18F-FDG PET/CT Radiomic Analysis with Machine Learning for Identifying Bone Marrow Involvement in the Patients with Suspected Relapsed Acute Leukemia

PET/CT in leukemia: utility and future directions

2-Deoxy-2-[ 18 F]fluoro- d -glucose PET/computed tomography ([ 18 F]FDG PET/CT) has proven to be a sensitive method for the detection and evaluation of hematologic malignancies, especially lymphoma. The increasing incidence and mortality rates of leukemia have raised significant concerns. Through the utilization of whole-body imaging, [ 18 F]FDG PET/CT provides a thorough assessment of the entire bone marrow, complementing the limited insights provided by biopsy samples. In this regard, [ 18 F]FDG PET/CT has the ability to assess diverse types of leukemia The utilization of [ 18 F]FDG PET/CT has been found to be effective in evaluating leukemia spread beyond the bone marrow, tracking disease relapse, identifying Richter's transformation, and assessing the inflammatory activity associated with acute graft versus host disease. However, its role in various clinical scenarios in leukemia remains unacknowledged. Despite their less common use, some novel PET/CT radiotracers are being researched for potential use in specific scenarios in leukemia patients. Therefore, the objectives of this review are to provide a thorough assessment of the current applications of [ 18 F]FDG PET/CT in the staging and monitoring of leukemia patients, as well as the potential for an expanding role of PET/CT in leukemia patients.

Applications of 18F-Fluorodesoxyglucose PET Imaging in Leukemia

The main finding that 18F-FDG PET imaging can reveal in patients with leukemias is the presence of bone marrow (BM) infiltration in both acute or chronic forms. This ability can influence and guide the use of BM biopsy but also assess to therapy response. Additionally 18F-FDG PET imaging has been reported as particularly useful for the diagnosis of leukemias in patients with non specific symptoms. In the case of acute leukemias it revealed also a role for the evaluation of extramedullary forms while in the case of chronic forms a role for the assessment of Richter transformation has been reported.

Identifying the risk factors of ICU-acquired fungal infections: clinical evidence from using machine learning

Background

Fungal infections are associated with high morbidity and mortality in the intensive care unit (ICU), but their diagnosis is difficult. In this study, machine learning was applied to design and define the predictive model of ICU-acquired fungi (ICU-AF) in the early stage of fungal infections using Random Forest.

Objectives

This study aimed to provide evidence for the early warning and management of fungal infections.

Methods

We analyzed the data of patients with culture-positive fungi during their admission to seven ICUs of the First Affiliated Hospital of Chongqing Medical University from January 1, 2015, to December 31, 2019. Patients whose first culture was positive for fungi longer than 48 h after ICU admission were included in the ICU-AF cohort. A predictive model of ICU-AF was obtained using the Least Absolute Shrinkage and Selection Operator and machine learning, and the relationship between the features within the model and the disease severity and mortality of patients was analyzed. Finally, the relationships between the ICU-AF model, antifungal therapy and empirical antifungal therapy were analyzed.

Results

A total of 1,434 cases were included finally. We used lasso dimensionality reduction for all features and selected six features with importance ≥0.05 in the optimal model, namely, times of arterial catheter, enteral nutrition, corticosteroids, broadspectrum antibiotics, urinary catheter, and invasive mechanical ventilation. The area under the curve of the model for predicting ICU-AF was 0.981 in the test set, with a sensitivity of 0.960 and specificity of 0.990. The times of arterial catheter (p = 0.011, OR = 1.057, 95% CI = 1.053-1.104) and invasive mechanical ventilation (p = 0.007, OR = 1.056, 95%CI = 1.015-1.098) were independent risk factors for antifungal therapy in ICU-AF. The times of arterial catheter (p = 0.004, OR = 1.098, 95%CI = 0.855-0.970) were an independent risk factor for empirical antifungal therapy.

Conclusion

The most important risk factors for ICU-AF are the six time-related features of clinical parameters (arterial catheter, enteral nutrition, corticosteroids, broadspectrum antibiotics, urinary catheter, and invasive mechanical ventilation), which provide early warning for the occurrence of fungal infection. Furthermore, this model can help ICU physicians to assess whether empiric antifungal therapy should be administered to ICU patients who are susceptible to fungal infections.

Prediction of coronary artery lesions in children with Kawasaki syndrome based on machine learning

OBJECTIVE

Kawasaki syndrome (KS) is an acute vasculitis that affects children < 5 years of age and leads to coronary artery lesions (CAL) in about 20-25% of untreated cases. Machine learning (ML) is a branch of artificial intelligence (AI) that integrates complex data sets on a large scale and uses huge data to predict future events. The purpose of the present study was to use ML to present the model for early risk assessment of CAL in children with KS by different algorithms.

METHODS

A total of 158 children were enrolled from Women and Children's Hospital, Qingdao University, and divided into 70-30% as the training sets and the test sets for modeling and validation studies. There are several classifiers are constructed for models including the random forest (RF), the logistic regression (LR), and the eXtreme Gradient Boosting (XGBoost). Data preprocessing is analyzed before applying the classifiers to modeling. To avoid the problem of overfitting, the 5-fold cross validation method was used throughout all the data.

RESULTS

The area under the curve (AUC) of the RF model was 0.925 according to the validation of the test set. The average accuracy was 0.930 (95% CI, 0.905 to 0.956). The AUC of the LG model was 0.888 and the average accuracy was 0.893 (95% CI, 0,837 to 0.950). The AUC of the XGBoost model was 0.879 and the average accuracy was 0.935 (95% CI, 0.891 to 0.980).

CONCLUSION

The RF algorithm was used in the present study to construct a prediction model for CAL effectively, with an accuracy of 0.930 and AUC of 0.925. The novel model established by ML may help guide clinicians in the initial decision to make a more aggressive initial anti-inflammatory therapy. Due to the limitations of external validation and regional population characteristics, additional research is required to initiate a further application in the clinic.

