Multi-phase contrast-enhanced magnetic resonance image-based radiomics-combined machine learning reveals microscopic ultra-early hepatocellular carcinoma lesions

Multiparameter MRI-based model integrating radiomics and deep learning for preoperative staging of laryngeal squamous cell carcinoma

The accurate preoperative staging of laryngeal squamous cell carcinoma (LSCC) provides valuable guidance for clinical decision-making. The objective of this study was to establish a multiparametric MRI model using radiomics and deep learning (DL) to preoperatively distinguish between Stages I-II and III-IV of LSCC. Data from 401 histologically confirmed LSCC patients were collected from two centers (training set: 213; internal test set: 91; external test set: 97). Radiomics features were extracted from the MRI images, and seven radiomics models based on single and combined sequences were developed via random forest (RF). A DL model was constructed via ResNet 18, where DL features were extracted from its final fully connected layer. These features were fused with crucial radiomics features to create a combined model. The performance of the models was assessed using the area under the receiver operating characteristic (ROC) curve (AUC) and compared with the radiologist performances. The predictive capability of the combined model for Progression-Free Survival (PFS) was evaluated via Kaplan-Meier survival analysis and the Harrell's Concordance Index (C-index). In the external test set, the combined model had an AUC of 0.877 (95% CI 0.807-0.946), outperforming the DL model (AUC: 0.811) and the optimal radiomics model (AUC: 0.835). The combined model significantly outperformed both the DL (p = 0.017) and the optimal radiomics models (p = 0.039), and the radiologists (both p < 0.050). Moreover, the combined model demonstrated great prognostic predictive value in patients with LSCC, achieving a C-index of 0.624 for PFS. This combined model enhances preoperative LSCC staging, aiding in making more informed clinical decisions.

Artificial intelligence-assisted platform performs high detection ability of hepatocellular carcinoma in CT images: an external clinical validation study

BACKGROUND

Accurate detection of hepatocellular carcinoma (HCC) in multiphasic contrast CT is essential for effective treatment and surgical planning. However, the variety of CT images, the misdiagnosis and missed diagnosis, and the inconsistent diagnosis among different radiologists pose challenges to accurate detection which demands sufficient clinical experience and can be time-consuming and labor-intensive.

PURPOSE

To evaluate the detection performance of an artificial intelligence (AI)-assisted platform for HCC by the external validation dataset.

METHODS

CT images pathologically diagnosed with HCC from December 2021 to June 2023 were retrospectively analyzed to evaluate the detection ability of the AI-assisted platform. The AI-assisted platform is designed based on a two-phase segmentation approach, integrating coarse and fine segmentation techniques to accurately identify and delineate hepatic lesions. The CT images were annotated and confirmed by the experienced radiologists using InferScholar software as the "gold standard". The automatic HCC segmentation performed by the AI-assisted platform was used to compare with the annotation of radiologists. Furthermore, we also did subgroup analysis depending on the size and location of HCC to explore the impact factors of HCC detectability. The segmentation accuracies were evaluated by Dice coefficient (Dice), accuracy, recall, precision, and F1-score. Our study focused on evaluating the efficacy of the AI-assisted platform in clinical settings.

RESULTS

One Hundred Forty HCC patients were finally included in this study. The artificial intelligence (AI)-assisted platform's performance was rigorously assessed by comparing the segmentation outcomes with standard diagnostic criteria. The average dice score of the AI-assisted platform is 0.8819, which showed a high detection performance for HCC. Besides, for the subgroup analysis, the model also demonstrated high performance in diameter greater than 20 mm with all results exceeding 0.9, and all final evaluation index values for the location analysis were consistently exceeding 0.97. All the results showed comparable performance with radiologists. Our results demonstrate that the product not only accurately segments HCC lesions but also provides valuable insights into lesion characteristics that are essential for effective treatment planning.

CONCLUSION

This study validates the effectiveness of the artificial intelligence-assisted platform in detecting HCC lesions and analyzing lesion size and location. It can serve as an auxiliary tool to help radiologists identify, locate, and assess lesions.

