Interpretative diagnostic model for neuroblastoma metastases using bone marrow cytology.
The 1 match
- [1] § Materials and methods › Construction and validation of diagnostic cytology model for BM metastasis in neuroblastoma ↔ feature_extraction/deep_learning_training/Model building and evaluation.ipynb, lines 286 to 325 · score 0.50 · NPV, PPV, thresholds, sensitivity, metric, WSI · 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
Jupyter notebook · 464 lines · 15 KB · no license · 1 match
- # %% [markdown]
- # %%
- # Model building and evaluation
- 1. data verification and review
- 2. data regularization which changes the data to obey N~(0, 1)
- 3. construct the training and test sets
- 4. features screening through Lasso and select non-zero items for the subsequent model
- 5. Use machine learning algorithms for clinical task
- 6. Model visualization
- # feature_file: the path of the feature data
- # label_file: label information file for each sample
- # labels: targets to be learned by the AI system
- import os
- from IPython.display import display
- os.environ['KMP_DUPLICATE_LIB_OK'] = 'TRUE'
- from onekey_algo import OnekeyDS as okds
- import pandas as pd
- os.makedirs('img', exist_ok=True)
- os.makedirs('results', exist_ok=True)
- os.makedirs('features', exist_ok=True)
- # file settings
- label_file = r'label information file for each sample'
- feature_file = r'the path of your feature data'
- labels = ['label']
- # %% [markdown]
- # %%
- # read the column name of labeled file
- # read data, the data is stored in CSV forma
- # the required label_data is a 'DataFrame' format including the ID column and the subsequent labels column which can be multiple columns
- feature_data = pd.read_csv(feature_file)
- display(feature_data)
- label_data = pd.read_csv(label_file)
- label_data.head()
- # %% [markdown]
- # %%
- # feature splicing
- from onekey_algo.custom.utils import print_join_info
- print_join_info(feature_data, label_data)
- combined_data = pd.merge(feature_data, label_data, on=['ID'], how='inner')
- ids = combined_data['ID']
- combined_data = combined_data.drop(['ID'], axis=1)
- print(combined_data[labels].value_counts())
- combined_data.columns
- # %% [markdown]
- # %%
- combined_data.describe()
- # %% [markdown]
- # #normalize_df,change the data to a mean of 0,a variance of 1
- #
- # $column = \frac{column - mean}{std}$
- # %%
- # regularization
- from onekey_algo.custom.components.comp1 import normalize_df
- data = normalize_df(combined_data, not_norm=labels, group='group')
- data = data.dropna(axis=1)
- data.describe()
- # %% [markdown]
- # %%
- # correlation coefficient,there are three methods to choose from for calculating the correlation coefficient
- 1. pearson: standard correlation coefficient
- 2. kendall: Kendall Tau correlation coefficient
- 3. spearman: Spearman rank correlation
- pearson_corr = data[data['group'] == 'train'][[c for c in data.columns if c not in labels]].corr('pearson')
- # kendall_corr = data[[c for c in data.columns if c not in labels]].corr('kendall')
- # spearman_corr = data[[c for c in data.columns if c not in labels]].corr('spearman')
- # %% [markdown]
- # %%
- # visualization of correlation coefficient
- import seaborn as sns
- import matplotlib.pyplot as plt
- from onekey_algo.custom.components.comp1 import draw_matrix
- if combined_data.shape[1] < 100:
- plt.figure(figsize=(50.0, 40.0))
- # select the correlation coefficient for visualization
- draw_matrix(pearson_corr, annot=True, cmap='YlGnBu', cbar=False)
- plt.savefig(f'img/feature_corr.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # cluster analysis
- import seaborn as sns
- import matplotlib.pyplot as plt
- if combined_data.shape[1] < 100:
- pp = sns.clustermap(pearson_corr, linewidths=.5, figsize=(50.0, 40.0), cmap='YlGnBu')
- plt.setp(pp.ax_heatmap.get_yticklabels(), rotation=0)
- plt.savefig(f'img/feature_cluster.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # feature filtering--correlation coefficient
- def select_feature(corr threshold: float = 0.9 keep: int = 1 topn=10 verbose=False):
- from onekey_algo.custom.components.comp1
- import select_feature
- sel_feature = select_feature(pearson_corr, threshold=0.9, topn=32, verbose=False)
- sel_feature = sel_feature + labels + ['group']
- sel_feature
- # %% [markdown]
- # %%
- # features screening
- sel_data = data[sel_feature]
- sel_data.describe()
- # %%
- # %%
- # construct datasets
- import numpy as np
- import onekey_algo.custom.components as okcomp
- n_classes = 2
- train_data = sel_data[(sel_data['group'] == 'train')]
- train_ids = ids[train_data.index]
- train_data = train_data.reset_index()
