Predicting oncological outcomes in early-stage cervical cancer: follow-up analysis using medical informatics methods