OSCR

Modality-level attribution and redundancy-aware radiomics for MRI-based differentiation of melanoma and NSCLC brain metastases.

Code ↔ Paper

11 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

9 blocks of code here have a permanent identifier (oscr:b:…), listed with the matches below: citable for good, with a page of its own. Select a match and press Cite to copy it. What a block identifier is.

The 11 matches
  1. [1] § Methodology › Radiomic features importance ranking ↔ rank_and_split__4.py, lines 99 to 190 · score 0.88 · class_weight, logistic regression, train CV, correlation filter, missingness, saga · block oscr:b:0G24BKD7MA645-9 · Report a mismatch
  2. [2] § Methodology › Radiomic features importance ranking ↔ logreg__5a.py, lines 173 to 183 · score 0.76 · class_weight, logistic regression, correlation filter, saga, L1, solver · block oscr:b:ZBZZGV2EMDH2P-O · Report a mismatch
  3. [3] § Methodology › Explainability ↔ svc__5d.py, lines 554 to 634 · score 0.72 · KernelSHAP, model agnostic, bounded, Platt, explanations, SVC · block oscr:b:WFWHT0E8YR3WP-B · Report a mismatch
  4. [4] § Methodology › Model development and hyperparameter optimization ↔ rank_and_split__4.py, lines 99 to 190 · score 0.70 · logistic regression, train CV, correlation threshold, penalty, stratified, fold · block oscr:b:0G24BKD7MA645-9 · Report a mismatch
  5. [5] § Methodology › Model development and hyperparameter optimization ↔ randfor__5c.py, lines 221 to 310 · score 0.70 · train CV, correlation threshold, Optuna, Hyperparameter, depth, optimization · block oscr:b:MZGBXRW7GPGZN-A · Report a mismatch
  6. [6] § Methodology › Model development and hyperparameter optimization ↔ collect_val_test_reports.py, lines 225 to 333 · score 0.68 · imbalance aware, PR AUC, precision, Youden, sensitivity, validation · block oscr:b:6AW6G58SMQEB8-O · Report a mismatch
  7. [7] § Results › Classification ↔ collect_val_test_reports.py, lines 225 to 333 · score 0.63 · imbalance aware metrics, PR AUC, sensitivity, probability, classification, threshold · block oscr:b:6AW6G58SMQEB8-O · Report a mismatch
  8. [8] § Methodology › Explainability ↔ lightGBM__5b.py, lines 656 to 701 · score 0.57 · Pareto style, cumulative importance · block oscr:b:RW2Z8C14ZZEFV-U · Report a mismatch
  9. [9] § Methodology › Explainability ↔ randfor__5c.py, lines 670 to 723 · score 0.57 · Pareto style, cumulative importance · block oscr:b:73C560VMY8R8S-V · Report a mismatch
  10. [10] § Methodology › Radiomics extraction ↔ all_lesions__1/pipeline_extract_catalogue__1.py, lines 63 to 94 · score 0.57 · bin width, Radiomic feature extraction, spline, resampled, mask · block oscr:b:740Z2DHA9QNKP-C · Report a mismatch
  11. [11] § Methodology › Radiomics extraction ↔ pipeline_extract_catalogue__1.py, lines 153 to 273 · score 0.56 · PyRadiomics, Bias, isotropic, N4, resampled, bin · block oscr:b:8871D86HP9SE2-W · Report a mismatch

Paper

Loaded from Europe PMC by your browser; OSCR does not store it: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