Application value of different imaging methods in the early diagnosis of small hepatocellular carcinoma: a network meta-analysis

Objective

To determine the diagnostic value of ultrasound, multi-phase enhanced computed tomography, and magnetic resonance imaging of small hepatocellular carcinoma.

Methods

Experimental studies on diagnosing small hepatocellular carcinoma in four databases: PubMed, Cochrane Library, Web of Science, and Embase, were comprehensively searched from October 2007 to October 2024. Relevant diagnostic accuracy data were extracted and a Bayesian model that combined direct and indirect evidence was used for analysis.

Results

16 original studies were included and data from 2,447 patients were collated to assess the diagnostic value of 10 different methods. The methodological quality of the included studies was good and there was no obvious publication bias. The pooled DOR of all diagnostic methods was 19.61, which was statistically significant (I2 = 76.0%, P < 0.01, 95% CI:13.30 - 28.92). Normal US + CEUS + ultrasonic elastic imaging had the highest specificity (92.9), accuracy (93.6), and positive predictive value (94.4). Unenhanced MRI + Contrast-enhanced MRI had the highest sensitivity (96.6) and negative predictive value (96.6), but specificity (12.5) and positive predictive value (34.4) were extremely poor. Contrast-enhanced MRI had the highest diagnostic value in individual imaging methods (sensitivity: 66, specificity: 55.5, accuracy: 67.9, positive predictive value: 64.4, negative predictive value: 66.5). There was significant inconsistency and high heterogeneity in this study.

Systematic review registration

https://www.crd.york.ac.uk/PROSPERO/, identifier CRD42024507883.

Comparison of multiple machine learning models for predicting prognosis of pancreatic ductal adenocarcinoma based on contrast-enhanced CT radiomics and clinical features

Objective

The aim of this study was to evaluate the prognostic potential of combining clinical features and radiomics with multiple machine learning (ML) algorithms in pancreatic ductal adenocarcinoma (PDAC).

Methods

A total of 116 patients with PDAC who met the eligibility criteria were randomly assigned to a training or validation cohort. Seven ML algorithms, including Supervised Principal Components, stepwise Cox, Random Survival Forest, CoxBoost, Least absolute shrinkage and selection operation (Lasso), Ridge, and Elastic network, were integrated into 43 algorithm combinations. Forty-three radiomics models were constructed separately using radiomics features extracted from arterial phase (AP), venous phase (VP), and combined arterial and venous phase (AP+VP) images. The concordance index (C-index) of each model was calculated. The model with the highest mean C-index was identified as the best model for calculating the radiomics score (Radscore). Univariate and multivariate Cox analyses were used to identify independent prognostic indicators and create a clinical model for prognosis prediction. The multivariable Cox regression was used to combine Radscore with clinical features to create a combined model. The efficacy of the model was evaluated using the C-index, calibration curves, and decision curve analysis (DCA).

Results

The model based on the Lasso+StepCox[both] algorithm constructed using AP+VP radiomics features showed the best predictive ability among the 114 radiomics models. The C-indices of the model in the training and validation cohorts were 0.742 and 0.722, respectively. Based on the results of the univariate and multivariate Cox regression analyses, sex, Tumor-Node-Metastasis (TNM) stage, and systemic inflammation response index were included to build the clinical model. The combined model, incorporating three clinical factors and AP+VP-Radscore, achieved the highest C-indices of 0.764 and 0.746 in the training and validation cohorts, respectively. In terms of preoperative prognosis prediction for PDAC, the calibration curve and DCA showed that the combined model had a good consistency and greatest net benefit.

Conclusion

A combined model of clinical features and AP+VP-Radscore screened using multiple ML algorithms has an excellent ability to predict the prognosis of PDAC and may provide a noninvasive and effective method for clinical decision-making.

