Quantitative CT radiomics-based models for prediction of haematoma expansion and poor functional outcome in primary intracerebral haemorrhage

Enhancing Outcome Prediction in Intracerebral Hemorrhage Through Deep Learning: A Retrospective Multicenter Study

RATIONALE AND OBJECTIVES

This study aimed to employ deep learning techniques to analyze and validate an automatic prognostic biomarker for predicting outcomes following intracerebral hemorrhage (ICH).

MATERIALS AND METHODS

This study included patients with ICH whose onset-to-imaging time (OIT) was less than 6 h. Patients were randomly divided into training and test sets at a 7:3 ratio. Using the Resnet50 deep learning method, we extracted features from the hematoma and perihematomal edema (PHE) areas and constructed a 90-day prognosis prediction model using logistic regression. To evaluate predictive efficacy and clinical significance, we employed logistic regression to train three models: Clinical, Deep Score, and the combined Clinical-Deep Learning (Merge).

RESULTS

Our study comprised 1098 patients (652 male, 446 female), with a mean Glasgow Coma Scale (GCS) score of 10. Univariate and multivariate analyses identified age, intraventricular hemorrhage (IVH), hematoma and PHE volume, and admission GCS score as independent prognostic factors. Additionally, 15 deep learning features were retained through LASSO regression. In the training set, the AUC values for the three models were as follows: Clinical model (0.88), Deep Score (0.91), and Merge model (0.94). In the test set, the Merge model exhibited a significantly higher AUC value than the other models. Calibration curves revealed satisfactory calibration of the Merge model nomogram in both training and test sets.

CONCLUSION

Our Merge model nomogram is an objective and effective prognostic tool, offering personalized risk assessments for 90-day functional outcomes in patients with ICH.

CT radiomics combined with clinical and radiological factors predict hematoma expansion in hypertensive intracerebral hemorrhage

OBJECTIVES

This study aimed to establish a hematoma expansion (HE) prediction model for hypertensive intracerebral hemorrhage (HICH) patients by combining CT radiomics, clinical information, and conventional imaging signs.

METHODS

A retrospective continuous collection of HICH patients from three medical centers was divided into a training set (n = 555), a validation set (n = 239), and a test set (n = 77). Extract radiomics features from baseline CT plain scan images and combine them with clinical information and conventional imaging signs to construct radiomics models, clinical imaging sign models, and hybrid models, respectively. The models will be evaluated using the area under the curve (AUC), clinical decision curve analysis (DCA), net reclassification index (NRI), and integrated discrimination improvement (IDI).

RESULTS

In the training, validation, and testing sets, the radiomics model predicts an AUC of HE of 0.885, 0.827, and 0.894, respectively, while the clinical imaging sign model predicts an AUC of HE of 0.759, 0.725, and 0.765, respectively. Glasgow coma scale score at admission, first CT hematoma volume, irregular hematoma shape, and radiomics score were used to construct a hybrid model, with AUCs of 0.901, 0.838, and 0.917, respectively. The DCA shows that the hybrid model had the highest net profit rate. Compared with the radiomics model and the clinical imaging sign model, the hybrid model showed an increase in NRI and IDI.

CONCLUSION

The hybrid model based on CT radiomics combined with clinical and radiological factors can effectively individualize the evaluation of the risk of HE in patients with HICH.

CLINICAL RELEVANCE STATEMENT

CT radiomics combined with clinical information and conventional imaging signs can identify HICH patients with a high risk of HE and provide a basis for clinical-targeted treatment.

KEY POINTS

HE is an important prognostic factor in patients with HICH. The hybrid model predicted HE with training, validation, and test AUCs of 0.901, 0.838, and 0.917, respectively. This model provides a tool for a personalized clinical assessment of early HE risk.

CT-based deep learning model for predicting hospital discharge outcome in spontaneous intracerebral hemorrhage

OBJECTIVES

To predict the functional outcome of patients with intracerebral hemorrhage (ICH) using deep learning models based on computed tomography (CT) images.

METHODS

A retrospective, bi-center study of ICH patients was conducted. Firstly, a custom 3D convolutional model was built for predicting the functional outcome of ICH patients based on CT scans from randomly selected ICH patients in H training dataset collected from H hospital. Secondly, clinical data and radiological features were collected at admission and the Extreme Gradient Boosting (XGBoost) algorithm was used to establish a second model, named the XGBoost model. Finally, the Convolution model and XGBoost model were fused to build the third "Fusion model." Favorable outcome was defined as modified Rankin Scale score of 0-3 at discharge. The prognostic predictive accuracy of the three models was evaluated using an H test dataset and an external Y dataset, and compared with the performance of ICH score and ICH grading scale (ICH-GS).

RESULTS

A total of 604 patients with ICH were included in this study, of which 450 patients were in the H training dataset, 50 patients in the H test dataset, and 104 patients in the Y dataset. In the Y dataset, the areas under the curve (AUCs) of the Convolution model, XGBoost model, and Fusion model were 0.829, 0.871, and 0.905, respectively. The Fusion model prognostic performance exceeded that of ICH score and ICH-GS (p = 0.043 and p = 0.045, respectively).

CONCLUSIONS

Deep learning models have good accuracy for predicting functional outcome of patients with spontaneous intracerebral hemorrhage.

CLINICAL RELEVANCE STATEMENT

The proposed deep learning Fusion model may assist clinicians in predicting functional outcome and developing treatment strategies, thereby improving the survival and quality of life of patients with spontaneous intracerebral hemorrhage.

KEY POINTS

• Integrating clinical presentations, CT images, and radiological features to establish deep learning model for functional outcome prediction of patients with intracerebral hemorrhage. • Deep learning applied to CT images provides great help in prognosing functional outcome of intracerebral hemorrhage patients. • The developed deep learning model performs better than clinical prognostic scores in predicting functional outcome of patients with intracerebral hemorrhage.

Quantitative imaging for predicting hematoma expansion in intracerebral hemorrhage: A multimodel comparison

BACKGROUND

Several studies report that radiomics provides additional information for predicting hematoma expansion in intracerebral hemorrhage (ICH). However, the comparison of diagnostic performance of radiomics for predicting revised hematoma expansion (RHE) remains unclear.

METHODS

The cohort comprised 312 consecutive patients with ICH. A total of 1106 radiomics features from seven categories were extracted using Python software. Support vector machines achieved the best performance in both the training and validation datasets. Clinical factors models were constructed to predict RHE. Receiver operating characteristic curve analysis was used to assess the abilities of non-contrast computed tomography (NCCT) signs, radiomics features, and combined models to predict RHE.

RESULTS

We finally selected the top 21 features for predicting RHE. After univariate analysis, 4 clinical factors and 5 NCCT signs were selected for inclusion in the prediction models. In the training and validation dataset, radiomics features had a higher predictive value for RHE (AUC = 0.83) than a single NCCT sign and expansion-prone hematoma. The combined prediction model including radiomics features, clinical factors, and NCCT signs achieved higher predictive performances for RHE (AUC = 0.88) than other combined models.