Python · 248 lines · 10 KB · BSD-3-Clause · 2 matches

  1. #!/usr/bin/env python3
  2. import os
  3. os.environ["OMP_NUM_THREADS"] = "1"
  4. os.environ["MKL_NUM_THREADS"] = "1"
  5. os.environ["OPENBLAS_NUM_THREADS"] = "1"
  6. import numpy as np
  7. import pandas as pd
  8. from pathlib import Path
  9. import argparse
  10. import json
  11. from sklearn.base import BaseEstimator, TransformerMixin
  12. from sklearn.model_selection import StratifiedKFold
  13. from sklearn.pipeline import Pipeline
  14. from sklearn.preprocessing import StandardScaler, MinMaxScaler
  15. from sklearn.impute import SimpleImputer
  16. from sklearn.linear_model import LogisticRegression
  17. from sklearn.feature_selection import VarianceThreshold, SelectPercentile, f_classif, chi2, mutual_info_classif
  18. from sklearn.exceptions import NotFittedError
  19. from pandas.api.types import is_numeric_dtype
  20. # ============================== CONFIG ==============================
  21. parser = argparse.ArgumentParser()
  22. parser.add_argument("--topk", type=int, required=True, help="Top-k features to use")
  23. parser.add_argument("--config", type=str, required=True, help="Path to config JSON")
  24. args = parser.parse_args()
  25. TOPK_FEATURES = args.topk
  26. with open(args.config, "r") as f:
  27. cfg = json.load(f)["code4"]
  28. INPUT_XLSX = cfg["input_xlsx"]
  29. OUTPUT_DIR = f"./top_{TOPK_FEATURES}_features"
  30. os.makedirs(OUTPUT_DIR, exist_ok=True)
  31. OUTPUT_TRAIN_RAW = OUTPUT_DIR + "/traincv_set__4.xlsx"
  32. OUTPUT_TEST_RAW = OUTPUT_DIR + "/test_set__4.xlsx"
  33. OUTPUT_IMPORTANCE = OUTPUT_DIR + "/importance_traincv__4.csv"
  34. OUTPUT_FEATURELIST = OUTPUT_DIR + "/feature_list__4.csv"
  35. OUTPUT_TRAIN_TOPK = OUTPUT_DIR + f"/traincv_set__4_top{TOPK_FEATURES}.xlsx"
  36. OUTPUT_TEST_TOPK = OUTPUT_DIR + f"/test_set__4_top{TOPK_FEATURES}.xlsx"
  37. OUTPUT_FAMILY_SUM = OUTPUT_DIR + f"/radiomics_family_summary_top{TOPK_FEATURES}__4.csv"
  38. RANDOM_SEED = cfg.get("random_seed", 42)
  39. N_TEST_MINORITY = cfg["test_patients"]["minority"]
  40. N_TEST_MAJORITY = cfg["test_patients"]["majority"]
  41. N_SPLITS = cfg["n_splits"]
  42. TARGET_COL = cfg["grouping"]["target_col"]
  43. CLASSES = list(cfg["grouping"]["classes"].values())
  44. PRIMARY_KEY_1 = "subject_id"
  45. PRIMARY_KEY_2 = "roi"
  46. FS_CFG = cfg.get("feature_selection", {})
  47. # ===================== helpers: split =====================
  48. def choose_patients_by_class(df: pd.DataFrame, target_col: str, n_min: int, n_maj: int, rng: np.random.Generator):
  49. subjects = df[[PRIMARY_KEY_1, target_col]].drop_duplicates()
  50. grouped = subjects.groupby(target_col)[PRIMARY_KEY_1].unique().to_dict()
  51. min_class, maj_class = CLASSES[0], CLASSES[1]
  52. min_subj = grouped.get(min_class, [])
  53. maj_subj = grouped.get(maj_class, [])
  54. min_sample = rng.choice(min_subj, size=min(n_min, len(min_subj)), replace=False)
  55. maj_sample = rng.choice(maj_subj, size=min(n_maj, len(maj_subj)), replace=False)
  56. return np.concatenate([min_sample, maj_sample])
  57. # ===================== Corr filter =====================
  58. class CorrelationFilter(BaseEstimator, TransformerMixin):
  59. def __init__(self, threshold=0.95, method="spearman"):
  60. self.threshold = threshold
  61. self.method = method
  62. self.keep_features_ = None
  63. def fit(self, X, y=None):
  64. Xdf = pd.DataFrame(X).copy()
  65. med = pd.Series(np.nanmedian(Xdf.values, axis=0), index=Xdf.columns)
  66. Xdf = Xdf.fillna(med)
  67. valid = Xdf.notna().sum(axis=0) >= 2