- train_data = train_data.drop('index', axis=1)
- y_data = train_data[labels]
- X_data = train_data.drop(labels + ['group'], axis=1)
- test_data = sel_data[sel_data['group'] != 'train']
- test_ids = ids[test_data.index]
- test_data = test_data.reset_index()
- test_data = test_data.drop('index', axis=1)
- y_test_data = test_data[labels]
- X_test_data = test_data.drop(labels + ['group'], axis=1)
- y_all_data = sel_data[labels]
- X_all_data = sel_data.drop(labels + ['group'], axis=1)
- column_names = X_data.columns
- print(f"sample size in the training set:{X_data.shape}, sample size in the validation set:{X_test_data.shape}")
- # %% [markdown]
- # %%
- # Lasso, initialize the Lasso model with alpha as the penalty coefficient
- # reference (https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.Lasso.html?highlight=lasso#sklearn.linear_model.Lasso)
- alpha = okcomp.comp1.lasso_cv_coefs(X_data, y_data, column_names=None, alpha_logmin=-3)
- plt.savefig(f'img/feature_lasso.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- okcomp.comp1.lasso_cv_efficiency(X_data, y_data, points=50, alpha_logmin=-3)
- plt.savefig(f'img/feature_mse_label.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # penalty factor, use the penalty factor of cross-validation as the basis for model training
- from sklearn import linear_model
- models = []
- for label in labels:
- clf = linear_model.Lasso(alpha=alpha)
- clf.fit(X_data, y_data[label])
- models.append(clf)
- # %% [markdown]
- # %%
- # feature screening, screened features with coef > 0 and print
- COEF_THRESHOLD = 1e-8 # feature thresholds after filtering
- scores = []
- selected_features = []
- for label, model in zip(labels, models):
- feat_coef = [(feat_name, coef) for feat_name, coef in zip(column_names, model.coef_)
- if COEF_THRESHOLD is None or abs(coef) > COEF_THRESHOLD]
- selected_features.append([feat for feat, _ in feat_coef])
- formula = ' '.join([f"{coef:+.6f} * {feat_name}" for feat_name, coef in feat_coef])
- score = f"{label} = {model.intercept_} {'+' if formula[0] != '-' else ''} {formula}"
- scores.append(score)
- print(scores[0])
- # %% [markdown]
- # %%
- # feature weights
- feat_coef = sorted(feat_coef, key=lambda x: x[1])
- feat_coef_df = pd.DataFrame(feat_coef, columns=['feature_name', 'Coefficients'])
- feat_coef_df.plot(x='feature_name', y='Coefficients', kind='barh')
- plt.savefig(f'img/feature_weights.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # screening features, use the features with high coefficients screened by Lasso as training data
- X_data = X_data[selected_features[0]]
- X_test_data = X_test_data[selected_features[0]]
- X_data.columns
- # %% [markdown]
- # %%
- # model selection
- model_names = ['ExtraTrees','RandomForest','LightGBM','AdaBoost', 'LR', 'MLP']
- models = okcomp.comp1.create_clf_model(model_names)
- model_names = list(models.keys())
- # %% [markdown]
- # %%
- # cross validation (if the study needed, no need in multicenterc ohort)
- # import seaborn as sns
- # import matplotlib.pyplot as plt
- # import pandas as pd
- # from sklearn.metrics import accuracy_score, roc_auc_score
- # results = okcomp.comp1.get_bst_split(X_data, y_data, models, test_size=0.2, metric_fn=roc_auc_score, cv=True, random_state=0)
- _, (X_train_sel, X_test_sel, y_train_sel, y_test_sel) = results['results'][results['max_idx']]
- # X_train_sel, X_test_sel, y_train_sel, y_test_sel = X_data, X_test_data, y_data, y_test_data
- # trails, _ = zip(*results['results'])
- # cv_results = pd.DataFrame(trails, columns=model_names)
- # sns.boxplot(data=cv_results)
- # plt.ylabel('AUC %')
- # plt.xlabel('Model Nmae')
- # plt.savefig(f'img/model_cv.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # model selection and evaluation
- import joblib
- from onekey_algo.custom.components.comp1 import plot_feature_importance, plot_learning_curve, smote_resample
- targets = []
- os.makedirs('models', exist_ok=True)
- for l in labels:
- new_models = list(okcomp.comp1.create_clf_model(model_names).values())
- for mn, m in zip(model_names, new_models):
- X_train_smote, y_train_smote = X_train_sel, y_train_sel
- # X_train_smote, y_train_smote = smote_resample(X_train_sel, y_train_sel)
- m.fit(X_train_smote, y_train_smote[l])
- # save result
- joblib.dump(m, f'models/{mn}_{l}.pkl')
- plot_feature_importance(m, selected_features[0], save_dir='img')
- # plot_learning_curve(m, X_train_sel, y_train_sel, title=f'Learning Curve {mn}')
- # plt.savefig(f"img/Rad_{mn}_learning_curve.svg", bbox_inches='tight')
- plt.show()
- targets.append(new_models)
- # %% [markdown]
- # %%
- # WSI-level predictions
- # predictions for prediction results of each model corresponding to each label.
