Table 1_Nomogram model for predicting papillary thyroid carcinoma based on clinical-ultras

A multicenter study from Institutions A, B, C, and D developed a nomogram model to differentiate papillary thyroid carcinoma from benign nodules. The model integrates patient age, ultrasound features, and the inflammatory biomarker NLR, validated across multiple cohorts. The dataset, authored by Caihong Wang and last updated in April 2026, contains the model's construction and validation details.

Use Cases

• Training a binary classifier for thyroid nodule malignancy based on clinical and ultrasound features.
• Benchmarking new predictive models against the reported AUC performance metrics (e.g., 0.841 in training).
• Investigating the role of inflammatory biomarkers like NLR in cancer prediction models.
• Developing clinical decision support tools based on nomogram methodology described in the study.

Strengths

• Model validated across four distinct cohorts with reported AUCs, including an internal validation cohort (n=220) and two external validation cohorts (n=164 and n=143).
• Performance assessed with multiple metrics including AUC, decision curve analysis (DCA), and clinical impact curve (CIC).
• Includes bootstrap resampling validation (1000 resamples) on the entire dataset, reporting an AUC of 0.826.

Limitations

• Row count is unknown, which may limit suitability assessment for direct ML training.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is a 1.2 MB DOCX file, which may require parsing to extract structured data.

Provenance

Source

figshare, authored by Caihong Wang.

Collection Method

Data collected from a multicenter study across four institutions, with cohorts divided for training and validation.

Time Range

null

Freshness

Last updated 2026-04-13 04:34:58

Geography

null