  68. Xdf = Xdf.loc[:, valid]
  69. corr = Xdf.corr(method=self.method).abs()
  70. upper = corr.where(np.triu(np.ones(corr.shape), k=1).astype(bool))
  71. drop = set()
  72. for col in upper.columns:
  73. if col in drop:
  74. continue
  75. high = upper.index[upper[col] >= self.threshold].tolist()
  76. drop.update(high)
  77. self.keep_features_ = [c for c in Xdf.columns if c not in drop]
  78. return self
  79. def transform(self, X):
  80. Xdf = pd.DataFrame(X)
  81. return Xdf[self.keep_features_].values
  82. # ===================== importance on train-cv =====================
  83. def compute_importance_traincv(df_traincv: pd.DataFrame):
  84. df = df_traincv.copy()
  85. df.columns = df.columns.map(lambda c: str(c).strip())
  86. y = df[TARGET_COL].map({CLASSES[0]: 0, CLASSES[1]: 1}).values
  87. feat_cols = [c for c in df.columns if c not in {PRIMARY_KEY_1, PRIMARY_KEY_2, TARGET_COL} and is_numeric_dtype(df[c])]
  88. print(f"Numeric features considered: {len(feat_cols)}")
  89. original_count = len(feat_cols)
  90. X = df[feat_cols].copy()
  91. X.columns = X.columns.astype(str)
  92. # Variance Threshold
  93. if FS_CFG.get("variance_threshold", {}).get("enabled", True):
  94. thresh = FS_CFG["variance_threshold"].get("threshold", 1e-8)
  95. vt = VarianceThreshold(threshold=thresh)
  96. X = X.loc[:, vt.fit(X).get_support()]
  97. print(f"[Variance Threshold] Features remaining: {X.shape[1]} ({100*X.shape[1]/original_count:.1f}%)")
  98. # Missingness Filter
  99. if FS_CFG.get("missingness_filter", {}).get("enabled", True):
  100. max_nan_frac = FS_CFG["missingness_filter"].get("max_nan_fraction", 0.2)
  101. keep_mask = X.isna().mean() <= max_nan_frac
  102. X = X.loc[:, keep_mask]
  103. print(f"[Missingness Filter] Features remaining: {X.shape[1]} ({100*X.shape[1]/original_count:.1f}%)")
  104. # Correlation Filter
  105. if FS_CFG.get("correlation_filter", {}).get("enabled", True):
  106. corr_thresh = FS_CFG["correlation_filter"].get("threshold", 0.95)
  107. corr = CorrelationFilter(threshold=corr_thresh)
  108. corr.fit(X)
  109. X = pd.DataFrame(X[corr.keep_features_], columns=corr.keep_features_)
  110. print(f"[Correlation Filter] Features remaining: {X.shape[1]} ({100*X.shape[1]/original_count:.1f}%)")
  111. # ANOVA (f_classif)
  112. if FS_CFG.get("anova_f_classif", {}).get("enabled", False):
  113. sel = SelectPercentile(f_classif, percentile=FS_CFG["anova_f_classif"].get("percentile", 50))
  114. sel.fit(X.fillna(0), y)
  115. X = X.loc[:, sel.get_support()]
  116. print(f"[ANOVA F-test] Features remaining: {X.shape[1]} ({100*X.shape[1]/original_count:.1f}%)")
  117. # Chi-squared
  118. if FS_CFG.get("chi2", {}).get("enabled", False):
  119. X_safe = X.fillna(0)
  120. if (X_safe >= 0).all().all():
  121. chi = SelectPercentile(chi2, percentile=FS_CFG["chi2"].get("percentile", 50))
  122. chi.fit(X_safe, y)
  123. X = X.loc[:, chi.get_support()]
  124. print(f"[Chi2 Test] Features remaining: {X.shape[1]} ({100*X.shape[1]/original_count:.1f}%)")
  125. else:
  126. print("[Chi2 Test] Skipped: negative values detected")
  127. # Mutual Information
  128. if FS_CFG.get("mutual_info", {}).get("enabled", False):
  129. mi = SelectPercentile(mutual_info_classif, percentile=FS_CFG["mutual_info"].get("percentile", 50))
  130. mi.fit(X.fillna(0), y)
  131. X = X.loc[:, mi.get_support()]
  132. print(f"[Mutual Info] Features remaining: {X.shape[1]} ({100*X.shape[1]/original_count:.1f}%)")
  133. feat_cols_final = X.columns.tolist()
  134. pipe = Pipeline([
  135. ("imp", SimpleImputer(strategy="median")),