- # pred_scores for predicted probability value for each model for each label.
- from sklearn.metrics import accuracy_score
- from sklearn.preprocessing import OneHotEncoder
- from onekey_algo.custom.components.delong import calc_95_CI
- from onekey_algo.custom.components.metrics import analysis_pred_binary
- predictions = [[(model.predict(X_train_sel), model.predict(X_test_sel))
- for model in target] for label, target in zip(labels, targets)]
- pred_scores = [[(model.predict_proba(X_train_sel), model.predict_proba(X_test_sel))
- for model in target] for label, target in zip(labels, targets)]
- metric = []
- pred_sel_idx = []
- for label, prediction, scores in zip(labels, predictions, pred_scores):
- pred_sel_idx_label = []
- for mname, (train_pred, test_pred), (train_score, test_score) in zip(model_names, prediction, scores):
- acc, auc, ci, tpr, tnr, ppv, npv, precision, recall, f1, thres = analysis_pred_binary(y_train_sel[label],
- train_score[:, 1])
- ci = f"{ci[0]:.4f} - {ci[1]:.4f}"
- metric.append((mname, acc, auc, ci, tpr, tnr, ppv, npv, precision, recall, f1, thres, f"{label}-train"))
- acc, auc, ci, tpr, tnr, ppv, npv, precision, recall, f1, thres = analysis_pred_binary(y_test_sel[label],
- test_score[:, 1])
- ci = f"{ci[0]:.4f} - {ci[1]:.4f}"
- metric.append((mname, acc, auc, ci, tpr, tnr, ppv, npv, precision, recall, f1, thres, f"{label}-test"))
- pred_sel_idx_label.append(np.logical_or(test_score[:, 0] >= thres, test_score[:, 1] >= thres))
- pred_sel_idx.append(pred_sel_idx_label)
- metric = pd.DataFrame(metric, index=None, columns=['model_name', 'Accuracy', 'AUC', '95% CI',
- 'Sensitivity', 'Specificity',
- 'PPV', 'NPV', 'Precision', 'Recall', 'F1',
- 'Threshold', 'Task'])
- metric
- # %% [markdown]
- # %%
- # draw curves
- import seaborn as sns
- plt.figure(figsize=(10, 10))
- plt.subplot(211)
- sns.barplot(x='model_name', y='Accuracy', data=metric, hue='Task')
- plt.subplot(212)
- sns.lineplot(x='model_name', y='Accuracy', data=metric, hue='Task')
- plt.savefig(f'img/model_acc.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # ROC
- sel_model = model_names
- for sm in sel_model:
- if sm in model_names:
- sel_model_idx = model_names.index(sm)
- # Plot all ROC curves
- plt.figure(figsize=(8, 8))
- for pred_score, label in zip(pred_scores, labels):
- okcomp.comp1.draw_roc([np.array(y_train_sel[label]), np.array(y_test_sel[label])],
- pred_score[sel_model_idx],
- labels=['Train', 'Test'], title=f"Model: {sm}")
- plt.savefig(f'img/model_{sm}_roc.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # model result
- sel_model = model_names
- for pred_score, label in zip(pred_scores, labels):
- pred_test_scores = []
- for sm in sel_model:
- if sm in model_names:
- sel_model_idx = model_names.index(sm)
- pred_test_scores.append(pred_score[sel_model_idx][1])
- okcomp.comp1.draw_roc([np.array(y_test_sel[label])] * len(pred_test_scores),
- pred_test_scores,
- labels=sel_model, title=f"Model AUC")
- plt.savefig(f'img/model_roc.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # DCA
- from onekey_algo.custom.components.comp1 import plot_DCA
- for pred_score, label in zip(pred_scores, labels):
- pred_test_scores = []
- for sm in sel_model:
- if sm in model_names:
- sel_model_idx = model_names.index(sm)
- okcomp.comp1.plot_DCA(pred_score[sel_model_idx][1][:,1], np.array(y_test_sel[label]),
- title=f'Rad Model {sm} DCA')
- plt.savefig(f'img/model_{sm}_dca.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # confusion matrix
- # set the drawing parameters
- sel_model = model_names
- c_matrix = {}
- for sm in sel_model:
- if sm in model_names:
