TRV-2026-3002Version 1 · Published

Saved · see the current record

Why this version exists

Certified into the record

The exact text we saved

This is the version exactly as it was saved, character for character. The SHA-256 hash below is a 64-character code worked out from that text. Run the same text through SHA-256 and you get the same code; change a single letter and the code comes out completely different. Each version also stores the previous version’s hash, which chains them together so no version can be swapped out quietly.

TRUVACE RECORD VERSION
record: TRV-2026-3002
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-11-17T18:02:34.098705Z
status: published
lens: g_space
sector: health
headline: A Machine Learning Model Based on Ultrasound Radiomics and Multi-modal Clinical Features for Predicting Lymph Node Metastasis of Papillary Thyroid Carcinoma in the Setting of Hashimoto's Thyroiditis
dek: To investigate the value of multi-modal features (ultrasound characteristics, BRAF V600E mutation, and TPO-Ab) in predicting lymph node metastasis (LNM) of papillary thyroid carcinoma (PTC) in the context of Hashimoto's thyroiditis (HT), and to construct machine learning models for individualized risk assessment. A retrospective cohort of 377 patients with pathologically confirmed PTC and concurrent HT was enrolled and divided into LNM n=186 and non-LNM n=191 groups. Preoperative ultrasound features (e.g., Minim…
gain_title: In a study of 377 patients with papillary thyroid cancer and Hashimoto's thyroiditis, an XGBoost model combining ultrasound radiomics and clinical markers reached 0.879 accuracy at predicting lymph node metastasis in an independent external cohort of 107 patients.
problem_title: (none)
trace_subject: (none)
gain_reading: In a study of 377 patients with papillary thyroid cancer and Hashimoto's thyroiditis, an XGBoost model combining ultrasound radiomics and clinical markers reached 0.879 accuracy at predicting lymph node metastasis in an independent external cohort of 107 patients.
gain_evidence: the optimal XGBoost model achieved a stable accuracy of 0.879 in an independent external cohort (n = 107)
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers built machine learning models to estimate the risk that papillary thyroid carcinoma has spread to lymph nodes in patients who also have Hashimoto's thyroiditis. The best model, XGBoost, reached an accuracy of 0.879 in an independent external cohort of 107 patients.

The models were trained on a retrospective cohort of 377 patients, 186 with lymph node metastasis and 191 without. They used preoperative ultrasound features, TPO-Ab blood levels and BRAF V600E mutation status. Radiomics features accounted for 81.5% of the model's predictive contribution.

The authors say the approach could support individualized risk stratification and clinical decision-making for these patients. The evidence comes from retrospective records rather than a prospective clinical deployment.
limitation: The model was built from a retrospective cohort of 377 patients and checked in an external cohort of only 107 patients.
tag: Evidence-backed gain
key_points: The study enrolled 377 patients with pathologically confirmed papillary thyroid carcinoma and concurrent Hashimoto's thyroiditis: 186 with lymph node metastasis and 191 without. | Five algorithms were compared: random forest, support vector machine, XGBoost, logistic regression and k-nearest neighbors. | In univariate analysis, calcification type, nodule diameter, minimum tumor-to-capsule distance and aspect ratio were significantly associated with metastasis; nodule diameter contributed most. | Radiomics features supplied 81.5% of predictive contribution, morphological features 12.7%, TPO-Ab 3.8% and BRAF V600E 2.0%. | The best XGBoost model reached 0.879 accuracy in an independent external cohort of 107 patients.
rundown: The retrospective cohort split into 186 patients with lymph node metastasis and 191 without. Preoperative ultrasound features such as minimum tumor-to-capsule distance, calcification type and ratio, lesion location and nodule morphology were collected alongside serological and pathological indicators.

Five models (random forest, support vector machine, XGBoost, logistic regression and k-nearest neighbors) were built and assessed with ROC and calibration curves. Combining several feature types improved prediction accuracy, the authors report.

Feature importance analysis showed radiomics features contributing 81.5%, morphological features 12.7%, TPO-Ab 3.8% and BRAF V600E 2.0%. The authors conclude the models enable individualized risk stratification for lymph node metastasis in this patient group.
sources:
- peer_reviewed | Ultrasound in Medicine & Biology | https://doi.org/10.1016/j.ultrasmedbio.2026.10.008 | 2026-11-16
prev: 0000000000000000000000000000000000000000000000000000000000000000
SHA-256 hash
aca42b45c3343f42a260e1a2c98b85b23b906a872e345c31d2b152ce4a306df2
Previous version’s hash
0000000000000000000000000000000000000000000000000000000000000000
Check this record hasn’t been altered
How to check it yourself, without trusting this page

Download any version from /api/record/TRV-2026-3002 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.