  136. ("sc", StandardScaler()),
  137. ("clf", LogisticRegression(
  138. penalty="l1", solver="saga", C=1.0, max_iter=10000,
  139. class_weight="balanced", random_state=RANDOM_SEED, n_jobs=-1
  140. )),
  141. ])
  142. skf = StratifiedKFold(n_splits=N_SPLITS, shuffle=True, random_state=RANDOM_SEED)
  143. coef_list = []
  144. for fold_i, (tr, te) in enumerate(skf.split(X, y), start=1):
  145. Xtr, ytr = X.iloc[tr], y[tr]
  146. if len(np.unique(ytr)) < 2:
  147. print(f"[Fold {fold_i}] Skipped due to single-class train split")
  148. continue
  149. pipe.fit(Xtr, ytr)
  150. coefs = pd.Series(pipe.named_steps["clf"].coef_[0], index=X.columns)
  151. coef_list.append(coefs)
  152. if not coef_list:
  153. raise RuntimeError("All CV folds failed due to class imbalance.")
  154. coef_df = pd.concat(coef_list, axis=1)
  155. coef_df.columns = [f"fold{i+1}" for i in range(len(coef_list))]
  156. median_abs_coef = coef_df.abs().median(axis=1, skipna=True)
  157. imp = pd.DataFrame({"median_abs_coef": median_abs_coef}).sort_values("median_abs_coef", ascending=False)
  158. return imp, feat_cols
  159. # ===================== main =====================
  160. def main():
  161. rng = np.random.default_rng(RANDOM_SEED)
  162. df_all = pd.read_excel(INPUT_XLSX)
  163. df_all[PRIMARY_KEY_1] = df_all[PRIMARY_KEY_1].astype(str)
  164. df_all = df_all[df_all[TARGET_COL].isin(CLASSES)].copy()
  165. print(f"Total samples after filtering target classes: {df_all.shape[0]}")
  166. test_subjects = choose_patients_by_class(df_all, TARGET_COL, N_TEST_MINORITY, N_TEST_MAJORITY, rng)
  167. mask_test = df_all[PRIMARY_KEY_1].isin(test_subjects)
  168. df_test_raw = df_all.loc[mask_test].copy()
  169. df_train_raw = df_all.loc[~mask_test].copy()
  170. print(f"Train patients: {df_train_raw[PRIMARY_KEY_1].nunique()} | Test patients: {df_test_raw[PRIMARY_KEY_1].nunique()}")
  171. df_train_raw.to_excel(OUTPUT_TRAIN_RAW, index=False)
  172. df_test_raw.to_excel(OUTPUT_TEST_RAW, index=False)
  173. imp, feat_cols_all = compute_importance_traincv(df_train_raw)
  174. imp.to_csv(OUTPUT_IMPORTANCE, index=True)
  175. remaining_feats = [f for f in imp.index.tolist() if f in df_all.columns]
  176. if len(remaining_feats) < TOPK_FEATURES:
  177. print(f"Warning: Less than {TOPK_FEATURES} features remaining: Keeping only {len(remaining_feats)}.")
  178. TOPK_FINAL = len(remaining_feats)
  179. else:
  180. TOPK_FINAL = TOPK_FEATURES
  181. top_feats = remaining_feats[:TOPK_FINAL]
  182. pd.Series(top_feats, name="feature").to_csv(OUTPUT_FEATURELIST, index=False)
  183. cols_final = [PRIMARY_KEY_1, PRIMARY_KEY_2, TARGET_COL] + top_feats
  184. for c in cols_final:
  185. if c not in df_train_raw.columns:
  186. df_train_raw[c] = np.nan
  187. if c not in df_test_raw.columns:
  188. df_test_raw[c] = np.nan
  189. df_train_topk = df_train_raw.loc[:, cols_final]
  190. df_test_topk = df_test_raw.loc[:, cols_final]
  191. df_train_topk.to_excel(OUTPUT_TRAIN_TOPK, index=False)
  192. df_test_topk.to_excel(OUTPUT_TEST_TOPK, index=False)
  193. def family_of(name: str) -> str:
  194. return name.split("__", 1)[0] if isinstance(name, str) and "__" in name else "misc"
  195. print(f"Top-{TOPK_FINAL} features kept.")
  196. topk_for_family = imp.head(TOPK_FINAL).copy()
  197. families = topk_for_family.index.to_series().apply(family_of)
  198. family_summary = families.value_counts().rename_axis("family").reset_index(name=f"count_in_top{TOPK_FINAL}")
  199. family_summary.to_csv(OUTPUT_FAMILY_SUM, index=False)
  200. if __name__ == "__main__":
  201. main()