- sel_model_idx = model_names.index(sm)
- for idx, label in enumerate(labels):
- cm = okcomp.comp1.calc_confusion_matrix(predictions[idx][sel_model_idx][-1], y_test_sel[label],
- # sel_idx = pred_sel_idx[idx][sel_model_idx],
- class_mapping={1:'1', 0:'0'}, num_classes=2)
- c_matrix[label] = cm
- plt.figure(figsize=(5, 4))
- plt.title(f'Rad Model:{sm}')
- okcomp.comp1.draw_matrix(cm, norm=False, annot=True, cmap='Blues', fmt='.3g')
- plt.savefig(f'img/model_{sm}_cm.svg', bbox_inches = 'tight')
- # %% [markdown]
- # %%
- # sample prediction histogram
- # plot the predicted results and the corresponding true results for each sample
- sel_model = model_names
- c_matrix = {}
- for sm in sel_model:
- if sm in model_names:
- sel_model_idx = model_names.index(sm)
- for idx, label in enumerate(labels):
- okcomp.comp1.draw_predict_score(pred_scores[idx][sel_model_idx][-1], y_test_sel[label])
- plt.title(f'{sm} sample predict score')
- plt.legend(labels=["label=0","label=1"],loc="lower right")
- plt.savefig(f'img/model_{sm}_sample_dis.svg', bbox_inches = 'tight')
- plt.show()
- # %% [markdown]
- # %% [markdown]
- # #save model result
- # import os
- # import numpy as np
- #
- # os.makedirs('results', exist_ok=True)
- # sel_model = sel_model
- #
- # for idx, label in enumerate(labels):
- # for sm in sel_model:
- # if sm in model_names:
- # sel_model_idx = model_names.index(sm)
- # target = targets[idx][sel_model_idx]
- # # sample prediction result
- # train_indexes = np.reshape(np.array(train_ids), (-1, 1)).astype(str)
- # test_indexes = np.reshape(np.array(test_ids), (-1, 1)).astype(str)
- # y_train_pred_scores = target.predict_proba(X_train_sel)
- # y_test_pred_scores = target.predict_proba(X_test_sel)
- # columns = ['ID'] + [f"{label}-{i}"for i in range(y_test_pred_scores.shape[1])]
- # # save results
- # result_train = pd.DataFrame(np.concatenate([train_indexes, y_train_pred_scores], axis=1), columns=columns)
- # result_train.to_csv(f'results/{sm}_Rad_train.csv', index=False)
- # result_test = pd.DataFrame(np.concatenate([test_indexes, y_test_pred_scores], axis=1), columns=columns)
- # result_test.to_csv(f'results/{sm}_Rad_test.csv', index=False)
- # %%
Model building and evaluation.ipynb at commit d1a39a2, no license · at the source
Overview
13 affiliations
- Department of Laboratory Medicine, Xin Hua Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, 200092 China
- Faculty of Medical Laboratory Science, Shanghai Jiao Tong University School of Medicine, Shanghai, 200025 China
- Institute of Artificial Intelligence Medicine, Shanghai Academy of Experimental Medicine, Shanghai, 200092 China
- Department of Laboratory Medicine, Shenzhen Children’s Hospital, Affiliated to Shantou University Medical College, Shenzhen, 518038 China
- National Clinical Trial Institute, Children’s Hospital, Zhejiang University School of Medicine, National Clinical Research Center for Children and Adolescents’ Health and Diseases, Hangzhou, Zhejiang 310052 China
- Research Center for Clinical Pharmacy, College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, Zhejiang 310058 China
- School of Communication and Information Engineering, Shanghai University, Shanghai, 200444 China
- Department of Pediatric Hematology/Oncology, Xin Hua Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, 200092 China
- Xinhua Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, 200092 China
- Shanghai Engineering Research Center of Intelligence Pediatrics (SERCIP), Shanghai, 200127 China
