A Novel Application of Unsupervised Machine Learning and Supervised Machine Learning-Derived Radiomics in Anterior Cruciate Ligament Rupture

Exploring tumor heterogeneity in colorectal liver metastases by imaging: Unsupervised machine learning of preoperative CT radiomics features for prognostic stratification

OBJECTIVES

This study aimed to investigate tumor heterogeneity of colorectal liver metastases (CRLM) and stratify the patients into different risk groups of prognoses following liver resection by applying an unsupervised radiomics machine-learning approach to preoperative CT images.

METHODS

This retrospective study retrieved clinical information and CT images of 197 patients with CRLM from The Cancer Imaging Archive (TCIA) database. Radiomics features were extracted from a segmented liver lesion identified at the portal venous phase. Those features which showed high stability, non-redundancy, and indicative information were selected. An unsupervised consensus clustering analysis on these features was adopted to identify subgroups of CRLM patients. Overall survival (OS), disease-free survival (DFS), and liver-specific DFS were compared between the identified subgroups. Cox regression analysis was applied to evaluate prognostic risk factors.

RESULTS

A total of 851 radiomics features were extracted, and 56 robust features were finally selected for unsupervised clustering analysis which identified two distinct subgroups (96 and 101 patients respectively). There were significant differences in the OS, DFS, and liver-specific DFS between the subgroups (all log-rank p < 0.05). The subgroup with worse outcome using the proposed radiomics model was consistently associated with shorter OS, DFS, and liver-specific DFS, with hazard ratios of 1.78 (95 %CI: 1.12-2.83), 1.72 (95 %CI: 1.16-2.54), and 1.59 (95 %CI: 1.10-2.31), respectively. The general performance of this radiomics model outperformed the traditional Clinical Risk Score and Tumor Burden Score in the prognosis prediction after surgery for CRLM.

CONCLUSION

Radiomics features derived from preoperative CT images can reveal the heterogeneity of CRLM and stratify the patients with CRLM into subgroups with significantly different clinical outcomes.

Radiomics signature of osteoarthritis: Current status and perspective

Osteoarthritis (OA) is one of the fast-growing disability-related diseases worldwide, which has significantly affected the quality of patients' lives and brings about substantial socioeconomic burdens in medical expenditure. There is currently no cure for OA once the bone damage is established. Unfortunately, the existing radiological examination is limited to grading the disease's severity and is insufficient to precisely diagnose OA, detect early OA or predict OA progression. Therefore, there is a pressing need to develop novel approaches in medical image analysis to detect subtle changes for identifying early OA development and rapid progressors. Recently, radiomics has emerged as a unique approach to extracting high-dimensional imaging features that quantitatively characterise visible or hidden information from routine medical images. Radiomics data mining via machine learning has empowered precise diagnoses and prognoses of disease, mainly in oncology. Mounting evidence has shown its great potential in aiding the diagnosis and contributing to the study of musculoskeletal diseases. This paper will summarise the current development of radiomics at the crossroads between engineering and medicine and discuss the application and perspectives of radiomics analysis for OA diagnosis and prognosis.

The translational potential of this article

Radiomics is a novel approach used in oncology, and it may also play an essential role in the diagnosis and prognosis of OA. By transforming medical images from qualitative interpretation to quantitative data, radiomics could be the solution for precise early OA detection, progression tracking, and treatment efficacy prediction. Since the application of radiomics in OA is still in the early stages and primarily focuses on fundamental studies, this review may inspire more explorations and bring more promising diagnoses, prognoses, and management results of OA.

Application of machine learning-based multi-sequence MRI radiomics in diagnosing anterior cruciate ligament tears

OBJECTIVE

To compare the diagnostic power among various machine learning algorithms utilizing multi-sequence magnetic resonance imaging (MRI) radiomics in detecting anterior cruciate ligament (ACL) tears. Additionally, this research aimed to create and validate the optimal diagnostic model.

METHODS

In this retrospective analysis, 526 patients were included, comprising 178 individuals with ACL tears and 348 with a normal ACL. Radiomics features were derived from multi-sequence MRI scans, encompassing T1-weighted imaging and proton density (PD)-weighted imaging. The process of selecting the most reliable radiomics features involved using interclass correlation coefficient (ICC) testing, t tests, and the least absolute shrinkage and selection operator (LASSO) technique. After the feature selection process, five machine learning classifiers were created. These classifiers comprised logistic regression (LR), support vector machine (SVM), K-nearest neighbors (KNN), light gradient boosting machine (LightGBM), and multilayer perceptron (MLP). A thorough performance evaluation was carried out, utilizing diverse metrics like the area under the receiver operating characteristic curve (ROC), specificity, accuracy, sensitivity positive predictive value, and negative predictive value. The classifier exhibiting the best performance was chosen. Subsequently, three models were developed: the PD model, the T1 model, and the combined model, all based on the optimal classifier. The diagnostic performance of these models was assessed by employing AUC values, calibration curves, and decision curve analysis.

RESULTS

Out of 2032 features, 48 features were selected. The SVM-based multi-sequence radiomics outperformed all others, achieving AUC values of 0.973 and 0.927, sensitivities of 0.933 and 0.857, and specificities of 0.930 and 0.829, in the training and validation cohorts, respectively.

CONCLUSION

The multi-sequence MRI radiomics model, which is based on machine learning, exhibits exceptional performance in diagnosing ACL tears. It provides valuable insights crucial for the diagnosis and treatment of knee joint injuries, serving as an accurate and objective supplementary diagnostic tool for clinical practitioners.

A radiomics nomogram based on MSCT and clinical factors can stratify fibrosis in inflammatory bowel disease

Intestinal fibrosis is one of the major complications of inflammatory bowel disease (IBD) and a pathological process that significantly impacts patient prognosis and treatment selection. Although current imaging assessment and clinical markers are widely used for the diagnosis and stratification of fibrosis, these methods suffer from subjectivity and limitations. In this study, we aim to develop a radiomics diagnostic model based on multi-slice computed tomography (MSCT) and clinical factors. MSCT images and relevant clinical data were collected from 218 IBD patients, and a large number of quantitative image features were extracted. Using these features, we constructed a radiomics model and transformed it into a user-friendly diagnostic nomogram. A nomogram was developed to predict fibrosis in IBD by integrating multiple factors. The nomogram exhibited favorable discriminative ability, with an AUC of 0.865 in the validation sets, surpassing both the logistic regression (LR) model (AUC = 0.821) and the clinical model (AUC = 0.602) in the test set. In the train set, the LR model achieved an AUC of 0.975, while the clinical model had an AUC of 0.735. The nomogram demonstrated superior performance with an AUC of 0.971, suggesting its potential as a valuable tool for predicting fibrosis in IBD and improving clinical decision-making. The radiomics nomogram, incorporating MSCT and clinical factors, demonstrates promise in stratifying fibrosis in IBD. The nomogram outperforms traditional clinical models and offers personalized risk assessment. However, further validation and addressing identified limitations are necessary to enhance its applicability.