rank_and_split__4.py at commit 0e87668, under BSD-3-Clause · at the source

Overview

Authors: Rafail C Christodoulou1, Georgios Vamvouras2, Amal El Masri3, Platon S Papageorgiou4, Evros Vassiliou5, Elena E Solomou6, Sokratis G Papageorgiou7, Michalis F Georgiou8
  1. Division of Neuroimaging and Neurointervention, Department of Radiology, Stanford University, Stanford, CA, United States
  2. Department of Electrical and Computer Engineering, National Technical University of Athens (NTUA), Athens, Greece
  3. Faculty of Medicine, American University of Beirut, Beirut, Lebanon
  4. Department of Medicine, National and Kapodistrian University of Athens, Athens, Greece
  5. Department of Biological Sciences, Kean University, Union, NJ, United States
  6. Internal Medicine-Hematology, University of Patras Medical School, Rion, Greece
  7. 1st Department of Neurology, Medical School, National and Kapodistrian University of Athens, Eginition Hospital, Athens, Greece
  8. Department of Radiology, University of Miami, Miami, FL, United States
Journal: Frontiers in oncology, volume 16, article 1816015
Dates: received 23 February 2026; accepted 8 June 2026; published online 29 June 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.3389/fonc.2026.1816015 · PMID 42444810 · PMCID PMC13357173 · OpenAlex W7166529421
Open access: gold, a free copy (OpenAlex)
Status: code verified (how the code is found and verified)
Categories: structural MRI / diffusion (modality), other condition (population)
Methods: Connectivity, Statistics, Machine learning, Preprocessing, fMRI & imaging, Physiology & signal measures
Keywords: arteficial intelligence, explainable AI (SHAP), melanoma, NSCLC, personalized & precision medicine, radiomics
Topic: Radiomics and Machine Learning in Medical Imaging (Radiology, Nuclear Medicine and Imaging, Medicine), according to OpenAlex
Citations: not cited yet (Europe PMC); 23 references in the paper

Abstract

Background: Differentiating the primary source of brain metastases using imaging alone is difficult, especially when intracranial disease is the initial sign of cancer. This study aimed to develop and validate a redundancy-aware, explainable radiomics framework for distinguishing melanoma from non–small cell lung cancer (NSCLC) brain metastases through multiparametric MRI.

Methods: Lesion-level radiomic features were extracted from T1, contrast-enhanced T1 (T1CE), T2, and FLAIR sequences. Correlation-based redundancy filtering and L1-regularized feature ranking were applied to minimize collinearity among features. A structured parametric analysis across progressively larger top-K feature subsets was performed to evaluate performance stability and modality-specific contribution patterns. Multiple machine learning classifiers were tested using cross-validation and an independent test set, with model interpretability assessed via TreeSHAP to quantify feature- and modality-level contributions.

Results: The best-performing configuration achieved a test AUC of 0.75, while RF-200 was selected for the primary tree-based explainability analysis due to its stable AUC-based discrimination. SHAP analysis showed that classification depended on a small set of predictors combining texture and intensity distribution features. FLAIR and T2 features contributed most to model attribution, suggesting that fluid-sensitive sequences capture discrimination-relevant heterogeneity and microenvironmental signals beyond contrast enhancement.

Discussion: A redundancy-aware, interpretable radiomics approach shows promising preliminary discrimination between melanoma and NSCLC brain metastases and provides structured insight into modality-specific sequence contributions. External validation and prospective reader studies are required before clinical translation.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repository

Its files are read in the Code ↔ Paper reader above, with 11 matches between paragraphs and lines of code.

georgeDV2002/metastasis_radiomics

License: BSD-3-Clause
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: 0e87668ae65c74b0fc6f9185e9a872f5b97ff989, 1 May 2026
Languages: Python (23)
Size: 30 files, 23 scripts
Software Heritage: not archived
Found in: “Code availability”
Holds: README, license file, environment (requirements.txt)
Not found: CITATION.cff, tests, continuous integration, documentation
Tools: pandas (21 files), NumPy (20 files), Matplotlib (12 files), scikit-learn (9 files), SciPy (3 files), SHAP (3 files), NiBabel (2 files), PyRadiomics (2 files), SimpleITK (2 files), LightGBM (1 file), seaborn (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
25 files

Code availability

The complete implementation of the preprocessing, radiomics extraction, feature selection, model training, calibration, and explainability pipeline is publicly available at: https://github.com/georgeDV2002/metastasis_radiomics.