The added diagnostic value of 18 F-FDG PET/CT radiomic analysis in multiple myeloma patients with negative visual analysis

PURPOSE

A small number of patients diagnosed with multiple myeloma (MM) by bone marrow aspiration reported as being disease-free on 18 F-FDG PET/CT. We aim to evaluate the diagnostic value of radiomics approach in patients with MM who were negative by visual analysis.

MATERIALS AND METHODS

Thirty-three patients judged negative by visual analysis were assigned to the MM group. Contemporaneous 31 disease-free patients served as the control group. 70% of the whole data set was used as training set (23 from MM group and 22 from control group) and 30% as testing set (10 from MM group and 9 from control group). Axial skeleton volumes were automatically segmented and high-dimensional imaging features were extracted from PET and CT. The unsupervised machine learning method was used to filter and reduce the dimensions of the extracted features. Random forest was used to construct the prediction model and then validated with 10-fold cross-validation and evaluated on the independent testing set.

RESULTS

One thousand seven hundred two quantitative features were extracted from PET and CT. Of those, three first-order and one high-order imaging features were uncorrelated. With the cross-validation on the training group, the sensitivity, specificity, accuracy and area under the curve of random forest were 0.850, 0.792, 0.818 and 0.894, respectively. On the independent testing set, the accuracy of the model was 0.850 and the area under the curve was 0.909.

CONCLUSION

Radiomic analysis based on 18 F-FDG PET/CT using machine learning model provides a quantitative, objective and efficient mechanism for diagnosing patients with MM who were negative by visual analysis.

Clinical application of 18F-fluorodeoxyglucose positron emission tomography/computed tomography radiomics-based machine learning analyses in the field of oncology

Machine learning (ML) analyses using 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET)/computed tomography (CT) radiomics features have been applied in the field of oncology. The current review aimed to summarize the current clinical articles about 18F-FDG PET/CT radiomics-based ML analyses to solve issues in classifying or constructing prediction models for several types of tumors. In these studies, lung and mediastinal tumors were the most commonly evaluated lesions, followed by lymphatic, abdominal, head and neck, breast, gynecological, and other types of tumors. Previous studies have commonly shown that 18F-FDG PET radiomics-based ML analysis has good performance in differentiating benign from malignant tumors, predicting tumor characteristics and stage, therapeutic response, and prognosis by examining significant differences in the area under the receiver operating characteristic curves, accuracies, or concordance indices (> 0.70). However, these studies have reported several ML algorithms. Moreover, different ML models have been applied for the same purpose. Thus, various procedures were used in 18F-FDG PET/CT radiomics-based ML analysis in oncology, and 18F-FDG PET/CT radiomics-based ML models, which are easy and universally applied in clinical practice, would be expected to be established.

Nuclear medicine radiomics in digestive system tumors: Concept, applications, challenges, and future perspectives

Radiomics aims to develop novel biomarkers and provide relevant deeper subvisual information about pathology, immunophenotype, and tumor microenvironment. It uses automated or semiautomated quantitative analysis of high-dimensional images to improve characterization, diagnosis, and prognosis. Recent years have seen a rapid increase in radiomics applications in nuclear medicine, leading to some promising research results in digestive system oncology, which have been driven by big data analysis and the development of artificial intelligence. Although radiomics advances one step further toward the non-invasive precision medical analysis, it is still a step away from clinical application and faces many challenges. This review article summarizes the available literature on digestive system tumors regarding radiomics in nuclear medicine. First, we describe the workflow and steps involved in radiomics analysis. Subsequently, we discuss the progress in clinical application regarding the utilization of radiomics for distinguishing between various diseases and evaluating their prognosis, and demonstrate how radiomics advances this field. Finally, we offer our viewpoint on how the field can progress by addressing the challenges facing clinical implementation.

Machine‐learning radiomics to predict bone marrow metastasis of neuroblastoma using magnetic resonance imaging

Background

Neuroblastoma is one common pediatric malignancy notorious for high temporal and spatial heterogeneities. More than half of its patients develop distant metastases involving vascularized organs, especially the bone marrow. It is thus necessary to have an economical, noninvasive method without much radiation for follow‐ups. Radiomics has been used in many cancers to assist accurate diagnosis but not yet in bone marrow metastasis in neuroblastoma.

Methods

A total of 182 patients with neuroblastoma were retrospectively collected and randomly divided into the training and validation sets. Five‐hundred and seventy‐two radiomics features were extracted from magnetic resonance imaging, among which 41 significant ones were selected via T‐test for model development. We attempted 13 machine‐learning algorithms and eventually chose three best‐performed models. The integrative performance evaluations are based on the area under the curves (AUCs), calibration curves, risk deciles plots, and other indexes.

Results

Extreme gradient boosting, random forest (RF), and adaptive boosting were the top three to predict bone marrow metastases in neuroblastoma while RF was the most accurate one. Its AUC was 0.90 (0.86–0.93), F1 score was 0.82, sensitivity was 0.76, and negative predictive value was 0.79 in the training set. The values were 0.82 (0.71–0.93), 0.80, 0.75, and 0.92 in the validation set, respectively.

Conclusions

Radiomics models are likely to contribute more to metastatic diagnoses and the formulation of personalized healthcare strategies in clinics. It has great potential of being a revolutionary method to replace traditional interventions in the future.

Correlations between glucose metabolism of bone marrow on 18 F-fluoro-D-glucose PET/computed tomography and hematopoietic cell populations in autoimmune diseases

PURPOSE

This study aims to investigate which hematopoieticcell populations, clinical factors, and laboratory values are associated with FDG uptake in bone marrow (BM) on FDG PET/CT in patients with autoimmune diseases.

METHODS

Forty-six patients with autoimmune disease who underwent FDG PET/CT and BM aspiration (BMA) between 2017 and 2022 were enrolled. The max and mean standard uptake values (SUVmax and SUVmean, SUVs) of FDG in BM, liver, and spleen were measured, and the bone marrow-to-liver SUVs ratios (BLRmax and BLRmean, BLRs) and spleen-to-liver SUVs ratios (SLRmax and SLRmean, SLRs) were calculated. BMA and clinical and laboratory parameters were collected and evaluated for association with BLRs and SLRs.