- Pediatric Cancer Research Center, Zhejiang University School of Medicine, National Clinical Research Center for Children and Adolescents’ Health and Diseases, Hangzhou, Zhejiang 310052 China
- Department of Surgical Oncology, Children’s Hospital, Zhejiang University School of Medicine, National Clinical Research Center for Children and Adolescents’ Health and Diseases, Hangzhou, Zhejiang 310052 China
- Department of Clinical Laboratory, Children’s Hospital, Zhejiang University School of Medicine, National Clinical Research Center for Children and Adolescents’ Health and Diseases, Hangzhou, Zhejiang 310052 China
Abstract
The abstract is not reproduced here: the paper's license (CC BY-NC-ND) does not allow it. Read it in the paper, at the publisher or on Europe PMC.
Repository
Its files are read in the Code ↔ Paper reader above, with 1 match between paragraphs and lines of code.
dr-jenna/cMIL
d1a39a23789e2f53cefcf05af2dc6b2fa59b42e1, 16 January 2025Availability: 1 check, the latest on 29 September 2026: the link answers
- 29 September 2026: the link answers
33 files
- feature_extraction/
deep_learning_training/ , Jupyter, 45 linesDataset splitting.ipynb - feature_extraction/
deep_learning_training/ , Jupyter, 464 lines, 1 matchModel building and evaluation.ipynb - feature_extraction/
deep_learning_training/ , Python, 362 linesModel building and evaluation.py - feature_extraction/
deep_learning_training/ , Jupyter, 47 linesMulti Instance Learning.ipynb - feature_extraction/
deep_learning_training/ , Python, 49 linesMulti Instance Learning.py - feature_extraction/
deep_learning_training/ , Jupyter, 83 linesPatch feature extraction.ipynb - feature_extraction/
deep_learning_training/ , Python, 50 linesPatch feature extraction.py - feature_extraction/
deep_learning_training/ , Jupyter, 45 linesPatch level prediction/ Dataset splitting.ipynb - feature_extraction/
deep_learning_training/ , Jupyter, 120 linesPatch level prediction/ What-feature.ipynb - feature_extraction/
onekey_core/ , Python, 30 lines__init__.py - feature_extraction/
onekey_core/ , Python, 13 linescore/ __init__.py - feature_extraction/
onekey_core/ , Python, 40 linescore/ image_loader.py - feature_extraction/
onekey_core/ , Python, 68 linescore/ losses_factory.py - feature_extraction/
onekey_core/ , Python, 35 linescore/ model_factory.py - feature_extraction/
onekey_core/ , Python, 59 linescore/ optimizer_lr_factory.py - feature_extraction/
onekey_core/ , Python, 58 linescore/ test_factory.py - feature_extraction/
onekey_core/ , Python, 31 linesextension.py - feature_extraction/
onekey_core/ , Python, 12 linesmodels/ __init__.py - feature_extraction/
onekey_core/ , Python, 63 linesmodels/ _utils.py - feature_extraction/
onekey_core/ , Python, 232 linesmodels/ res2net_v1b.py - feature_extraction/
onekey_core/ , Python, 395 linesmodels/ resnet.py - feature_extraction/
onekey_core/ , Python, 227 linesmodels/ resnet3d.py - feature_extraction/
onekey_core/ , Python, 4 linesmodels/ utils.py - feature_extraction/
onekey_core/ , Python, 1 linetransforms/ __init__.py - feature_extraction/
onekey_core/ , Python, 846 linestransforms/ functional.py - feature_extraction/
onekey_core/ , Python, 1,295 linestransforms/ transforms.py - feature_extraction/
onekey_core/ , Python, 107 linesutils.py - quality_supervision/
create_patch_dataset_csv , Python, 20 lines.py - quality_supervision/
patch_quality_supervisio , Python, 75 linesn.py - quality_supervision/
patch_reading.py , Python, 51 lines - quality_supervision/
quality_supervision_mode , Python, 38 linesl.py - LICENSE, License, 1 line
- README.md, Text, 147 lines
The paper's code and data availability statement is in the Data section.