CONCLUSIONS

NCCT radiomics features have a good degree of discrimination for predicting RHE in ICH patients. Combined prediction models that include quantitative imaging significantly improve the prediction of RHE, which may assist in the risk stratification of ICH patients for anti-expansion treatments.

Machine learning-based radiomics for predicting outcomes in cervical cancer patients undergoing concurrent chemoradiotherapy

PURPOSES

To investigate the value of machine learning-based radiomics for predicting disease-free survival (DFS) and overall survival (OS) undergoing concurrent chemoradiotherapy (CCRT) for patients with locally advanced cervical cancer (LACC).

MATERIALS AND METHODS

In this multicentre study, 700 patients with IB2-IVA cervical cancer who underwent CCRT with ongoing follow-up were retrospectively analyzed. Three-dimensional radiomics features of primary lesions and its surrounding 5 mm region in T2WI sequences were collected. Six machine learning methods were used to construct the optimal radiomics model for accurate prediction of DFS and OS after CCRT in LACC patients. Eventually, TCGA and GEO databases were used to explore the mechanisms of radiomics in predicting the progression and survival of cervical cancer. This study adhered CLEAR for reporting and its quality was assessed using RQS and METRICS.

RESULTS

In the prediction of DFS, the RSF model combined tumor and peritumor radiomics demonstrated the best predictive efficacy, with the AUC for predicting 1-year, 3-year, and 5-year DFS in the training, validation, and test sets of 0.986, 0.989, 0.990, and 0.884, 0.838, 0.823, and 0.829, 0.809, 0.841, respectively. In the prediction of OS, the GBM model best performer, with AUC of 0.999, 0.995, 0.978, and 0.981, 0.975, 0.837, and 0.904, 0.860, 0.905. Differential genes in TCGA and GEO suggest that the prediction of radiomics model may be associated with KDELR2 and HK2.

CONCLUSION

Machine learning-based radiomics models help to predict DFS and OS after CCRT in LACC patients, and the combination of tumor and peritumor information has higher predictive efficacy, which can provide a reliable basis for therapeutic decision-making in cervical cancer patients.

HE-Mind: A model for automatically predicting hematoma expansion after spontaneous intracerebral hemorrhage

PURPOSE

To develop and validate an end-to-end model for automatically predicting hematoma expansion (HE) after spontaneous intracerebral hemorrhage (sICH) using a novel deep learning framework.

METHODS

This multicenter retrospective study collected cranial noncontrast computed tomography (NCCT) images of 490 patients with sICH at admission for model training (n = 236), internal testing (n = 60), and external testing (n = 194). A HE-Mind model was designed to predict HE, which consists of a densely connected U-net for segmentation process, a multi-instance learning strategy for resolving label ambiguity and a Siamese network for classification process. Two radiomics models based on support vector machine or logistic regression and two deep learning models based on residual network or Swin transformer were developed for performance comparison. Reader experiments including physician diagnosis mode and artificial intelligence mode were conducted for efficiency comparison.

RESULTS

The HE-Mind model showed better performance compared to the comparative models in predicting HE, with areas under the curve of 0.849 and 0.809 in the internal and external test sets respectively. With the assistance of the HE-Mind model, the predictive accuracy and work efficiency of the emergency physician, junior radiologist, and senior radiologist were significantly improved, with accuracies of 0.768, 0.789, and 0.809 respectively, and reporting times of 7.26 s, 5.08 s, and 3.99 s respectively.

CONCLUSIONS

The HE-Mind model could rapidly and automatically process the NCCT data and predict HE after sICH within three seconds, indicating its potential to assist physicians in the clinical diagnosis workflow of HE.

Noninvasive prediction of lymph node metastasis in pancreatic cancer using an ultrasound-based clinicoradiomics machine learning model

OBJECTIVES

This study was designed to explore and validate the value of different machine learning models based on ultrasound image-omics features in the preoperative diagnosis of lymph node metastasis in pancreatic cancer (PC).

METHODS

This research involved 189 individuals diagnosed with PC confirmed by surgical pathology (training cohort: n = 151; test cohort: n = 38), including 50 cases of lymph node metastasis. Image-omics features were extracted from ultrasound images. After dimensionality reduction and screening, eight machine learning algorithms, including logistic regression (LR), support vector machine (SVM), K-nearest neighbors (KNN), random forest (RF), extra trees (ET), extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), and multilayer perceptron (MLP), were used to establish image-omics models to predict lymph node metastasis in PC. The best omics prediction model was selected through ROC curve analysis. Machine learning models were used to analyze clinical features and determine variables to establish a clinical model. A combined model was constructed by combining ultrasound image-omics and clinical features. Decision curve analysis (DCA) and a nomogram were used to evaluate the clinical application value of the model.

RESULTS

A total of 1561 image-omics features were extracted from ultrasound images. 15 valuable image-omics features were determined by regularization, dimension reduction, and algorithm selection. In the image-omics model, the LR model showed higher prediction efficiency and robustness, with an area under the ROC curve (AUC) of 0.773 in the training set and an AUC of 0.850 in the test set. The clinical model constructed by the boundary of lesions in ultrasound images and the clinical feature CA199 (AUC = 0.875). The combined model had the best prediction performance, with an AUC of 0.872 in the training set and 0.918 in the test set. The combined model showed better clinical benefit according to DCA, and the nomogram score provided clinical prediction solutions.

CONCLUSION

The combined model established with clinical features has good diagnostic ability and can be used to predict lymph node metastasis in patients with PC. It is expected to provide an effective noninvasive method for clinical decision-making, thereby improving the diagnosis and treatment of PC.

Machine learning for predicting hematoma expansion in spontaneous intracerebral hemorrhage: a systematic review and meta-analysis

PURPOSE

Early identification of hematoma enlargement and persistent hematoma expansion (HE) in patients with cerebral hemorrhage is increasingly crucial for determining clinical treatments. However, due to the lack of clinically effective tools, radiomics has been gradually introduced into the early identification of hematoma enlargement. Though, radiomics has limited predictive accuracy due to variations in procedures. Therefore, we conducted a systematic review and meta-analysis to explore the value of radiomics in the early detection of HE in patients with cerebral hemorrhage.

METHODS

Eligible studies were systematically searched in PubMed, Embase, Cochrane and Web of Science from inception to April 8, 2024. English articles are considered eligible. The radiomics quality scoring (RQS) tool was used to evaluate included studies.

RESULTS

A total of 34 studies were identified with sample sizes ranging from 108 to 3016. Eleven types of models were involved, and the types of modeling contained mainly clinical, radiomic, and radiomic plus clinical features. The radiomics models seem to have better performance (0.77 and 0.73 C-index in the training cohort and validation cohort, respectively) than the clinical models (0.69 C-index in the training cohort and 0.70 C-index in the validation cohort) in discriminating HE. However, the C-index was the highest for the combined model in both the training (0.82) and validation (0.79) cohorts.

