Target:
This study explored the use of ultrasound for assessing inguinofemoral lymph nodes in vulvar cancer patients and aimed to develop a machine learning-based model for distinguishing between metastatic and non-metastatic nodes. They examined 237 inguinal regions in 127 women, using ultrasound prior to surgery. Fourteen informative features were collected and used to train the machine learning model, known as the Morphonode Predictive Model, which integrated various data classifiers.
The results were promising: the random forest classifier (RFC) achieved 93.3% accuracy and a 97.1% negative predictive value, while the decision tree (DT) identified four specific signatures linked to metastasis risk, each with different predictive values. This Morphonode Predictive Model shows potential for clinical integration, aiding in the preoperative stratification of vulvar cancer patients, offering an accurate and non-invasive method for assessing inguinofemoral lymph nodes.
doi: 10.3390/cancers15041121