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;
- 31 scripts, each with its path and the digest of its content;
- 1 match 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.
Code and data availability statement
The paper has a code and data availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:
- it points to the authors' code: dr-jenna/
cMIL - it says that the data are available on request
Read it in the paper: doi.org/10.1186/s12967-026-08069-2.
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, 29 September 2026: the first record
Recorded: type, language, journal, volume, issue, pages, dates, 15 authors, 6 keywords, 14 MeSH terms, 3 funders, 33 references.
Cite
This paper
Ma, J., Yao, Q., Fu, X., Xia, Z., Yue, Y., Xu, D., Yuan, X., Zhao, L., Wang, J., Dong, A., Gao, L., Yang, J., Shen, L., Zheng, Y., & Ni, S. (2026). Interpretative diagnostic model for neuroblastoma metastases using bone marrow cytology. Journal of translational medicine, 24(1), 691. https://
BibTeX
@article{ma2026interpret
author = {Ma, Juan and Yao, Qiang and Fu, Xiaoying and Xia, Zhouqi and Yue, Yaoting and Xu, Dongqing and Yuan, Xiaojun and Zhao, Liebin and Wang, Jinhu and Dong, Ao and Gao, Limei and Yang, Junyao and Shen, Lisong and Zheng, Yingxia and Ni, Shaoqing},
title = {{Interpretative diagnostic model for neuroblastoma metastases using bone marrow cytology}},
journal = {Journal of translational medicine},
year = {2026},
month = apr,
volume = {24},
number = {1},
pages = {691},
publisher = {BMC},
issn = {1479-5876},
doi = {10.1186/
url = {https://
pmid = {41947227},
pmcid = {PMC13188757}
}
RIS
TY - JOUR
AU - Ma, Juan
AU - Yao, Qiang
AU - Fu, Xiaoying
AU - Xia, Zhouqi
AU - Yue, Yaoting
AU - Xu, Dongqing
AU - Yuan, Xiaojun
AU - Zhao, Liebin
AU - Wang, Jinhu
AU - Dong, Ao
AU - Gao, Limei
AU - Yang, Junyao
AU - Shen, Lisong
AU - Zheng, Yingxia
AU - Ni, Shaoqing
TI - Interpretative diagnostic model for neuroblastoma metastases using bone marrow cytology
T2 - Journal of translational medicine
J2 - J Transl Med
PY - 2026
DA - 2026/
VL - 24
IS - 1
SP - 691
SN - 1479-5876
PB - BMC
DO - 10.1186/
UR - https://
LA - en
ER -
CSL-JSON
{
"id": "10.1186/
"type": "article-journal",
"title": "Interpretative diagnostic model for neuroblastoma metastases using bone marrow cytology",
"container-title": "Journal of translational medicine",
"author": [
{
"family": "Ma",
"given": "Juan"
},
{
"family": "Yao",
"given": "Qiang"
},
{
"family": "Fu",
"given": "Xiaoying"
},
{
"family": "Xia",
"given": "Zhouqi"
},
{
"family": "Yue",
"given": "Yaoting"
},
{
"family": "Xu",
"given": "Dongqing"
},
{
"family": "Yuan",
"given": "Xiaojun"
},
{
"family": "Zhao",
"given": "Liebin"
},
{
"family": "Wang",
"given": "Jinhu"
},
{
"family": "Dong",
"given": "Ao"
},
{
"family": "Gao",
"given": "Limei"
},
{
"family": "Yang",
"given": "Junyao"
},
{
"family": "Shen",
"given": "Lisong"
},
{
"family": "Zheng",
"given": "Yingxia"
},
{
"family": "Ni",
"given": "Shaoqing"
}
],
"container-title-short":
"volume": "24",
"issue": "1",
"page": "691",
"DOI": "10.1186/
"PMID": "41947227",
"PMCID": "PMC13188757",
"ISSN": "1479-5876",
"publisher": "BMC",
"URL": "https://
"language": "en",
"issued": {
"date-parts": [
[
2026,
4,
7
]
]
}
}
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:10.1038/s41598-025-26004-1 [code]
- Human visual attention-inspired knowledge distillation underlying interpretable computational pathology.Journal: Scientific reportsIn common: Pillow, PyTorch, pandas, 2 other tools, other condition, 2 references
- [2] doi:10.1016/j.xgen.2024.100739 [code]
- CSI-GEP: A GPU-based unsupervised machine learning approach for recovering gene expression programs in atlas-scale single-cell RNA-seq data.Journal: Cell genomicsIn common: Pillow, PyTorch, seaborn, 4 other tools, 1 reference