Artificial intelligence in hepatocellular carcinoma diagnosis: a comprehensive review of current literature

BACKGROUND AND AIM

Hepatocellular carcinoma (HCC) diagnosis mainly relies on its pathognomonic radiological profile, obviating the need for biopsy. The project of incorporating artificial intelligence (AI) techniques in HCC aims to improve the performance of image recognition. Herein, we thoroughly analyze and evaluate proposed AI models in the field of HCC diagnosis.

METHODS

A comprehensive review of the literature was performed utilizing MEDLINE/PubMed and Web of Science databases with the end of search date being the 30th of September 2023. The MESH terms "Artificial Intelligence," "Liver Cancer," "Hepatocellular Carcinoma," "Machine Learning," and "Deep Learning" were searched in the title and/or abstract. All references of the obtained articles were also evaluated for any additional information.

RESULTS

Our search resulted in 183 studies meeting our inclusion criteria. Across all diagnostic modalities, reported area under the curve (AUC) of most developed models surpassed 0.900. A B-mode US and a contrast-enhanced US model achieved AUCs of 0.947 and 0.957, respectively. Regarding the more challenging task of HCC diagnosis, a 2021 deep learning model, trained with CT scans, classified hepatic malignant lesions with an AUC of 0.986. Finally, a MRI machine learning model developed in 2021 displayed an AUC of 0.975 when differentiating small HCCs from benign lesions, while another MRI-based model achieved HCC diagnosis with an AUC of 0.970.

CONCLUSIONS

AI tools may lead to significant improvement in diagnostic management of HCC. Many models fared better or comparable to experienced radiologists while proving capable of elevating radiologists' accuracy, demonstrating promising results for AI implementation in HCC-related diagnostic tasks.

CTA-Based Radiomics and Area Change Rate Predict Infrarenal Abdominal Aortic Aneurysms Patients Events: A Multicenter Study

RATIONALE AND OBJECTIVES

Clinical assessment of abdominal aortic aneurysm (AAA) intervention and rupture risk relies primarily on maximum diameter, but studies have shown that sole dependence on diameter has limitations. CTA-based radiomics, aneurysm and lumen area change rates (AACR, LACR) are measured to predict potential AAA events.

MATERIALS AND METHODS

Between January 2017 and November 2022, 260 AAA patients from four centers who underwent two preoperative CTA examinations were included in this retrospective study. The endpoint event is defined as AAA rupture or repair. Patients were categorized into event and no-event groups based on the occurrence of endpoint event during follow-up. AACR and LACR were assessed using baseline and follow-up CTA, with radiomics features extracted from the baseline images. C-statistics and the Kaplan-Meier analysis were used to evaluate the predictive performance.

RESULTS

A total of 193 eligible infrarenal AAA patients were included, 176 (91.2%) were man and 17 (8.8%) were woman. The median follow-up was 33.4 (14.2, 57.4) months. Seven models were constructed, comprising the aneurysm-based Radscore model, lumen-based Radscore model, intraluminal thrombus (ILT)-based Radscore model, AACR model, LACR model, clinical model (including high-density lipoprotein, D-dimer, and baseline aneurysm diameter), and a merged model. On the external validation set, the C-index of seven models were 0.713 (0.574-0.853), 0.642 (0.499-0.786), 0.727 (0.600-0.854), 0.619 (0.484-0.753), 0.680 (0.530-0.830), 0.690 (0.557-0.824) and 0.760 (0.651-0.869), in that order. In the Kaplan-Meier analysis, the merged model was best-divided patients into high/low-risk groups with Log-rank p < 0.0001. The AARC and LARC between non-event and event groups have significant differences (AACR: 1.4 cm2/y vs. 2.3 cm2/y, p < 0.0001; LACR: 0.3 cm2/y vs. 1.1 cm2/y, p < 0.0001).

CONCLUSION

CTA-based radiomics, AACR and LACR have good predictive value for outcome event in infrarenal AAA patients.