RESULTS

The patients were divided into the Grade II group (20; 43.5%) and Grade III groups (26; 56.5%) according to hemopoietic activity. The BLRmax ( P = 0.021), proportion of granulocytes ( P = 0.011), metamyelocytes ( P = 0.009), myelocytes ( P = 0.024), and monocytes ( P = 0.037) in BM were significantly higher in the Grade II group. Multivariate (stepwise) linear regression analyses showed that the proportion of granulocytes in BM was the strongest and only independent factor ( P < 0.0001) associated with BLRmax with an adjusted R2 of 0.431 in model 1. In model 2, ferritin ( P = 0.018), CRP ( P = 0.025), and the proportion of metamyelocytes ( P = 0.043) in BM were correlated with BLRmax with an adjusted R2 of 0.414.

CONCLUSION

The FDG uptake in BM is associated with hemopoietic activity and is regulated by hyperplastic granulocytes, particularly immature metamyelocytes, in patients with autoimmune diseases. Glucose metabolism in the BM correlates with the severity of systemic inflammation.

A radiomics approach for predicting acute hematologic toxicity in patients with cervical or endometrial cancer undergoing external-beam radiotherapy

PURPOSE

This study is purposed to establish a predictive model for acute severe hematologic toxicity (HT) during radiotherapy in patients with cervical or endometrial cancer and investigate whether the integration of clinical features and computed tomography (CT) radiomics features of the pelvic bone marrow (BM) could define a more precise model.

METHODS

A total of 207 patients with cervical or endometrial cancer from three cohorts were retrospectively included in this study. Forty-one clinical variables and 2226 pelvic BM radiomic features that were extracted from planning CT scans were included in the model construction. Following feature selection, model training was performed on the clinical and radiomics features via machine learning, respectively. The radiomics score, which was the output of the final radiomics model, was integrated with the variables that were selected by the clinical model to construct a combined model. The performance of the models was evaluated using the area under the receiver operating characteristic curve (AUC).

RESULTS

The best-performing prediction model comprised two clinical features (FIGO stage and cycles of postoperative chemotherapy) and radiomics score and achieved an AUC of 0.88 (95% CI, 0.81-0.93) in the training set, 0.80 (95% CI, 0.62-0.92) in the internal-test set and 0.85 (95% CI, 0.71-0.94) in the external-test dataset.

CONCLUSION

The proposed model which incorporates radiomics signature and clinical factors outperforms the models based on clinical or radiomics features alone in terms of the AUC. The value of the pelvic BM radiomics in chemoradiotherapy-induced HT is worthy of further investigation.

The application of machine learning to imaging in hematological oncology: A scoping review

Background

Here, we conducted a scoping review to (i) establish which machine learning (ML) methods have been applied to hematological malignancy imaging; (ii) establish how ML is being applied to hematological cancer radiology; and (iii) identify addressable research gaps.

Methods

The review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Extension for Scoping Reviews guidelines. The inclusion criteria were (i) pediatric and adult patients with suspected or confirmed hematological malignancy undergoing imaging (population); (ii) any study using ML techniques to derive models using radiological images to apply to the clinical management of these patients (concept); and (iii) original research articles conducted in any setting globally (context). Quality Assessment of Diagnostic Accuracy Studies 2 criteria were used to assess diagnostic and segmentation studies, while the Newcastle-Ottawa scale was used to assess the quality of observational studies.

Results

Of 53 eligible studies, 33 applied diverse ML techniques to diagnose hematological malignancies or to differentiate them from other diseases, especially discriminating gliomas from primary central nervous system lymphomas (n=18); 11 applied ML to segmentation tasks, while 9 applied ML to prognostication or predicting therapeutic responses, especially for diffuse large B-cell lymphoma. All studies reported discrimination statistics, but no study calculated calibration statistics. Every diagnostic/segmentation study had a high risk of bias due to their case-control design; many studies failed to provide adequate details of the reference standard; and only a few studies used independent validation.

Conclusion

To deliver validated ML-based models to radiologists managing hematological malignancies, future studies should (i) adhere to standardized, high-quality reporting guidelines such as the Checklist for Artificial Intelligence in Medical Imaging; (ii) validate models in independent cohorts; (ii) standardize volume segmentation methods for segmentation tasks; (iv) establish comprehensive prospective studies that include different tumor grades, comparisons with radiologists, optimal imaging modalities, sequences, and planes; (v) include side-by-side comparisons of different methods; and (vi) include low- and middle-income countries in multicentric studies to enhance generalizability and reduce inequity.

Bone marrow segmentation and radiomics analysis of [18F]FDG PET/CT images for measurable residual disease assessment in multiple myeloma

BACKGROUND AND OBJECTIVES

The last few years have been crucial in defining the most appropriate way to quantitatively assess [¹⁸F]FDG PET images in Multiple Myeloma (MM) patients to detect persistent tumor burden. The visual evaluation of images complements the assessment of Measurable Residual Disease (MRD) in bone marrow samples by multiparameter flow cytometry (MFC) or next-generation sequencing (NGS). The aim of this study was to quantify MRD by analyzing quantitative and texture [¹⁸F]FDG PET features.

METHODS

Whole body [¹⁸F]FDG PET of 39 patients with newly diagnosed MM were included in the database, and visually evaluated by experts in nuclear medicine. A segmentation methodology of the skeleton from CT images and an additional manual segmentation tool were proposed, implemented in a software solution including a graphical user interface. Both the compact bone and the spinal canal were removed from the segmentation to obtain only the bone marrow mask. SUV metrics, GLCM, GLRLM, and NGTDM parameters were extracted from the PET images and evaluated by Mann-Whitney U-tests and Spearman ρ rank correlation as valuable features differentiating PET+/PET- and MFC+/MFC- groups. Seven machine learning algorithms were applied for evaluating the classification performance of the extracted features.