Reproduced under the paper's license (CC BY), from the paper cited above.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 23 scripts, each with its path and the digest of its content;
  • 11 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data availability statement

Publicly available datasets were analyzed in this study. This data can be found here: https://www.cancerimagingarchive.net/collection/pretreat-metstobrain-masks/.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, pages, dates, 8 authors, 6 keywords, 23 references.

Cite

This paper

Christodoulou, R. C., Vamvouras, G., El Masri, A., Papageorgiou, P. S., Vassiliou, E., Solomou, E. E., Papageorgiou, S. G., & Georgiou, M. F. (2026). Modality-level attribution and redundancy-aware radiomics for MRI-based differentiation of melanoma and NSCLC brain metastases. Frontiers in oncology, 16, 1816015. https://doi.org/10.3389/fonc.2026.1816015

BibTeX

@article{christodoulou2026modality,
author = {Christodoulou, Rafail C and Vamvouras, Georgios and El Masri, Amal and Papageorgiou, Platon S and Vassiliou, Evros and Solomou, Elena E and Papageorgiou, Sokratis G and Georgiou, Michalis F},
title = {{Modality-level attribution and redundancy-aware radiomics for MRI-based differentiation of melanoma and NSCLC brain metastases}},
journal = {Frontiers in oncology},
year = {2026},
month = jun,
volume = {16},
pages = {1816015},
publisher = {Frontiers Media SA},
issn = {2234-943X},
doi = {10.3389/fonc.2026.1816015},
url = {https://doi.org/10.3389/fonc.2026.1816015},
pmid = {42444810},
pmcid = {PMC13357173}
}

RIS

TY - JOUR
AU - Christodoulou, Rafail C
AU - Vamvouras, Georgios
AU - El Masri, Amal
AU - Papageorgiou, Platon S
AU - Vassiliou, Evros
AU - Solomou, Elena E
AU - Papageorgiou, Sokratis G
AU - Georgiou, Michalis F
TI - Modality-level attribution and redundancy-aware radiomics for MRI-based differentiation of melanoma and NSCLC brain metastases
T2 - Frontiers in oncology
J2 - Front Oncol
PY - 2026
DA - 2026/06/29
VL - 16
SP - 1816015
SN - 2234-943X
PB - Frontiers Media SA
DO - 10.3389/fonc.2026.1816015
UR - https://doi.org/10.3389/fonc.2026.1816015
LA - en
ER -