Multiparametric magnetic resonance imaging (MRI)-based radiomics model explained by the Shapley Additive exPlanations (SHAP) method for predicting complete response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer: a multicenter retrospective study

BACKGROUND

Predicting the response to neoadjuvant chemoradiotherapy (nCRT) before initiating treatment is essential for tailoring therapeutic strategies and monitoring prognosis in locally advanced rectal cancer (LARC). In this study, we aimed to develop and validate radiomic-based models to predict clinical and pathological complete responses (cCR and pCR, respectively) by incorporating the Shapley Additive exPlanations (SHAP) method for model interpretation.

METHODS

A total of 285 patients with complete pretreatment clinical characteristics and T1-weighted (T1W) and T2-weighted (T2W) magnetic resonance imaging (MRI) at 3 centers were retrospectively recruited. The features of tumor lesions were extracted by PyRadiomics and selected using least absolute shrinkage and selection operator (LASSO) algorithm. The selected features were used to build multilayer perceptron (MLP) models alone or combined with clinical features. Area under the receiver operating characteristic curve (AUC), decision curve, and calibration curve were applied to evaluate performance of models. The SHAP method was adopted to explain the prediction models.

RESULTS

The radiomic-based models all showed better performances than clinical models. The clinical-radiomic models showed the best differentiation on cCR and pCR with mean AUCs of 0.718 and 0.810 in the validation set, respectively. The decision curves of the clinical-radiomic models showed its values in clinical application. The SHAP method powerfully interpreted the prediction models both at a holistic and individual levels.

CONCLUSIONS

Our study highlights that the radiomic-based prediction models have more excellent abilities than clinical models and can effectively predict treatment response and optimize therapeutic strategies for patients with LARCs.

Prediction of glypican-3 expression in hepatocellular carcinoma using multisequence magnetic resonance imaging-based histology nomograms

Background

Hepatocellular carcinoma (HCC) is often associated with the overexpression of multiple proteins and genes. For instance, patients with HCC and a high expression of the glypican-3 (GPC3) gene have a poor prognosis, and noninvasive assessment of GPC3 expression before surgery is helpful for clinical decision-making. Therefore, our primary aim in this study was to develop and validate multisequence magnetic resonance imaging (MRI) radiomics nomograms for predicting the expression of GPC3 in individuals diagnosed with HCC.

Methods

We conducted a retrospective analysis of 143 patients with HCC, including 123 cases from our hospital and 20 cases from The Cancer Genome Atlas (TCGA) or The Cancer Imaging Archive (TCIA) public databases. We used preoperative multisequence MRI images of the patients for the radiomics analysis. We extracted and screened the imaging histologic features using fivefold cross-validation, Pearson correlation coefficient, and the least absolute shrinkage and selection operator (LASSO) analysis method. We used logistic regression (LR) to construct a radiomics model, developed nomograms based on the radiomics scores and clinical parameters, and evaluated the predictive performance of the nomograms using receiver operating characteristic (ROC) curves, calibration curves, and decision curves.

Results

Our multivariate analysis results revealed that tumor morphology (P=0.015) and microvascular (P=0.007) infiltration could serve as independent predictors of GPC3 expression in patients with HCC. The nomograms integrating multisequence radiomics radiomics score, tumor morphology, and microvascular invasion had an area under the curve (AUC) value of 0.989. This approach was superior to both the radiomics model (AUC 0.979) and the clinical model (AUC 0.793). The sensitivity, specificity, and accuracy of 0.944, 0.800, and 0.913 for the test set, respectively, and the model's calibration curve demonstrated good consistency (Brier score =0.029). The decision curve analysis (DCA) indicated that the nomogram had a higher net clinical benefit for predicting the expression of GPC3. External validation of the model's prediction yielded an AUC value of 0.826.

Conclusions

Our study findings highlight the close association of multisequence MRI imaging and radiomic features with GPC3 expression. Incorporating clinical parameters into nomograms can offer valuable preoperative insights into tailoring personalized treatment plans for patients diagnosed with HCC.

Development of a multi-phase CT-based radiomics model to differentiate heterotopic pancreas from gastrointestinal stromal tumor

BACKGROUND

To investigate whether CT-based radiomics can effectively differentiate between heterotopic pancreas (HP) and gastrointestinal stromal tumor (GIST), and whether different resampling methods can affect the model's performance.