CONCLUSIONS

Machine learning based on radiomic plus clinical features has the best predictive performance for HE, followed by machine learning based on radiomic features, and can be used as a potential tool to assist clinicians in early judgment.

Research advances in predicting the expansion of hypertensive intracerebral hemorrhage based on CT images: an overview

Hematoma expansion (HE) is an important risk factor for death or poor prognosis in patients with hypertensive intracerebral hemorrhage (HICH). Accurately predicting the risk of HE in patients with HICH is of great clinical significance for timely intervention and improving patient prognosis. Many imaging signs reported in literatures showed the important clinical value for predicting HE. In recent years, the development of radiomics and artificial intelligence has provided new methods for HE prediction with high accuracy. Therefore, this article reviews the latest research progress in CT imaging, radiomics, and artificial intelligence of HE, in order to help identify high-risk patients for HE in clinical practice.

An Online Dynamic Radiomics-Clinical Nomogram to Predict Recurrence in Patients with Spontaneous Intracerebral Hemorrhage

OBJECTIVE

Radiomics can reflect the heterogeneity within the focus. We aim to explore whether radiomics can predict recurrent intracerebral hemorrhage (RICH) and develop an online dynamic nomogram to predict it.

METHODS

This retrospective study collected the clinical and radiomics features of patients with spontaneous intracerebral hemorrhage seen in our hospital from October 2013 to October 2016. We used the minimum redundancy maximum relevancy and the least absolute shrinkage and selection operator methods to screen radiomics features and calculate the Rad-score. We use the univariate and multivariate analyses to screen clinical predictors. Optimal clinical features and Rad-score were used to construct different logistics regression models called the clinical model, radiomics model, and combined-logistic regression model. DeLong testing was performed to compare performance among different models. The model with the best predictive performance was used to construct an online dynamic nomogram.

RESULTS

Overall, 304 patients with intracerebral hemorrhage were enrolled in this study. Fourteen radiomics features were selected to calculate the Rad-score. The patients with RICH had a significantly higher Rad-score than those without (0.5 vs. -0.8; P< 0.001). The predictive performance of the combined-logistic regression model with Rad-score was better than that of the clinical model for both the training (area under the receiver operating curve, 0.81 vs. 0.71; P = 0.02) and testing (area under the receiver operating curve, 0.65 vs. 0.58; P = 0.04) cohorts statistically.

CONCLUSIONS

Radiomics features were determined related to RICH. Adding Rad-score into conventional clinical models significantly improves the prediction efficiency. We developed an online dynamic nomogram to accurately and conveniently evaluate RICH.

Machine-learning-based performance comparison of two-dimensional (2D) and three-dimensional (3D) CT radiomics features for intracerebral haemorrhage expansion

AIM

To investigate the value of non-contrast CT (NCCT)-based two-dimensional (2D) radiomics features in predicting haematoma expansion (HE) after spontaneous intracerebral haemorrhage (ICH) and compare its predictive ability with the three-dimensional (3D) signature.

MATERIALS AND METHODS

Three hundred and seven ICH patients who received baseline NCCT within 6 h of ictus from two stroke centres were analysed retrospectively. 2D and 3D radiomics features were extracted in the manner of one-to-one correspondence. The 2D and 3D models were generated by four different machine-learning algorithms (regularised L1 logistic regression, decision tree, support vector machine and AdaBoost), and the receiver operating characteristic (ROC) curve was used to compare their predictive performance. A robustness analysis was performed according to baseline haematoma volume.

RESULTS

Each feature type of 2D and 3D modalities used for subsequent analyses had excellent consistency (mean ICC >0.9). Among the different machine-learning algorithms, pairwise comparison showed no significant difference in both the training (mean area under the ROC curve [AUC] 0.858 versus 0.802, all p>0.05) and validation datasets (mean AUC 0.725 versus 0.678, all p>0.05), and the 10-fold cross-validation evaluation yielded similar results. The AUCs of the 2D and 3D models were comparable either in the binary or tertile volume analysis (all p>0.5).

CONCLUSION

NCCT-derived 2D radiomics features exhibited acceptable and similar performance to the 3D features in predicting HE, and this comparability seemed unaffected by initial haematoma volume. The 2D signature may be preferred in future HE-related radiomic works given its compatibility with emergency condition of ICH.

Clinical Features, Non-Contrast CT Radiomic and Radiological Signs in Models for the Prediction of Hematoma Expansion in Intracerebral Hemorrhage

PURPOSE

Rapid identification of hematoma expansion (HE) risk at baseline is a priority in intracerebral hemorrhage (ICH) patients and may impact clinical decision making. Predictive scores using clinical features and Non-Contract Computed Tomography (NCCT)-based features exist, however, the extent to which each feature set contributes to identification is limited. This paper aims to investigate the relative value of clinical, radiological, and radiomics features in HE prediction.

METHODS

Original data was retrospectively obtained from three major prospective clinical trials ["Spot Sign" Selection of Intracerebral Hemorrhage to Guide Hemostatic Therapy (SPOTLIGHT)NCT01359202; The Spot Sign for Predicting and Treating ICH Growth Study (STOP-IT)NCT00810888] Patients baseline and follow-up scans following ICH were included. Clinical, NCCT radiological, and radiomics features were extracted, and multivariate modeling was conducted on each feature set.

RESULTS

317 patients from 38 sites met inclusion criteria. Warfarin use (p=0.001) and GCS score (p=0.046) were significant clinical predictors of HE. The best performing model for HE prediction included clinical, radiological, and radiomic features with an area under the curve (AUC) of 87.7%. NCCT radiological features improved upon clinical benchmark model AUC by 6.5% and a clinical & radiomic combination model by 6.4%. Addition of radiomics features improved goodness of fit of both clinical (p=0.012) and clinical & NCCT radiological (p=0.007) models, with marginal improvements on AUC. Inclusion of NCCT radiological signs was best for ruling out HE whereas the radiomic features were best for ruling in HE.

CONCLUSION

NCCT-based radiological and radiomics features can improve HE prediction when added to clinical features.

Radiomic-based nonlinear supervised learning classifiers on non-contrast CT to predict functional prognosis in patients with spontaneous intracerebral hematoma

PURPOSE

To evaluate if nonlinear supervised learning classifiers based on non-contrast CT can predict functional prognosis at discharge in patients with spontaneous intracerebral hematoma.