RESULTS

Quantitative analysis for PET+/PET- differentiating demonstrated to be significant for most of the variables assessed with Mann-Whitney U-test such as Variance, Energy, and Entropy (p-value = 0.001). Moreover, the quantitative analysis with a balanced database evaluated by Mann-Whitney U-test revealed in even better results with 19 features with p-values < 0.001. On the other hand, radiomics analysis for MFC+/MFC- differentiating demonstrated the necessity of combining MFC evaluation with [¹⁸F]FDG PET assessment in the MRD diagnosis. Machine learning algorithms using the image features for the PET+/PET- classification demonstrated high performance metrics but decreasing for the MFC+/MFC- classification.

CONCLUSIONS

A proof-of-concept for the extraction and evaluation of bone marrow radiomics features of [¹⁸F]FDG PET images was proposed and implemented. The validation showed the possible use of these features for the image-based assessment of MRD.

Diagnostic Value of 18F-FDG PET/CT-Based Radiomics Nomogram in Bone Marrow Involvement of Pediatric Neuroblastoma

OBJECTIVES

To develop and validate an ¹⁸F-FDG PET/CT-based radiomics nomogram and evaluate the value of the ¹⁸F-FDG PET/CT-based radiomics nomogram for the diagnosis of bone marrow involvement (BMI) in pediatric neuroblastoma.

MATERIALS AND METHODS

A total of 144 patients with neuroblastoma (100 in the training cohort and 44 in the validation cohort) were retrospectively included. The PET/CT images of patients were visually assessed. The results of bone marrow aspirates or biopsies were used as the gold standard for BMI. Radiomics features and conventional PET parameters were extracted using the 3D slicer. Features were selected by the least absolute shrinkage and selection operator regression, and radiomics signature was constructed. Univariate and multivariate logistic regression analyses were applied to identify the independent clinical risk factors and construct the clinical model. Other different models, including the conventional PET model, combined PET-clinical model and combined radiomics model, were built using logistic regression. The combined radiomics model was based on clinical factors, conventional PET parameters and radiomics signature, which was presented as a radiomics nomogram. The diagnostic performance of the different models was evaluated by receiver operating characteristic (ROC) curves and decision curve analysis (DCA).

RESULTS

By visual assessment, BMI was observed in 80 patients. Four conventional PET parameters (SUVmax, SUVmean, metabolic tumor volume, and total lesion glycolysis) were extracted. And 15 radiomics features were selected to build the radiomics signature. The 11q aberration, neuron-specific enolase and vanillylmandelic acid were identified as the independent clinical risk factors to establish the clinical model. The radiomics nomogram incorporating the radiomics signature, the independent clinical risk factors and SUVmean demonstrated the best diagnostic value for identifying BMI, with an area under the curve (AUC) of 0.963 and 0.931 in the training and validation cohorts, respectively. And the DCA demonstrated that the radiomics nomogram was clinically useful.

CONCLUSION

The ¹⁸F-FDG PET/CT-based radiomics nomogram which incorporates radiomics signature, independent clinical risk factors and conventional PET parameters could improve the diagnostic performance for BMI of pediatric neuroblastoma without additional medical costs and radiation exposure.

A Review of Artificial Intelligence Applications in Hematology Management: Current Practices and Future Prospects

Background

Machine learning (ML) and deep learning (DL) methods have recently garnered a great deal of attention in the field of cancer research by making a noticeable contribution to the growth of predictive medicine and modern oncological practices. Considerable focus has been particularly directed toward hematologic malignancies because of the complexity in detecting early symptoms. Many patients with blood cancer do not get properly diagnosed until their cancer has reached an advanced stage with limited treatment prospects. Hence, the state-of-the-art revolves around the latest artificial intelligence (AI) applications in hematology management.

Objective

This comprehensive review provides an in-depth analysis of the current AI practices in the field of hematology. Our objective is to explore the ML and DL applications in blood cancer research, with a special focus on the type of hematologic malignancies and the patient’s cancer stage to determine future research directions in blood cancer.

Methods

We searched a set of recognized databases (Scopus, Springer, and Web of Science) using a selected number of keywords. We included studies written in English and published between 2015 and 2021. For each study, we identified the ML and DL techniques used and highlighted the performance of each model.

Results

Using the aforementioned inclusion criteria, the search resulted in 567 papers, of which 144 were selected for review.

Conclusions

The current literature suggests that the application of AI in the field of hematology has generated impressive results in the screening, diagnosis, and treatment stages. Nevertheless, optimizing the patient’s pathway to treatment requires a prior prediction of the malignancy based on the patient’s symptoms or blood records, which is an area that has still not been properly investigated.

Radiomic Analysis of Positron-Emission Tomography and Computed Tomography Images to Differentiate between Multiple Myeloma and Skeletal Metastases

Context

Multiple myeloma and extensive lytic skeletal metastases may appear similar on positron-emission tomography and computed tomography (PET-CT) in the absence of an obvious primary site or occult malignancy. Radiomic analysis extracts a large number of quantitative features from medical images with the potential to uncover disease characteristics below the human visual threshold.

Aim

This study aimed to evaluate the diagnostic capability of PET and CT radiomic features to differentiate skeletal metastases from multiple myeloma.

Settings and Design

Forty patients (20 histopathologically proven cases of multiple myeloma and 20 cases of a variety of bone metastases) underwent staging 18F-fluorodeoxyglucose PET-CT at our institute.

Methodology

A total of 138 PET and 138 CT radiomic features were extracted by manual semi-automatic segmentation and standardized. The original dataset was subject separately to receiver operating curve analysis and correlation matrix filtering. The former showed 16 CT and 19 PET parameters to be significantly related to the outcome at 5%, whereas the latter resulted in 16 CT and 14 PET features. Feature selection was done with 7 evaluators with stratified 10-fold cross-validation. The selected features of each evaluator were subject to 14 machine-learning algorithms. In view of small sample size, two approaches for model performance were adopted: The first using 10-fold stratified cross-validation and the second using independent random training and test samples (26:14). In both approaches, the highest area under the curve (AUC) values were selected for 5 CT and 5 PET features. These 10 features were combined and the same process was repeated.