METHODS

Multi-phase CT radiological data were retrospectively collected from 94 patients. Of these, 40 with HP and 54 with GISTs were enrolled between April 2017 and November 2021. One experienced radiologist manually delineated the volume of interest and then resampled the voxel size of the images to 0.5 × 0.5 × 0.5 mm3, 1 × 1 × 1 mm3, and 2 × 2 × 2 mm3, respectively. Radiomics features were extracted using PyRadiomics, resulting in 1218 features from each phase image. The datasets were randomly divided into training set (n = 66) and validation set (n = 28) at a 7:3 ratio. After applying multiple feature selection methods, the optimal features were screened. Radial basis kernel function-based support vector machine (RBF-SVM) was used as the classifier, and model performance was evaluated using the area under the receiver operating curve (AUC) analysis, as well as accuracy, sensitivity, and specificity.

RESULTS

The combined phase model performed better than the other phase models, and the resampling method of 0.5 × 0.5 × 0.5 mm3 achieved the highest performance with an AUC of 0.953 (0.881-1), accuracy of 0.929, sensitivity of 0.938, and specificity of 0.917 in the validation set. The Delong test showed no significant difference in AUCs among the three resampling methods, with p > 0.05.

CONCLUSIONS

Radiomics can effectively differentiate between HP and GISTs on CT images, and the diagnostic performance of radiomics is minimally affected by different resampling methods.

Machine learning methods for accurately predicting survival and guiding treatment in stage I and II hepatocellular carcinoma

Accurately predicting survival in patients with early hepatocellular carcinoma (HCC) is essential for making informed decisions about treatment and prognosis. Herein, we have developed a machine learning (ML) model that can predict patient survival and guide treatment decisions. We obtained patient demographic information, tumor characteristics, and treatment details from the SEER database. To analyze the data, we employed a Cox proportional hazards (CoxPH) model as well as 3 ML algorithms: neural network multitask logistic regression (N-MLTR), DeepSurv, and random survival forest (RSF). Our evaluation relied on the concordance index (C-index) and Integrated Brier Score (IBS). Additionally, we provided personalized treatment recommendations regarding surgery and chemotherapy choices and validated models' efficacy. A total of 1136 patients with early-stage (I, II) hepatocellular carcinoma (HCC) who underwent liver resection or transplantation were randomly divided into training and validation cohorts at a ratio of 3:7. Feature selection was conducted using Cox regression analyses. The ML models (NMLTR: C-index = 0.6793; DeepSurv: C-index = 0.7028; RSF: C-index = 0.6890) showed better discrimination in predicting survival than the standard CoxPH model (C-index = 0.6696). Patients who received recommended treatments had higher survival rates than those who received unrecommended treatments. ML-based surgery treatment recommendations yielded higher hazard ratios (HRs): NMTLR HR = 0.36 (95% CI: 0.25-0.51, P < .001), DeepSurv HR = 0.34 (95% CI: 0.24-0.49, P < .001), and RSF HR = 0.37 (95% CI: 0.26-0.52, P = <.001). Chemotherapy treatment recommendations were associated with significantly improved survival for DeepSurv (HR: 0.57; 95% CI: 0.4-0.82, P = .002) and RSF (HR: 0.66; 95% CI: 0.46-0.94, P = .020). The ML survival model has the potential to benefit prognostic evaluation and treatment of HCC. This novel analytical approach could provide reliable information on individual survival and treatment recommendations.

Precision diagnosis of hepatocellular carcinoma

ABSTRACT

Hepatocellular carcinoma (HCC) is the most common type of primary hepatocellular carcinoma (PHC). Early diagnosis of HCC remains the key to improve the prognosis. In recent years, with the promotion of the concept of precision medicine and more in-depth analysis of the biological mechanism underlying HCC, new diagnostic methods, including emerging serum markers, liquid biopsies, molecular diagnosis, and advances in imaging (novel contrast agents and radiomics), have emerged one after another. Herein, we reviewed and analyzed scientific advances in the early diagnosis of HCC and discussed their application and shortcomings. This review aimed to provide a reference for scientific research and clinical practice of HCC.