METHODS

Retrospective, single-center, observational analysis of patients with a diagnosis of spontaneous intracerebral hematoma confirmed by non-contrast CT between January 2016 and April 2018. Patients with HIE > 18 years and with TCCSC performed within the first 24 h of symptom onset were included. Patients with secondary spontaneous intracerebral hematoma and in whom radiomic variables were not available were excluded. Clinical, demographic and admission variables were collected. Patients were classified according to the Modified Rankin Scale (mRS) at discharge into good (mRS 0-2) and poor prognosis (mRS 3-6). After manual segmentation of each spontaneous intracerebral hematoma, the radiomics variables were obtained. The sample was divided into a training and testing cohort and a validation cohort (70-30% respectively). Different methods of variable selection and dimensionality reduction were used, and different algorithms were used for model construction. Stratified 10-fold cross-validation were performed on the training and testing cohort and the mean area under the curve (AUC) were calculated. Once the models were trained, the sensitivity of each was calculated to predict functional prognosis at discharge in the validation cohort.

RESULTS

105 patients with spontaneous intracerebral hematoma were analyzed. 105 radiomic variables were evaluated for each patient. P-SVM, KNN-E and RF-10 algorithms, in combination with the ANOVA variable selection method, were the best performing classifiers in the training and testing cohort (AUC 0.798, 0.752 and 0.742 respectively). The predictions of these models, in the validation cohort, had a sensitivity of 0.897 (0.778-1;95%CI), with a false-negative rate of 0% for predicting poor functional prognosis at discharge.

CONCLUSION

The use of radiomics-based nonlinear supervised learning classifiers are a promising diagnostic tool for predicting functional outcome at discharge in HIE patients, with a low false negative rate, although larger and balanced samples are still needed to develop and improve their performance.

Development and validation of a novel clinical prediction model to predict the 90-day functional outcome of spontaneous intracerebral hemorrhage

Background

Spontaneous intracerebral hemorrhage (SICH) is associated with high mortality and disability. Accurately predicting adverse prognostic risks of SICH is helpful in developing risk stratification and precision medicine strategies for this phenomenon.

Methods

We analyzed 413 patients with SICH admitted to Hefei Second People's Hospital as a training cohort, considering 74 patients from the First Affiliated Hospital of Anhui Medical University for external validation. Univariate and multivariate logistic regression analyses were used to select risk factors for 90-day functional outcomes, and a nomogram was developed to predict their incidence in patients. Discrimination, fitting performance, and clinical utility of the resulting nomogram were evaluated through receiver operating characteristic (ROC) curves, accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), calibration plots, and decision curves analysis (DCA), respectively.

Results

Of the 413 patients, 180 had a poor prognosis. Univariate analysis showed significant variance of age, systolic pressure, intraventricular hemorrhage (IVH), Glasgow Coma Scale (GCS) scores, National Institute of Health Stroke Scale (NIHSS) scores, and hematoma volume between the groups (p < 0.05). Logistic multivariate regression analysis showed that age, IVH, NIHSS, and hematoma volume were associated with unfavorable outcomes. Based on the results, a nomogram model was developed with an area under the ROC curve of 0.91 (95% CI; 0.88-0.94) and 0.89 (95% CI; 0.80-0.95) in the training and validation sets, respectively. In the validation set, the accuracy, sensitivity, specificity, PPV, and NPV of the model were 0.851, 0.923, 0.812, 0.727, and 0.951, respectively. The calibration plot demonstrates the goodness of fit between the nomogram predictions and actual observations. Finally, DCA indicated significant clinical adaptability.

Conclusion

We developed and validated a short-term prognostic nomogram model for patients with SICH including NIHSS scores, age, hematoma volume, and IVH. This model has valuable potential in predicting the prognosis of patients with SICH.

Perihematomal edema-based CT-radiomics model to predict functional outcome in patients with intracerebral hemorrhage

PURPOSE

The purpose of this study was to identify possible association between noncontrast computed tomography (NCCT)-based radiomics features of perihematomal edema (PHE) and poor functional outcome at 90 days after intracerebral hemorrhage (ICH) and to develop a NCCT-based radiomics-clinical nomogram to predict 90-day functional outcomes in patients with ICH.

MATERIALS AND METHODS

In this multicenter retrospective study, 107 radiomics features were extracted from 1098 NCCT examinations obtained in 1098 patients with ICH. There were 652 men and 446 women with a mean age of 60 ± 12 (SD) years (range: 23-95 years). After harmonized and univariable and multivariable screening, seven of these radiomics features were closely associated with the 90-day functional outcome of patients with ICH. The radiomics score (Rad-score) was calculated based on the seven radiomics features. A clinical-radiomics nomogram was developed and validated in three cohorts. The model performance was evaluated using area under the curve analysis and decision and calibration curves.

RESULTS

Of the 1098 patients with ICH, 395 had a good outcome at 90 days. Hematoma hypodensity sign and intraventricular and subarachnoid hemorrhages were identified as risk factors for poor outcomes (P < 0.001). Age, Glasgow coma scale score, and Rad-score were independently associated with outcome. The clinical-radiomics nomogram showed good predictive performance with AUCs of 0.882 (95% CI: 0.859-0.905), 0.834 (95% CI: 0.776-0.891) and 0.905 (95% CI: 0.839-0.970) in the three cohorts and clinical applicability.

CONCLUSION

NCCT-based radiomics features from PHE are highly correlated with outcome. When combined with Rad-score, radiomics features from PHE can improve the predictive performance for 90-day poor outcome in patients with ICH.

Radiomic markers of intracerebral hemorrhage expansion on non-contrast CT: independent validation and comparison with visual markers

Objective

To devise and validate radiomic signatures of impending hematoma expansion (HE) based on admission non-contrast head computed tomography (CT) of patients with intracerebral hemorrhage (ICH).

Methods

Utilizing a large multicentric clinical trial dataset of hypertensive patients with spontaneous supratentorial ICH, we developed signatures predictive of HE in a discovery cohort (n = 449) and confirmed their performance in an independent validation cohort (n = 448). In addition to n = 1,130 radiomic features, n = 6 clinical variables associated with HE, n = 8 previously defined visual markers of HE, the BAT score, and combinations thereof served as candidate variable sets for signatures. The area under the receiver operating characteristic curve (AUC) quantified signatures' performance.

Results

A signature combining select radiomic features and clinical variables attained the highest AUC (95% confidence interval) of 0.67 (0.61-0.72) and 0.64 (0.59-0.70) in the discovery and independent validation cohort, respectively, significantly outperforming the clinical (pdiscovery = 0.02, pvalidation = 0.01) and visual signature (pdiscovery = 0.03, pvalidation = 0.01) as well as the BAT score (pdiscovery < 0.001, pvalidation < 0.001). Adding visual markers to radiomic features failed to improve prediction performance. All signatures were significantly (p < 0.001) correlated with functional outcome at 3-months, underlining their prognostic relevance.

Conclusion

Radiomic features of ICH on admission non-contrast head CT can predict impending HE with stable generalizability; and combining radiomic with clinical predictors yielded the highest predictive value. By enabling selective anti-expansion treatment of patients at elevated risk of HE in future clinical trials, the proposed markers may increase therapeutic efficacy, and ultimately improve outcomes.