Statistical Analysis Used

The quality of the performance of the models was assessed by MSE, RMSE, kappa statistic, AUC, area under the precision-recall curve, F-measure, and Matthews correlation coefficient.

Results

In the first approach, the highest AUC = 0.945 was seen with 5 CT parameters. In the second approach, the highest AUC = 0.9538 was seen with 4 CT and one PET parameter. CT neighborhood gray-level different matrix coarseness and CT gray-level run-length matrix LGRE were common parameters in both approaches. Comparison of AUC of the above models showed no significant difference (P = 0.9845). Feature selection by principal components analysis and feature classification by the multilayer perceptron machine-learning model using independent training and test samples yielded the overall highest AUC.

Conclusions

Machine-learning models using CT parameters were found to differentiate bone metastases from multiple myeloma better than models using PET parameters. Combined models using PET and CECT data showed better overall performance than models using only either PET or CECT data. Machine-learning models using independent training and test sets were performed on par with those using 10-fold stratified cross-validation with the former incorporating slightly more PET features. Certain first- and second-order CT and PET texture features contributed in differentiating these two conditions. Our findings suggested that, in general, metastases were finer in CT and PET texture and myelomas were more compact.

Radiomics in Oncological PET Imaging: A Systematic Review-Part 2, Infradiaphragmatic Cancers, Blood Malignancies, Melanoma and Musculoskeletal Cancers

The objective of this review was to summarize published radiomics studies dealing with infradiaphragmatic cancers, blood malignancies, melanoma, and musculoskeletal cancers, and assess their quality. PubMed database was searched from January 1990 to February 2022 for articles performing radiomics on PET imaging of at least 1 specified tumor type. Exclusion criteria includd: non-oncological studies; supradiaphragmatic tumors; reviews, comments, cases reports; phantom or animal studies; technical articles without a clinically oriented question; studies including <30 patients in the training cohort. The review database contained PMID, first author, year of publication, cancer type, number of patients, study design, independent validation cohort and objective. This database was completed twice by the same person; discrepant results were resolved by a third reading of the articles. A total of 162 studies met inclusion criteria; 61 (37.7%) studies included >100 patients, 13 (8.0%) were prospective and 61 (37.7%) used an independent validation set. The most represented cancers were esophagus, lymphoma, and cervical cancer (n = 24, n = 24 and n = 19 articles, respectively). Most studies focused on 18F-FDG, and prognostic and response to treatment objectives. Although radiomics and artificial intelligence are technically challenging, new contributions and guidelines help improving research quality over the years and pave the way toward personalized medicine.

Association of CT-Based Delta Radiomics Biomarker With Progression-Free Survival in Patients With Colorectal Liver Metastases Undergo Chemotherapy

Predicting the prognosis of patients in advance is conducive to providing personalized treatment for patients. Our aim was to predict the therapeutic efficacy and progression free survival (PFS) of patients with liver metastasis of colorectal cancer according to the changes of computed tomography (CT) radiomics before and after chemotherapy.

Methods

This retrospective study included 139 patients (397 lesions) with colorectal liver metastases who underwent neoadjuvant chemotherapy from April 2015 to April 2020. We divided the lesions into training cohort and testing cohort with a ratio of 7:3. Two - dimensional region of interest (ROI) was obtained by manually delineating the largest layers of each metastasis lesion. The expanded ROI (3 mm and 5 mm) were also included in the study to characterize microenvironment around tumor. For each of the ROI, 1,316 radiomics features were extracted from delineated plain scan, arterial, and venous phase CT images before and after neoadjuvant chemotherapy. Delta radiomics features were constructed by subtracting the radiomics features after treatment from the radiomics features before treatment. Univariate Cox regression and the Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression were applied in the training cohort to select the valuable features. Based on clinical characteristics and radiomics features, 7 Cox proportional-hazards model were constructed to predict the PFS of patients. C-index value and Kaplan Meier (KM) analysis were used to evaluate the efficacy of predicting PFS of these models. Moreover, the prediction performance of one-year PFS was also evaluated by area under the curve (AUC).

Results

Compared with the PreRad (Radiomics form pre-treatment CT images; C-index [95% confidence interval (CI)] in testing cohort: 0.614(0.552-0.675) and PostRad models (Radiomics form post-treatment CT images; 0.642(0.578-0.707), the delta model has better PFS prediction performance (Delta radiomics; 0.688(0.627-0.749). By incorporating clinical characteristics, CombDeltaRad obtains the best performance in both training cohort [C-index (95% CI): 0.802(0.772-0.832)] and the testing cohort (0.744(0.686-0.803). For 1-year PFS prediction, CombDeltaRad model obtained the best performance with AUC (95% CI) of 0.871(0.828-0.914) and 0.745 (0.651-0.838) in training cohort and testing cohort, respectively.

Conclusion

CT radiomics features have the potential to predict PFS in patients with colorectal cancer and liver metastasis who undergo neoadjuvant chemotherapy. By combining pre-treatment radiomics features, post-treatment radiomics features, and clinical characteristics better prediction results can be achieved.

Radiomic modeling to predict risk of vertebral compression fracture after stereotactic body radiation therapy for spinal metastases

OBJECTIVE

In the treatment of spinal metastases with stereotactic body radiation therapy (SBRT), vertebral compression fracture (VCF) is a common and potentially morbid complication. Better methods to identify patients at high risk of radiation-induced VCF are needed to evaluate prophylactic measures. Radiomic features from pretreatment imaging may be employed to more accurately predict VCF. The objective of this study was to develop and evaluate a machine learning model based on clinical characteristics and radiomic features from pretreatment imaging to predict the risk of VCF after SBRT for spinal metastases.