Radiomics: A radiological evidence-based artificial intelligence technique to facilitate personalized precision medicine in hepatocellular carcinoma

The high postoperative recurrence rates in hepatocellular carcinoma (HCC) remain a major hurdle in its management. Appropriate staging and treatment selection may alleviate the extent of fatal recurrence. However, effective methods to preoperatively evaluate pathophysiologic and molecular characteristics of HCC are lacking. Imaging plays a central role in HCC diagnosis and stratification due to the non-invasive diagnostic criteria. Vast and crucial information is hidden within image data. Other than providing a morphological sketch for lesion diagnosis, imaging could provide new insights to describe the pathophysiological and genetic landscape of HCC. Radiomics aims to facilitate diagnosis and prognosis of HCC using artificial intelligence techniques to harness the immense information contained in medical images. Radiomics produces a set of archetypal and robust imaging features that are correlated to key pathological or molecular biomarkers to preoperatively risk-stratify HCC patients. Inferred with outcome data, comprehensive combination of radiomic, clinical and/or multi-omics data could also improve direct prediction of response to treatment and prognosis. The evolution of radiomics is changing our understanding of personalized precision medicine in HCC management. Herein, we review the key techniques and clinical applications in HCC radiomics and discuss current limitations and future opportunities to improve clinical decision making.

CT-based radiomics nomogram for differentiation of adrenal hyperplasia from lipid-poor adenoma: an exploratory study

BACKGROUND

To establish and verify a radiomics nomogram for differentiating isolated micronodular adrenal hyperplasia (iMAD) from lipid-poor adenoma (LPA) based on computed tomography (CT)-extracted radiomic features.

METHODS

A total of 148 patients with iMAD or LPA were divided into three cohorts: a training cohort (n = 72; 37 iMAD and 35 LPA), a validation cohort (n = 36; 22 iMAD and 14 LPA), and an external validation cohort (n = 40; 20 iMAD and 20 LPA). Radiomics features were extracted from contrast-enhanced and non-contrast CT images. The least absolute shrinkage and selection operator (LASSO) method was applied to develop a triphasic radiomics model and unenhanced radiomics model using reproducible radiomics features. A clinical model was constructed using certain laboratory variables and CT findings. Radiomics nomogram was established by selected radiomics signature and clinical factors. Nomogram performance was assessed by calibration curve, the areas under receiver operating characteristic curves (AUC), and decision curve analysis (DCA).

RESULTS

Eleven and eight extracted features were finally selected to construct an unenhanced radiomics model and a triphasic radiomics model, respectively. There was no significant difference in AUC between the two models in the external validation cohort (0.838 vs. 0.843, p = 0.949). The radiomics nomogram inclusive of the unenhanced model, maximum diameter, and aldosterone showed the AUC of 0.951, 0.938, and 0.893 for the training, validation, and external validation cohorts, respectively. The nomogram showed good calibration, and the DCA demonstrated the superiority of the nomogram compared with the clinical factors model alone in terms of clinical usefulness.

CONCLUSIONS

A radiomics nomogram based on unenhanced CT images and clinical variables showed favorable performance for distinguishing iMAD from LPA. In addition, an efficient unenhanced model can help avoid extra contrast-enhanced scanning and radiation risk.

The Role of Radiomics and AI Technologies in the Segmentation, Detection, and Management of Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is the most common primary hepatic neoplasm. Thanks to recent advances in computed tomography (CT) and magnetic resonance imaging (MRI), there is potential to improve detection, segmentation, discrimination from HCC mimics, and monitoring of therapeutic response. Radiomics, artificial intelligence (AI), and derived tools have already been applied in other areas of diagnostic imaging with promising results. In this review, we briefly discuss the current clinical applications of radiomics and AI in the detection, segmentation, and management of HCC. Moreover, we investigate their potential to reach a more accurate diagnosis of HCC and to guide proper treatment planning.