Radiomics Outperforms Clinical and Radiologic Signs in Predicting Spontaneous Basal Ganglia Hematoma Expansion: A Pilot Study

Prediction of the hematoma expansion (HE) of spontaneous basal ganglia hematoma (SBH) from the first non-contrast CT can result in better management, which has the potential of improving outcomes. This study has been designed to compare the performance of "Radiomics analysis," "radiology signs," and "clinical-laboratory data" for this task. We retrospectively reviewed the electronic medical records for clinical, demographic, and laboratory data in patients with SBH. CT images were reviewed for the presence of radiologic signs, including black-hole, blend, swirl, satellite, and island signs. Radiomic features from the SBH on the first brain CT were extracted, and the most predictive features were selected. Different machine learning models were developed based on clinical, laboratory, and radiology signs and selected Radiomic features to predict hematoma expansion (HE). The dataset used for this analysis included 116 patients with SBH. Among different models and different thresholds to define hematoma expansion (10%, 20%, 25%, 33%, 40%, and 50% volume enlargement thresholds), the Random Forest based on 10 selected Radiomic features achieved the best performance (for 25% hematoma enlargement) with an area under the curve (AUC) of 0.9 on the training dataset and 0.89 on the test dataset. The models based on clinical-laboratory and radiology signs had low performance (AUCs about 0.5-0.6).

Difference of mean Hounsfield units (dHU) between follow-up and initial noncontrast CT scan predicts 90-day poor outcome in spontaneous supratentorial acute intracerebral hemorrhage with deep convolutional neural networks

OBJECTIVES

This study aimed to investigate the usefulness of a new non-contrast CT scan (NCCT) sign called the dHU, which represented the difference in mean Hounsfield unit values between follow-up and the initial NCCT for predicting 90-day poor functional outcomes in acute supratentorial spontaneous intracerebral hemorrhage(sICH) using deep convolutional neural networks.

METHODS

A total of 377 consecutive patients with sICH from center 1 and 91 patients from center 2 (external validation set) were included. A receiver operating characteristic (ROC) analysis was performed to determine the critical value of dHU for predicting poor outcome at 90 days. Modified Rankin score (mRS) >3 or >2 was defined as the primary and secondary poor outcome, respectively. Two multivariate models were developed to test whether dHU was an independent predictor of the two unfavorable functional outcomes.

RESULTS

The ROC analysis showed that a dHU >2.5 was a critical value to predict the poor outcomes (mRS >3) in sICH. The sensitivity, specificity, and accuracy of dHU >2.5 for poor outcome prediction were 37.5%, 86.0%, and 70.6%, respectively. In multivariate models developed after adjusting for all elements of the ICH score and hematoma expansion, dHU >2.5 was an independent predictor of both primary and secondary poor outcomes (OR = 2.61, 95% CI [1.32,5.13], P = 0.006; OR = 2.63, 95% CI [1.36,5.10], P = 0.004, respectively). After adjustment for all possible significant predictors (p < 0.05) by univariate analysis, dHU >2.5 had a positive association with primary and secondary poor outcomes (OR = 3.25, 95% CI [1.52,6.98], P = 0.002; OR = 3.42, 95% CI [1.64,7.15], P = 0.001).

CONCLUSIONS

The dHU of hematoma based on serial CT scans is independently associated with poor outcomes after acute sICH, which may help predict clinical evolution and guide therapy for sICH patients.

Outcome Prediction of Spontaneous Supratentorial Intracerebral Hemorrhage after Surgical Treatment Based on Non-Contrast Computed Tomography: A Multicenter Study

This study aims to explore the value of a machine learning (ML) model based on radiomics features and clinical features in predicting the outcome of spontaneous supratentorial intracerebral hemorrhage (sICH) 90 days after surgery. A total of 348 patients with sICH underwent craniotomy evacuation of hematoma from three medical centers. One hundred and eight radiomics features were extracted from sICH lesions on baseline CT. Radiomics features were screened using 12 feature selection algorithms. Clinical features included age, gender, admission Glasgow Coma Scale (GCS), intraventricular hemorrhage (IVH), midline shift (MLS), and deep ICH. Nine ML models were constructed based on clinical feature, and clinical features + radiomics features, respectively. Grid search was performed on different combinations of feature selection and ML model for parameter tuning. The averaged receiver operating characteristics (ROC) area under curve (AUC) was calculated and the model with the largest AUC was selected. It was then tested using multicenter data. The combination of lasso regression feature selection and logistic regression model based on clinical features + radiomics features had the best performance (AUC: 0.87). The best model predicted an AUC of 0.85 (95%CI, 0.75-0.94) on the internal test set and 0.81 (95%CI, 0.64-0.99) and 0.83 (95%CI, 0.68-0.97) on the two external test sets, respectively. Twenty-two radiomics features were selected by lasso regression. The second-order feature gray level non-uniformity normalized was the most important radiomics feature. Age is the feature with the greatest contribution to prediction. The combination of clinical features and radiomics features using logistic regression models can improve the outcome prediction of patients with sICH 90 days after surgery.

Development and validation of a novel radiomics-clinical model for predicting post-stroke epilepsy after first-ever intracerebral haemorrhage

OBJECTIVES

Post-stroke epilepsy (PSE) is associated with increased morbidity and mortality. This study aimed to develop and validate a novel prediction model combining clinical factors and radiomics features to accurately identify patients at high risk of developing PSE after intracerebral haemorrhage (ICH).

METHODS

Researchers performed a retrospective medical chart review to extract derivation and validation cohorts of patients with first-ever ICH that attended two tertiary hospitals in China between 2010 and 2020. Clinical data were extracted from electronic medical records and supplemented by tele-interview. Predictive clinical variables were selected by multivariable logistic regression to build the clinical model. Predictive radiomics features were identified, and a Rad-score was calculated according to the coefficient of the selected feature. Both clinical variables and radiomic features were combined to build the radiomics-clinical model. Performances of the clinical, Rad-score, and combined models were compared.

RESULTS

A total of 1571 patients were included in the analysis. Cortical involvement, early seizures within 7 days of ICH, NIHSS score, and ICH volume were included in the clinical model. Rad-score, instead of ICH volume, was included in the combined model. The combined model exhibited better discrimination ability and achieved an overall better benefit against threshold probability than the clinical model in the decision curve analysis (DCA).

CONCLUSIONS

The combined radiomics-clinical model was better able to predict ICH-associated PSE compared to the clinical model. This can help clinicians better predict an individual patient's risk of PSE following a first-ever ICH and facilitate earlier PSE diagnosis and treatment.

KEY POINTS

• Radiomics has not been used in predicting the risk of developing PSE. • Higher Rad-scores were associated with higher risk of developing PSE. • The combined model showed better performance of PSE prediction ability.