METHODS

Vertebral levels C2 through L5 containing metastases treated with SBRT were included if they were naive to prior surgery or radiation therapy, target delineation was based on consensus guidelines, and 1-year follow-up data were available. Clinical features, including characteristics of the patient, disease, and treatment, were obtained from chart review. Radiomic features were extracted from the planning target volume (PTV) on pretreatment CT and T1-weighted MRI. Clinical and radiomic features selected by least absolute shrinkage and selection operator (LASSO) regression were included in random forest classification models, which were trained to predict VCF within 1 year after SBRT. Model performance was assessed with leave-one-out cross-validation.

RESULTS

Within 1 year after SBRT, 15 of 95 vertebral levels included in the analysis demonstrated new or progressive VCF. Selected clinical features included BMI, performance status, total prescription dose, dose to 99% of the PTV, lumbar location, and 2 components of the Spine Instability Neoplastic Score (SINS): lytic tumor character and spinal misalignment. Selected radiomic features included 5 features from CT and 3 features from MRI. The best-performing classification model, derived from a combination of selected clinical and radiomic features, demonstrated a sensitivity of 0.844, specificity of 0.800, and area under the receiver operating characteristic (ROC) curve (AUC) of 0.878. This model was significantly more accurate than alternative models derived from only selected clinical features (AUC = 0.795, p = 0.048) or only components of the SINS (AUC = 0.579, p < 0.0001).

CONCLUSIONS

In the treatment of spinal metastases with SBRT, a machine learning model incorporating both clinical features and radiomic features from pretreatment imaging predicted VCF at 1 year after SBRT with excellent sensitivity and specificity, outperforming models developed from clinical features or components of the SINS alone. If validated, these findings may allow more judicious selection of patients for prophylactic interventions.

Comparison of FDG PET/CT and Bone Marrow Biopsy Results in Patients with Diffuse Large B Cell Lymphoma with Subgroup Analysis of PET Radiomics

Whether FDG PET/CT can replace bone marrow biopsy (BMBx) is undecided in patients with diffuse large B cell lymphoma (DLBCL). We compared the visual PET findings and PET radiomic features, with BMBx results. A total of 328 patients were included; 269 (82%) were PET-negative and 59 (18%) were PET-positive for bone lesions on visual assessment. A fair degree of agreement was present between PET and BMBx findings (ĸ = 0.362, p < 0.001). Bone involvement on PET/CT lead to stage IV in 12 patients, despite no other evidence of extranodal lesion. Of 35 discordant PET-positive and BMBx-negative cases, 22 (63%) had discrete bone uptake on PET/CT. A total of 144 patients were eligible for radiomic analysis, and two grey-level zone-length matrix derived parameters obtained from the iliac crests showed a trend for higher values in the BMBx-positive group compared to the BMBx-negative group (mean 436.6 ± 449.0 versus 227.2 ± 137.8, unadjusted p = 0.037 for high grey-level zone emphasis; mean 308.8 ± 394.4 versus 135.7 ± 97.2, unadjusted p = 0.048 for short-zone high grey-level emphasis), but statistical significance was not found after multiple comparison correction. Visual FDG PET/CT assessment and BMBx results were discordant in 17% of patients with newly diagnosed DLBCL, and the two tests are complementary in the evaluation of bone involvement.

Unsupervised Machine Learning-Based Analysis of Clinical Features, Bone Mineral Density Features and Medical Care Costs of Rotator Cuff Tears

Purpose

We aim to present unsupervised machine learning-based analysis of clinical features, bone mineral density (BMD) features, and medical care costs of Rotator cuff tears (RCT).

Patients and Methods

Fifty-three patients with RCT were reviewed, the clinical features, BMD features, and medical care costs were collected and analyzed by descriptive statistics. Furtherly, unsupervised machine learning (UML) algorithm was used for dimensionality reduction and cluster analysis of the RCT data.

Results

There were 26 males and 27 females. The patients were divided into four subgroups using the UML algorithm. There were significant differences among four subgroups regarding trauma exposure, full-thickness supraspinatus tendon tears, infraspinatus tendon tear, subscapularis tendon tear, BMD distribution, medial row anchors, lateral row anchors, total medical care costs, and consumables costs. We observed the highest frequency of trauma exposure, infraspinatus tendon tear, subscapularis tendon tear, osteoporosis, the highest number of medial row anchors, lateral row anchors, total medical care costs, and consumables costs in subgroup II.

Conclusion

The unsupervised machine learning-based analysis of RCT can provide clinically meaningful classification, which shows good interpretability and contribute to a better understanding of RCT. The significance of the results is limited due to the small number of samples, a larger follow-up study is needed to confirm the encouraging results.

Multi-parametric MRI phenotype with trustworthy machine learning for differentiating CNS demyelinating diseases

Background

Misdiagnosis of multiple sclerosis (MS) and neuromyelitis optica (NMO) may delay the treatment, resulting in poor prognosis. However, the precise identification of these two diseases is still challenging in clinical practice. We aimed to evaluate the value of quantitative radiomic features extracted from the brain white matter lesions for differential diagnosis of MS and NMO.

Methods

We recruited 116 CNS demyelinating patients including 78 MS, and 38 NMO. Three neuroradiologists performed visual differential diagnosis based on brain MRI for comparison purpose. A multi-level scheme was designed to harness the selection of discriminative and stable radiomics features extracted from brain while mater lesions in T1-MPRAGE, T2 sequences and clinical factors. Based on the imaging phenotype composed of the selected radiomic and clinical features, Multi-parametric Multivariate Random Forest (MM-RF) model was constructed and verified with both 10-fold cross-validation and independent testing. Result interpretation was provided to build trust in diagnostic decisions.

Results

Eighty-six patients were randomly selected to form the training set while the rest 30 patients for independent testing. On the training set, our MM-RF model achieved accuracy 0.849 and AUC 0.826 in 10-fold cross-validation, which were significantly higher than clinical visual analysis (0.709 and 0.683, p < 0.05). In the independent testing, the MM-RF model achieved AUC 0.902, accuracy 0.871, sensitivity 0.873, specificity 0.869, respectively. Furthermore, age, sex and EDSS were found mildly correlated with the radiomic features (p of all < 0.05).