CSL-JSON

{
"id": "10.3389/fonc.2026.1816015",
"type": "article-journal",
"title": "Modality-level attribution and redundancy-aware radiomics for MRI-based differentiation of melanoma and NSCLC brain metastases",
"container-title": "Frontiers in oncology",
"author": [
{
"family": "Christodoulou",
"given": "Rafail C"
},
{
"family": "Vamvouras",
"given": "Georgios"
},
{
"family": "El Masri",
"given": "Amal"
},
{
"family": "Papageorgiou",
"given": "Platon S"
},
{
"family": "Vassiliou",
"given": "Evros"
},
{
"family": "Solomou",
"given": "Elena E"
},
{
"family": "Papageorgiou",
"given": "Sokratis G"
},
{
"family": "Georgiou",
"given": "Michalis F"
}
],
"container-title-short": "Front Oncol",
"volume": "16",
"page": "1816015",
"DOI": "10.3389/fonc.2026.1816015",
"PMID": "42444810",
"PMCID": "PMC13357173",
"ISSN": "2234-943X",
"publisher": "Frontiers Media SA",
"URL": "https://doi.org/10.3389/fonc.2026.1816015",
"language": "en",
"issued": {
"date-parts": [
[
2026,
6,
29
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.3389/fneur.2024.1474461 [code]
Utilizing machine-learning techniques on MRI radiomics to identify primary tumors in brain metastases.
Journal: Frontiers in neurology
In common: PyRadiomics, SHAP, SimpleITK, 6 other tools, structural MRI / diffusion, other condition, 3 references
[2] doi:10.3390/life15040606 [code]
A Combined Approach Using T2*-Weighted Dynamic Susceptibility Contrast MRI Perfusion Parameters and Radiomics to Differentiate Between Radionecrosis and Glioma Progression: A Proof-of-Concept Study.
Journal: Life (Basel, Switzerland)
In common: PyRadiomics, LightGBM, SimpleITK, 6 other tools, structural MRI / diffusion, other condition, 2 references
[3] doi:10.1016/j.isci.2025.114100 [code]
Toward neuroimaging-based diagnostic support: a deep learning approach with a closed-loop system for psychiatric classification.
Journal: iScience
In common: PyRadiomics, LightGBM, SHAP, 7 other tools, 1 reference
[4] doi:10.1111/1759-7714.14386 [code]
A CT-based radiomics model to predict subsequent brain metastasis in patients with ALK-rearranged non-small cell lung cancer undergoing crizotinib treatment.
Journal: Thoracic cancer
In common: PyRadiomics, SimpleITK, seaborn, 5 other tools, other condition, 3 references
[5] doi:10.1038/s41598-022-20703-9 [code]
Robustness of radiomics to variations in segmentation methods in multimodal brain MRI.
Journal: Scientific reports
In common: PyRadiomics, LightGBM, NiBabel, 5 other tools, structural MRI / diffusion, other condition, 2 references
[6] doi:10.1093/noajnl/vdae060 [code]
Limited capability of MRI radiomics to predict primary tumor histology of brain metastases in external validation.
Journal: Neuro-oncology advances
In common: PyRadiomics, SimpleITK, seaborn, 5 other tools, structural MRI / diffusion, other condition, 2 references
[7] doi:10.1093/noajnl/vdag132 [code]
Spherical radiomics for radiogenomic assessment of glioblastoma heterogeneity.
Journal: Neuro-oncology advances
In common: PyRadiomics, SHAP, SimpleITK, 7 other tools, structural MRI / diffusion, other condition
[8] doi:10.1186/s12885-025-14454-9 [code]
Prediction of 1p/19q state in glioma by integrated deep learning method based on MRI radiomics.
Journal: BMC cancer
In common: PyRadiomics, SimpleITK, seaborn, 5 other tools, structural MRI / diffusion, other condition, 2 references
[9] doi:10.1016/j.nicl.2024.103707 [code]
Prediction of methylphenidate treatment response for ADHD using conventional and radiomics T1 and DTI features: Secondary analysis of a randomized clinical trial.
Journal: NeuroImage. Clinical
In common: PyRadiomics, SHAP, SimpleITK, 7 other tools, structural MRI / diffusion
[10] doi:10.3389/fpsyt.2024.1448145 [code]
Individualized multi-modal MRI biomarkers predict 1-year clinical outcome in first-episode drug-naïve schizophrenia patients.
Journal: Frontiers in psychiatry
In common: PyRadiomics, SHAP, SimpleITK, 7 other tools, structural MRI / diffusion

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Add the badge to the code's README

OSCR: the badge of this paper

If you are an author or a maintainer of this paper's code, you can add this badge. It says what OSCR knows of the paper (its code, its validated tracing map) and links the code to this page. Paste one of these snippets under the README's title. Keep the two marks around it, so that the badge can be replaced in place and is never added twice. OSCR changes nothing in your repository.

Markdown
<!-- oscr-badge:start -->
[![OSCR](https://openscicode.org/badge/doi/10.3389/fonc.2026.1816015.svg)](https://openscicode.org/badge/doi/10.3389/fonc.2026.1816015)
<!-- oscr-badge:end -->
reStructuredText
.. oscr-badge:start

.. image:: https://openscicode.org/badge/doi/10.3389/fonc.2026.1816015.svg
   :target: https://openscicode.org/badge/doi/10.3389/fonc.2026.1816015
   :alt: OSCR

.. oscr-badge:end
HTML
<!-- oscr-badge:start -->
<a href="https://openscicode.org/badge/doi/10.3389/fonc.2026.1816015"><img src="https://openscicode.org/badge/doi/10.3389/fonc.2026.1816015.svg" alt="OSCR"></a>
<!-- oscr-badge:end -->

On GitHub, open the README, choose Edit (the pencil), paste the snippet under the title, then "Commit changes".

If you choose to, you can have the badge kept up to date for you, with oscr badge --write from the command line or with the GitHub Action in your repository's workflows. Every other form and option (a card, AsciiDoc, LaTeX, shields.io) is on the badge page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: code errors and mismatches between the code and the paper about this paper and its code, from readers and its authors, beside the paper and its tracing map.

Report a mismatch between the code and the paper · Report a code error (signed in)

Reproductions: reports from readers who ran the authors' code and did not get the paper's result, what they ran, with which environment, commit and data.

Report a reproduction failure (signed in)

Versions of its code: the releases of its code tied to a version of this paper (the preprint, the accepted manuscript, the version of record…), each with the tracing map versioned with it.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.