- [3] doi:10.1038/s41598-026-40620-5 [code]
- mViSE: A visual search engine for analyzing multiplex IHC brain tissue images (spatial proteomics).Journal: Scientific reportsIn common: Pillow, PyTorch, seaborn, 4 other tools, 1 reference
- [4] doi:10.1093/gigascience/giae057 [code]
- Deep learning links localized digital pathology phenotypes with transcriptional subtype and patient outcome in glioblastoma.Journal: GigaScienceIn common: Pillow, seaborn, scikit-learn, 3 other tools, clinical / translational, other condition, 1 reference
- [5] doi:10.1093/jnen/nlae120 [code]
- A machine learning approach to automate microinfarct and microhemorrhage screening in hematoxylin and eosin-stained human brain tissues.Journal: Journal of neuropathology and experimental neurologyIn common: Pillow, PyTorch, scikit-learn, 3 other tools, clinical / translational, 1 reference
- [6] doi:10.1111/bpa.70029 [code]
- Pediatric brain tumor classification using digital pathology and deep learning: Evaluation of SOTA methods on a multi-center Swedish cohort.Journal: Brain pathology (Zurich, Switzerland)In common: Pillow, PyTorch, scikit-learn, 3 other tools, other condition, 1 reference
- [7] doi:10.1109/access.2025.3599857 [code]
- ADAM-1: An AI Reasoning and Bioinformatics Model for Alzheimer's Disease Detection and Microbiome-Clinical Data Integration.Journal: IEEE access : practical innovations, open solutionsIn common: PyTorch, seaborn, scikit-learn, 3 other tools, clinical / translational, 1 reference
- [8] doi:10.1038/s41598-026-48666-1 [code]
- Multimodal survival analysis of glioblastoma using whole-slide histopathology, gene expression, clinical variables and language-model-derived mutation features.Journal: Scientific reportsIn common: PyTorch, scikit-learn, pandas, 2 other tools, clinical / translational, other condition, 1 reference
- [9] doi:10.1038/s41598-025-05009-w [code]
- Lessons learned from RadiologyNET foundation models for transfer learning in medical radiology.Journal: Scientific reportsIn common: PyTorch, scikit-learn, pandas, 2 other tools, other condition, 1 reference
- [10] doi:10.1038/s41467-024-47945-7 [code]
- A human neural crest model reveals the developmental impact of neuroblastoma-associated
chromosomal aberrations. Journal: Nature communicationsIn common: pandas, other condition, 2 references
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.
Claim this paper
Correct its record
Say what each link of this record is, remove the ones that are not the paper's, add the ones that are missing. The correction becomes a new version of the record, in its Versions section.
Validate its tracing map
You validate the map as this page shows it: 1 repository of the authors' code, each at its verified commit and with its license, 31 scripts, and 1 match between paragraphs and code (see the Code and Map sections). It then receives a DOI on Zenodo, with you (your ORCID iD) and OSCR as its creators; the code itself is not deposited.
The map's fingerprint: sha256:a564eafad6074cf6…
Add the badge to the code's README
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 -->
[](https://openscicode.org/badge/doi/10.1186/s12967-026-08069-2)
<!-- oscr-badge:end -->reStructuredText
.. oscr-badge:start
.. image:: https://openscicode.org/badge/doi/10.1186/s12967-026-08069-2.svg
:target: https://openscicode.org/badge/doi/10.1186/s12967-026-08069-2
:alt: OSCR
.. oscr-badge:endHTML
<!-- oscr-badge:start -->
<a href="https://openscicode.org/badge/doi/10.1186/s12967-026-08069-2"><img src="https://openscicode.org/badge/doi/10.1186/s12967-026-08069-2.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.