Conclusions

Multi-parametric radiomic features have potential as practical quantitative imaging biomarkers for differentiating MS from NMO.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-021-03015-w.

Radiomics in the Setting of Neoadjuvant Radiotherapy: A New Approach for Tailored Treatment

Simple Summary

This review based on a literature search aims at showing the impact of Texture Analysis in the prediction of response to neoadjuvant radiotherapy and/or chemoradiotherapy. The manuscript explores radiomics approaches in different fields of neoadjuvant radiotherapy, including esophageal cancer, lung cancer, sarcoma and rectal cancer in order to shed a light in the setting of neoadjuvant radiotherapy that can be used to tailor the best subsequent therapeutical strategy.

Abstract

Introduction: Neoadjuvant radiotherapy is currently used mainly in locally advanced rectal cancer and sarcoma and in a subset of non-small cell lung cancer and esophageal cancer, whereas in other diseases it is under investigation. The evaluation of the efficacy of the induction strategy is made possible by performing imaging investigations before and after the neoadjuvant therapy and is usually challenging. In the last decade, texture analysis (TA) has been developed to help the radiologist to quantify and identify the parameters related to tumor heterogeneity, which cannot be appreciated by the naked eye. The aim of this narrative is to review the impact of TA on the prediction of response to neoadjuvant radiotherapy and or chemoradiotherapy. Materials and Methods: Key references were derived from a PubMed query. Hand searching and ClinicalTrials.gov were also used. Results: This paper contains a narrative report and a critical discussion of radiomics approaches in different fields of neoadjuvant radiotherapy, including esophageal cancer, lung cancer, sarcoma, and rectal cancer. Conclusions: Radiomics can shed a light on the setting of neoadjuvant therapies that can be used to tailor subsequent approaches or even to avoid surgery in the future. At the same, these results need to be validated in prospective and multicenter trials.

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

Purpose

We aim to present an unsupervised machine learning application in anterior cruciate ligament (ACL) rupture and evaluate whether supervised machine learning-derived radiomics features enable prediction of ACL rupture accurately.

Patients and Methods

Sixty-eight patients were reviewed. Their demographic features were recorded, radiomics features were extracted, and the input dataset was defined as a collection of demographic features and radiomics features. The input dataset was automatically classified by the unsupervised machine learning algorithm. Then, we used a supervised machine learning algorithm to construct a radiomics model. The t-test and least absolute shrinkage and selection operator (LASSO) method were used for feature selection, random forest and support vector machine (SVM) were used as machine learning classifiers. For each model, the sensitivity, specificity, accuracy, and the area under the curve (AUC) of receiver operating characteristic (ROC) curves were calculated to evaluate model performance.

Results

In total, 5 demographic features were recorded and 106 radiomics features were extracted. By applying the unsupervised machine learning algorithm, patients were divided into 5 groups. Group 5 had the highest incidence of ACL rupture and left knee involvement. There were significant differences in left knee involvement among the groups. Forty-three radiomics features were extracted using t-test and 7 radiomics features were extracted using LASSO method. We found that the combination of LASSO selection method and random forest classifier has the highest sensitivity, specificity, accuracy, and AUC. The 7 radiomics features extracted by LASSO method were potential predictors for ACL rupture.

Conclusion

We validated the clinical application of unsupervised machine learning involving ACL rupture. Moreover, we found 7 radiomics features which were potential predictors for ACL rupture. The study indicated that radiomics could be a valuable method in the prediction of ACL rupture.

Improved 18-FDG PET/CT diagnosis of multiple myeloma diffuse disease by radiomics analysis

OBJECTIVES

In multiple myeloma, the diagnosis of diffuse bone marrow infiltration on 18-FDG PET/CT can be challenging. We aimed to develop a PET/CT radiomics-based model that could improve the diagnosis of multiple myeloma diffuse disease on 18-FDG PET/CT.

METHODS

We prospectively performed PET/CT and whole-body diffusion-weighted MRI in 30 newly diagnosed multiple myeloma. MRI was the reference standard for diffuse disease assessment. Twenty patients were randomly assigned to a training set and 10 to an independent test set. Visual analysis of PET/CT was performed by two nuclear medicine physicians. Spine volumes were automatically segmented, and a total of 174 Imaging Biomarker Standardisation Initiative-compliant radiomics features were extracted from PET and CT. Selection of best features was performed with random forest features importance and correlation analysis. Machine-learning algorithms were trained on the selected features with cross-validation and evaluated on the independent test set.

RESULTS

Out of the 30 patients, 18 had established diffuse disease on MRI. The sensitivity, specificity and accuracy of visual analysis were 67, 75 and 70%, respectively, with a moderate kappa coefficient of agreement of 0.6. Five radiomics features were selected. On the training set, random forest classifier reached a sensitivity, specificity and accuracy of 93, 86 and 91%, respectively, with an area under the curve of 0.90 (95% confidence interval, 0.89-0.91). On the independent test set, the model achieved an accuracy of 80%.

CONCLUSIONS

Radiomics analysis of 18-FDG PET/CT images with machine-learning overcame the limitations of visual analysis, providing a highly accurate and more reliable diagnosis of diffuse bone marrow infiltration in multiple myeloma patients.

Automatic identification of myopic maculopathy related imaging features in optic disc region via machine learning methods

BACKGROUND

Myopic maculopathy (MM) is the most serious and irreversible complication of pathologic myopia, which is a major cause of visual impairment and blindness. Clinic proposed limited number of factors related to MM. To explore additional features strongly related with MM from optic disc region, we employ a machine learning based radiomics analysis method, which could explore and quantify more hidden or imperceptible MM-related features to the naked eyes and contribute to a more comprehensive understanding of MM and therefore may assist to distinguish the high-risk population in an early stage.

