed increased fibroid volume or symptomatic recurrence of fibroids; however, the treatment reintervention for these patients was not determined. Approximately 23.3% of the remaining patients were subjected to reintervention treatment, and the median duration of the interval between HIFU and reintervention was 24 months (interquartile range: 13–33 months).
Clinical Imaging Features and T2WI Radiomics in Postoperative Reintervention Risk
After eliminating patients lost to follow-ups, a total of 180 patients were present in the training and validation cohorts of the study. Compared to the non-reintervention group, patients in the intervention group were younger, had larger fibroid volumes, predominantly hyperintense fat-suppressed T2W (T2WI-FS) signals, and had a higher degree of fibroid enhancement. The differences across the intervention and re-intervention groups were statistically significant.
Of the total radiomic features, 798 had good inter-observer agreement, and nine optimal radiomic features were utilized in developing the clinical-imaging model. The predictive performance performed significantly better in the combined model compared to the clinical-imaging model in the training cohort, according to the results of the DeLong test.
Clinical Application of the Combined Model
The predictive accuracy was better in the combined model in the validation cohort, as represented by the Brier score of 0.174. The combined model was associated with a greater net benefit compared to the no-treatment, clinical imaging, radiomics, and all-treatment models.
This study emphasizes the importance of utilizing a complete prediction model that combines clinical imaging features and T2WI radiomics in order to evaluate the likelihood of postoperative reintervention after undergoing HIFU ablation for uterine fibroids. This novel methodology provides healthcare professionals with a mechanism to customize therapeutic approaches, potentially reducing the necessity for supplementary therapies. Moreover, the model’s capacity to improve the treatment of chronic diseases following ablation procedures has the potential to bring about long-term reductions in symptoms associated with fibroids and boost the overall well-being of patients.
Qin, S., Yang, L., Liu, N., Zheng, Y., Jia, Q., & Xiao-Hua, H. (2023). Prediction of postoperative reintervention risk for uterine fibroids using clinical-imaging features and T2WI radiomics before high-intensity focused ultrasound ablation. International Journal of Hyperthermia, 40(1). https://doi.org/10.1080/02656736.2023.2226847