CT Radiomics Model for Predicting Pain Relief After Radiotherapy in Bone Metastases

134 patient records from a dual-center study conducted between January 2022 and December 2024. The dataset was created by Zhiling Wan to develop and validate a CT-based radiomics model for predicting pain relief after palliative radiotherapy. It includes clinical variables and radiomic features, with results showing a k-nearest neighbors model achieving an AUC of up to 0.823.

Use Cases

• Training machine learning classifiers for pain response prediction based on radiomic features.
• Comparing the performance of different ML algorithms for a clinical prediction task.
• Validating a radiomics signature for bone metastasis treatment across different patient cohorts.
• Conducting decision curve analysis to assess the net clinical benefit of a predictive model.

Strengths

• Includes data from 134 patients across two medical centers, providing a basis for internal and external validation.
• Compares 11 different machine learning algorithms, with the KNN model's performance detailed across training and test sets.
• Features a clear clinical endpoint defined by international consensus (pain relief vs. non-relief).

Limitations

• Row count and column-level documentation are unknown, which may limit suitability assessment.
• The dataset is very small at 10.9 KB, indicating limited scope, likely containing summary results rather than raw patient data.
• Data may reflect geographic and temporal bias inherent to the two hospitals in Xuzhou used for the study.

Provenance

Source

Retrospectively enrolled patients from Xuzhou Central Hospital and Xuzhou First People’s Hospital.

Collection Method

Clinical variables and CT images were collected; radiomic features were extracted and screened from delineated tumor ROIs.

Time Range

January 2022 to December 2024

Freshness

Last updated 2026-04-21 04:20:13; freshness should be verified.

Geography

Patients from two hospitals in Xuzhou, China.