Prediction of short-term prognosis of patients with hypertensive intracerebral hemorrhage by radiomic-clinical nomogram

Hypertensive intracerebral hemorrhage (HICH) is the most common type of spontaneous intracerebral hemorrhage in China which is associated with high mortality and disability. We sought to develop and validate a noncontrast computed tomography (NCCT)-based nomogram model to achieve short-term prognostic prediction for patients with HICH. We retrospectively studied 292 patients with HICH from two medical centers, and they were divided into training (n = 151), validation (n = 66), and testing cohorts (n = 75). Based on radiomics, univariate and multivariate, and logistic regression analyses, four models (black hole sign, clinical, radiomics score, and combined models) were established to predict the prognosis of patients with HICH 30 days after the onset. The results suggested that the combined model had the best predictive performance with the area under the receiver operating characteristic curve (AUC) of 0.821, 0.816, and 0.815 in the training, validation, and testing cohorts, respectively. In addition, a radiomics-clinical (R-C) nomogram was visualized. A calibration curve analysis showed that the R-C nomogram had satisfactory calibration in the three cohorts. A decision curve analysis demonstrated that the R-C nomogram was clinically valuable. Our results suggest that the R-C nomogram can accurately and reliably predict the short-term prognosis of patients with HICH and provide a useful evaluation for making individualized treatment plans.

The "SALPARE study" of spontaneous intracerebral haemorrhage-part 2-early CT predictors of outcome in ICH: keeping it simple

BACKGROUND

The aim of this study was to investigate the prognostic role of hematoma characteristics and hematoma expansion (HE) in patients with spontaneous intracerebral hemorrhage (ICH).

METHODS

This multicenter prospective cohort study enrolled consecutive adult patients with non-traumatic ICH admitted to three Italian academic hospitals (Salerno, Padova, Reggio Emilia) over a 2-year period. Early noncontrast CT (NCCT) features of the hematoma, including markers of HE, and 3-month outcome were recorded. Multivariable logistic regression analysis was performed to identify predictors of poor outcome.

RESULTS

A total of 682 patients were included in the study [mean age: 73 ± 14 years; 316 (46.3%) females]. Pontine and massive hemorrhage, intraventricular bleeding, baseline hematoma volume > 15 mL, blend sign, swirl sign, margin irregularity ≥ 4, density heterogeneity ≥ 3, hypodensity ≥ 1, island sign, satellite sign, and black hole sign were associated with a higher risk of mortality and disability. However, at multivariate analysis only initial hematoma volume (OR 29.71) proved to be an independent predictor of poor functional outcome at 3 months.

CONCLUSION

Simple hematoma volume measured on baseline CT best identifies patients with a worse outcome, while early NCCT markers of HE do not seem to add any clinically significant information.

Efficacy of non-enhanced computer tomography-based radiomics for predicting hematoma expansion: A meta-analysis

Background

This meta-analysis aimed to assess the efficacy of radiomics using non-enhanced computed tomography (NCCT) for predicting hematoma expansion in patients with spontaneous intracerebral hemorrhage.

Methods

Throughout the inception of the project to April 11, 2022, a comprehensive search was conducted on PubMed, Embase, and Cochrane Central Register of Controlled Trials. The methodological quality of studies in this analysis was assessed by the radiomics quality scoring system (RQS). A meta-analysis of radiomic studies based on NCCT for predicting hematoma expansion in patients with intracerebral hemorrhage was performed. The efficacy of the radiomics approach and non-contrast CT markers was compared using network meta-analysis (NMA).

Results

Ten articles comprising a total of 1525 patients were quantitatively analyzed for hematoma expansion after cerebral hemorrhage using radiomics. Based on the included studies, the mean RQS was 14.4. The AUC value (95% confidence interval) of the radiomics model was 0.80 (0.76-0.83). Five articles comprising 846 patients were included in the NMA. The results synthesized according to Bayesian NMA revealed that the predictive ability of the radiomics model outperformed most of the NCCT biomarkers.

Conclusions

The NCCT-based radiomics approach has the potential to predict hematoma expansion. Compared to NCCT biomarkers, we recommend a radiomics approach. Standardization of the radiomics approach is required for further clinical implementation.

Systematic review registration

https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=324034, identifier [CRD42022324034].

Collinearity and Dimensionality Reduction in Radiomics: Effect of Preprocessing Parameters in Hypertrophic Cardiomyopathy Magnetic Resonance T1 and T2 Mapping

Radiomics and artificial intelligence have the potential to become a valuable tool in clinical applications. Frequently, radiomic analyses through machine learning methods present issues caused by high dimensionality and multicollinearity, and redundant radiomic features are usually removed based on correlation analysis. We assessed the effect of preprocessing-in terms of voxel size resampling, discretization, and filtering-on correlation-based dimensionality reduction in radiomic features from cardiac T1 and T2 maps of patients with hypertrophic cardiomyopathy. For different combinations of preprocessing parameters, we performed a dimensionality reduction of radiomic features based on either Pearson's or Spearman's correlation coefficient, followed by the computation of the stability index. With varying resampling voxel size and discretization bin width, for both T1 and T2 maps, Pearson's and Spearman's dimensionality reduction produced a slightly different percentage of remaining radiomic features, with a relatively high stability index. For different filters, the remaining features' stability was instead relatively low. Overall, the percentage of eliminated radiomic features through correlation-based dimensionality reduction was more dependent on resampling voxel size and discretization bin width for textural features than for shape or first-order features. Notably, correlation-based dimensionality reduction was less sensitive to preprocessing when considering radiomic features from T2 compared with T1 maps.

Semantic Segmentation of Spontaneous Intracerebral Hemorrhage, Intraventricular Hemorrhage, and Associated Edema on CT Images Using Deep Learning

This study evaluated deep learning algorithms for semantic segmentation and quantification of intracerebral hemorrhage (ICH), perihematomal edema (PHE), and intraventricular hemorrhage (IVH) on noncontrast CT scans of patients with spontaneous ICH. Models were assessed on 1732 annotated baseline noncontrast CT scans obtained from the Tranexamic Acid for Hyperacute Primary Intracerebral Haemorrhage (ie, TICH-2) international multicenter trial (ISRCTN93732214), and different loss functions using a three-dimensional no-new-U-Net (nnU-Net) were examined to address class imbalance (30% of participants with IVH in dataset). On the test cohort (n = 174, 10% of dataset), the top-performing models achieved median Dice similarity coefficients of 0.92 (IQR, 0.89-0.94), 0.66 (0.58-0.71), and 1.00 (0.87-1.00), respectively, for ICH, PHE, and IVH segmentation. U-Net-based networks showed comparable, satisfactory performances on ICH and PHE segmentations (P > .05), but all nnU-Net variants achieved higher accuracy than the Brain Lesion Analysis and Segmentation Tool for CT (BLAST-CT) and DeepLabv3+ for all labels (P < .05). The Focal model showed improved performance in IVH segmentation compared with the Tversky, two-dimensional nnU-Net, U-Net, BLAST-CT, and DeepLabv3+ models (P < .05). Focal achieved concordance values of 0.98, 0.88, and 0.99 for ICH, PHE, and ICH volumes, respectively. The mean volumetric differences between the ground truth and prediction were 0.32 mL (95% CI: -8.35, 9.00), 1.14 mL (-9.53, 11.8), and 0.06 mL (-1.71, 1.84), respectively. In conclusion, U-Net-based networks provide accurate segmentation on CT images of spontaneous ICH, and Focal loss can address class imbalance. International Clinical Trials Registry Platform (ICTRP) no. ISRCTN93732214 Supplemental material is available for this article. © RSNA, 2022 Keywords: Head/Neck, Brain/Brain Stem, Hemorrhage, Segmentation, Quantification, Convolutional Neural Network (CNN), Deep Learning Algorithms, Machine Learning Algorithms.

Radiomics for prediction of intracerebral hemorrhage outcomes: A retrospective multicenter study

BACKGROUND

Accurate risk stratification of patients with intracerebral hemorrhage (ICH) could help refine adjuvant therapy selection and better understand the clinical course. We aimed to evaluate the value of radiomics features from hematomal and perihematomal edema areas for prognosis prediction and to develop a model combining clinical and radiomic features for accurate outcome prediction of patients with ICH.

METHODS

This multicenter study enrolled patients with ICH from January 2016 to November 2021. Their outcomes at 3 months were recorded based on the modified Rankin Scale (good, 0-3; poor, 4-6). Independent clinical and radiomic risk factors for poor outcome were identified through multivariate logistic regression analysis, and predictive models were developed. Model performance and clinical utility were evaluated in both internal and external cohorts.

RESULTS

Among the 1098 ICH patients evaluated (mean age, 60 ± 13 years), 703 (64 %) had poor outcomes. Age, hemorrhage volume and location, and Glasgow Coma Scale (GCS) were independently associated with outcomes. The area under the receiver operating characteristic curve (AUC) of the clinical model was 0.881 in the external validation cohort. Addition of the Rad-score (combined hematoma and perihematomal edema area) improved predictive accuracy and model performance (AUC, 0.893), net reclassification improvement, 0.140 (P < 0.001), and integrated discrimination improvement, 0.050 (P < 0.001).

CONCLUSIONS

The radiomics features of hematomal and perihematomal edema area have additional value in prognostic prediction; moreover, addition of radiomic features significantly improves model accuracy.

A non-contrast computed tomography-based radiomics nomogram for the prediction of hematoma expansion in patients with deep ganglionic intracerebral hemorrhage

Background and purpose

Hematoma expansion (HE) is a critical event following acute intracerebral hemorrhage (ICH). We aimed to construct a non-contrast computed tomography (NCCT) model combining clinical characteristics, radiological signs, and radiomics features to predict HE in patients with spontaneous ICH and to develop a nomogram to assess the risk of early HE.

Materials and methods

We retrospectively reviewed 388 patients with ICH who underwent initial NCCT within 6 h after onset and follow-up CT within 24 h after initial NCCT, between January 2015 and December 2021. Using the LASSO algorithm or stepwise logistic regression analysis, five models (clinical model, radiological model, clinical-radiological model, radiomics model, and combined model) were developed to predict HE in the training cohort (n = 235) and independently verified in the test cohort (n = 153). The Akaike information criterion (AIC) and the likelihood ratio test (LRT) were used for comparing the goodness of fit of the five models, and the AUC was used to evaluate their ability in discriminating HE. A nomogram was developed based on the model with the best performance.

Results

The combined model (AIC = 202.599, χ2 = 80.6) was the best fitting model with the lowest AIC and the highest LRT chi-square value compared to the clinical model (AIC = 232.263, χ2 = 46.940), radiological model (AIC = 227.932, χ2 = 51.270), clinical-radiological model (AIC = 212.711, χ2 = 55.490) or radiomics model (AIC = 217.647, χ2 = 57.550). In both cohorts, the nomogram derived from the combined model showed satisfactory discrimination and calibration for predicting HE (AUC = 0.900, sensitivity = 83.87%; AUC = 0.850, sensitivity = 80.10%, respectively).

Conclusion

The NCCT-based model combining clinical characteristics, radiological signs, and radiomics features could efficiently discriminate early HE, and the nomogram derived from the combined model, as a non-invasive tool, exhibited satisfactory performance in stratifying HE risks.

Prediction of Intraparenchymal Hemorrhage Progression and Neurologic Outcome in Traumatic Brain Injury Patients Using Radiomics Score and Clinical Parameters

(1) Background: Radiomics analysis of spontaneous intracerebral hemorrhages on computed tomography (CT) images has been proven effective in predicting hematoma expansion and poor neurologic outcome. In contrast, there is limited evidence on its predictive abilities for traumatic intraparenchymal hemorrhage (IPH). (2) Methods: A retrospective analysis of 107 traumatic IPH patients was conducted. Among them, 45 patients (42.1%) showed hemorrhagic progression of contusion (HPC) and 51 patients (47.7%) had poor neurological outcome. The IPH on the initial CT was manually segmented for radiomics analysis. After feature extraction, selection and repeatability evaluation, several machine learning algorithms were used to derive radiomics scores (R-scores) for the prediction of HPC and poor neurologic outcome. (3) Results: The AUCs for R-scores alone to predict HPC and poor neurologic outcome were 0.76 and 0.81, respectively. Clinical parameters were used to build comparison models. For HPC prediction, variables including age, multiple IPH, subdural hemorrhage, Injury Severity Score (ISS), international normalized ratio (INR) and IPH volume taken together yielded an AUC of 0.74, which was significantly (p = 0.022) increased to 0.83 after incorporation of the R-score in a combined model. For poor neurologic outcome prediction, clinical variables of age, Glasgow Coma Scale, ISS, INR and IPH volume showed high predictability with an AUC of 0.92, and further incorporation of the R-score did not improve the AUC. (4) Conclusion: The results suggest that radiomics analysis of IPH lesions on initial CT images has the potential to predict HPC and poor neurologic outcome in traumatic IPH patients. The clinical and R-score combined model further improves the performance of HPC prediction.

Development and Validation of a Clinical-Based Signature to Predict the 90-Day Functional Outcome for Spontaneous Intracerebral Hemorrhage

We aimed to develop and validate an objective and easy-to-use model for identifying patients with spontaneous intracerebral hemorrhage (ICH) who have a poor 90-day prognosis. This three-center retrospective study included a large cohort of 1,122 patients with ICH who presented within 6 h of symptom onset [training cohort, n = 835; internal validation cohort, n = 201; external validation cohort (center 2 and 3), n = 86]. We collected the patients’ baseline clinical, radiological, and laboratory data as well as the 90-day functional outcomes. Independent risk factors for prognosis were identified through univariate analysis and multivariate logistic regression analysis. A nomogram was developed to visualize the model results while a calibration curve was used to verify whether the predictive performance was satisfactorily consistent with the ideal curve. Finally, we used decision curves to assess the clinical utility of the model. At 90 days, 714 (63.6%) patients had a poor prognosis. Factors associated with prognosis included age, midline shift, intraventricular hemorrhage (IVH), subarachnoid hemorrhage (SAH), hypodensities, ICH volume, perihematomal edema (PHE) volume, temperature, systolic blood pressure, Glasgow Coma Scale (GCS) score, white blood cell (WBC), neutrophil, and neutrophil-lymphocyte ratio (NLR) (p < 0.05). Moreover, age, ICH volume, and GCS were identified as independent risk factors for prognosis. For identifying patients with poor prognosis, the model showed an area under the receiver operating characteristic curve of 0.874, 0.822, and 0.868 in the training cohort, internal validation, and external validation cohorts, respectively. The calibration curve revealed that the nomogram showed satisfactory calibration in the training and validation cohorts. Decision curve analysis showed the clinical utility of the nomogram. Taken together, the nomogram developed in this study could facilitate the individualized outcome prediction in patients with ICH.

Can Artificial Intelligence Be Applied to Diagnose Intracerebral Hemorrhage under the Background of the Fourth Industrial Revolution? A Novel Systemic Review and Meta-Analysis

Aim

We intended to provide the clinical evidence that artificial intelligence (AI) could be used to assist doctors in the diagnosis of intracerebral hemorrhage (ICH).

Methods

Studies published in 2021 were identified after the literature search of PubMed, Embase, and Cochrane. Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) was used to perform the quality assessment of studies. Data extraction of diagnosis effect included accuracy (ACC), sensitivity (SEN), specificity (SPE), positive predictive value (PPV), negative predictive value (NPV), area under curve (AUC), and Dice scores (Dices). The pooled effect with its 95% confidence interval (95%CI) was calculated by the random effects model. I-Square (I 2) was used to test heterogeneity. To check the stability of the overall results, sensitivity analysis was conducted by recalculating the pooled effect of the remaining studies after omitting the study with the highest quality or the random effects model was switched to the fixed effects model. Funnel plot was used to evaluate publication bias. To reduce heterogeneity, recalculating the pooled effect of the remaining studies after omitting the study with the lowest quality or perform subgroup analysis.

Results

Twenty-five diagnostic tests of ICH via AI and doctors with overall high quality were included. Pooled ACC, SEN, SPE, PPV, NPV, AUC, and Dices were 0.88 (0.83∼0.93), 0.85 (0.81∼0.89), 0.90 (0.88∼0.92), 0.80 (0.75∼0.85), 0.93 (0.91∼0.95), 0.84 (0.80∼0.89), and 0.90 (0.85∼0.95), respectively. There was no publication bias. All of results were stable as revealed by sensitivity analysis and were accordant as outcomes via subgroups analysis.

Conclusion

Under the background of the fourth industrial revolution, AI might be an effective and efficient tool to assist doctors in the clinical diagnosis of ICH.

Location-Specific Radiomics Score: Novel Imaging Marker for Predicting Poor Outcome of Deep and Lobar Spontaneous Intracerebral Hemorrhage

Objective: To derive and validate a location-specific radiomics score (Rad-score) based on noncontrast computed tomography for predicting poor deep and lobar spontaneous intracerebral hemorrhage (SICH) outcome.

Methods: In total, 494 SICH patients from multiple centers were retrospectively reviewed. Poor outcome was considered mRS 3–6 at 6 months. The Rad-score was derived using optimal radiomics features. The optimal location-specific Rad-score cut-offs for poor deep and lobar SICH outcomes were identified using receiver operating characteristic curve analysis. Univariable and multivariable analyses were used to determine independent poor outcome predictors. The combined models for deep and lobar SICH were constructed using independent predictors of poor outcomes, including dichotomized Rad-score in the derivation cohort, which was validated in the validation cohort.

Results: Of 494 SICH patients, 392 (79%) had deep SICH, and 373 (76%) had poor outcomes. The Glasgow Coma Scale score, haematoma enlargement, haematoma location, haematoma volume and Rad-score were independent predictors of poor outcomes (all P < 0.05). Cut-offs of Rad-score, 82.90 (AUC = 0.794) in deep SICH and 80.77 (AUC = 0.823) in lobar SICH, were identified for predicting poor outcomes. For deep SICH, the AUCs of the combined model were 0.856 and 0.831 in the derivation and validation cohorts, respectively. For lobar SICH, the combined model AUCs were 0.866 and 0.843 in the derivation and validation cohorts, respectively.

Conclusion: Location-specific Rad-scores and combined models can identify subjects at high risk of poor deep and lobar SICH outcomes, which could improve clinical trial design by screening target patients.

Multi-Stage Harmonization for Robust AI across Breast MR Databases

Simple Summary

Batch harmonization of radiomic features extracted from magnetic resonance images of breast lesions from two databases was applied to an artificial intelligence/machine learning classification workflow. Training and independent test sets from the two databases, as well as the combination of them, were used in pre-harmonization and post-harmonization forms to investigate the generalizability of performance in the task of distinguishing between malignant and benign lesions. Most training and independent test scenarios were statistically equivalent, demonstrating that batch harmonization with feature selection harmonization can potentially develop generalizable classification models.

Abstract

Radiomic features extracted from medical images may demonstrate a batch effect when cases come from different sources. We investigated classification performance using training and independent test sets drawn from two sources using both pre-harmonization and post-harmonization features. In this retrospective study, a database of thirty-two radiomic features, extracted from DCE-MR images of breast lesions after fuzzy c-means segmentation, was collected. There were 944 unique lesions in Database A (208 benign lesions, 736 cancers) and 1986 unique lesions in Database B (481 benign lesions, 1505 cancers). The lesions from each database were divided by year of image acquisition into training and independent test sets, separately by database and in combination. ComBat batch harmonization was conducted on the combined training set to minimize the batch effect on eligible features by database. The empirical Bayes estimates from the feature harmonization were applied to the eligible features of the combined independent test set. The training sets (A, B, and combined) were then used in training linear discriminant analysis classifiers after stepwise feature selection. The classifiers were then run on the A, B, and combined independent test sets. Classification performance was compared using pre-harmonization features to post-harmonization features, including their corresponding feature selection, evaluated using the area under the receiver operating characteristic curve (AUC) as the figure of merit. Four out of five training and independent test scenarios demonstrated statistically equivalent classification performance when compared pre- and post-harmonization. These results demonstrate that translation of machine learning techniques with batch data harmonization can potentially yield generalizable models that maintain classification performance.