METHODS

A total of 457 eyes (313 patients) were enrolled and were divided into severe MM group and without severe MM group. Radiomics analysis was applied to depict features significantly correlated with severe MM from optic disc region. Receiver Operating Characteristic were used to evaluate these features' performance of classifying severe MM.

RESULTS

Eight new MM-related image features were discovered from the optic disc region, which described the shapes, textural patterns and intensity distributions of optic disc region. Compared with clinically reported MM-related features, these newly discovered features exhibited better abilities on severe MM classification. And the mean values of most features were markedly changed between patients with peripapillary diffuse chorioretinal atrophy (PDCA) and macular diffuse chorioretinal atrophy (MDCA).

CONCLUSIONS

Machine learning and radiomics method are useful tools for mining more MM-related features from the optic disc region, by which complex or even hidden MM-related features can be discovered and decoded. In this paper, eight new MM-related image features were found, which would be useful for further quantitative study of MM-progression. As a nontrivial byproduct, marked changes between PDCA and MDCA was discovered by both new image features and clinic features.

Case Report: Asymmetric Bone Marrow Involvement in Patients With Acute Leukemia After Allogeneic Hematopoietic Stem Cell Transplantation

After allogeneic hematopoietic stem cell transplantation (allo-HSCT), acute leukemia relapse is common, and asymmetric bone marrow recurrence hasn't been reported. Because the anatomical distribution of acute leukemia clones in the bone marrow after allo-HSCT is presumed to be diffuse, bone marrow aspirations are performed in single site. The first case was a 20-year-old man who was diagnosed with acute myelomonocytic leukemia and received haploidentical allo-HSCT. Routine bone marrow biopsy of his left posterior iliac bone marrow showed 52% leukemia blasts, while the right side had 0% blasts 10 days later. The second case was a 23-year-old woman who was diagnosed with acute B lymphoblastic leukemia and received HLA-identical sibling allo-HSCT. Although 62% of blasts were found in her left iliac marrow on day +122, 0% of blasts were found on a sample obtained from the right iliac crest on day +128. Bilateral iliac bone marrow pathology and whole-body ¹⁸F-FDG PET/CT scans confirmed that the leukemic infiltration in her bone marrow was asymmetric. To our knowledge, these are the first case reports of asymmetric bone marrow infiltration of blasts in acute leukemia patients after allo-HSCT. Bilateral posterior iliac crest aspirations or ¹⁸F-FDG-PET/CT scans may help distinguish such involvement.

A Systematic Review of PET Textural Analysis and Radiomics in Cancer

Background: Although many works have supported the utility of PET radiomics, several authors have raised concerns over the robustness and replicability of the results. This study aimed to perform a systematic review on the topic of PET radiomics and the used methodologies. Methods: PubMed was searched up to 15 October 2020. Original research articles based on human data specifying at least one tumor type and PET image were included, excluding those that apply only first-order statistics and those including fewer than 20 patients. Each publication, cancer type, objective and several methodological parameters (number of patients and features, validation approach, among other things) were extracted. Results: A total of 290 studies were included. Lung (28%) and head and neck (24%) were the most studied cancers. The most common objective was prognosis/treatment response (46%), followed by diagnosis/staging (21%), tumor characterization (18%) and technical evaluations (15%). The average number of patients included was 114 (median = 71; range 20–1419), and the average number of high-order features calculated per study was 31 (median = 26, range 1–286). Conclusions: PET radiomics is a promising field, but the number of patients in most publications is insufficient, and very few papers perform in-depth validations. The role of standardization initiatives will be crucial in the upcoming years.

Machine learning for predictive data analytics in medicine: A review illustrated by cardiovascular and nuclear medicine examples

The evidence-based medicine allows the physician to evaluate the risk-benefit ratio of a treatment through setting and data. Risk-based choices can be done by the doctor using different information. With the emergence of new technologies, a large amount of data is recorded offering interesting perspectives with machine learning for predictive data analytics. Machine learning is an ensemble of methods that process data to model a learning problem. Supervised machine learning algorithms consist in using annotated data to construct the model. This category allows to solve prediction data analytics problems. In this paper, we detail the use of supervised machine learning algorithms for predictive data analytics problems in medicine. In the medical field, data can be split into two categories: medical images and other data. For brevity, our review deals with any kind of medical data excluding images. In this article, we offer a discussion around four supervised machine learning approaches: information-based, similarity-based, probability-based and error-based approaches. Each method is illustrated with detailed cardiovascular and nuclear medicine examples. Our review shows that model ensemble (ME) and support vector machine (SVM) methods are the most popular. SVM, ME and artificial neural networks often lead to better results than those given by other algorithms. In the coming years, more studies, more data, more tools and more methods will, for sure, be proposed.

Artificial Intelligence in Hematology: Current Challenges and Opportunities

PURPOSE OF REVIEW

Artificial intelligence (AI), and in particular its subcategory machine learning, is finding an increasing number of applications in medicine, driven in large part by an abundance of data and powerful, accessible tools that have made AI accessible to a larger circle of investigators.

RECENT FINDINGS

AI has been employed in the analysis of hematopathological, radiographic, laboratory, genomic, pharmacological, and chemical data to better inform diagnosis, prognosis, treatment planning, and foundational knowledge related to benign and malignant hematology. As more widespread implementation of clinical AI nears, attention has also turned to the effects this will have on other areas in medicine. AI offers many promising tools to clinicians broadly, and specifically in the practice of hematology. Ongoing research into its various applications will likely result in an increasing utilization of AI by a broader swath of clinicians.

Artificial intelligence applications for oncological positron emission tomography imaging

Positron emission tomography (PET), a functional and dynamic molecular imaging technique, is generally used to reveal tumors' biological behavior. Radiomics allows a high-throughput extraction of multiple features from images with artificial intelligence (AI) approaches and develops rapidly worldwide. Quantitative and objective features of medical images have been explored to recognize reliable biomarkers, with the development of PET radiomics. This paper will review the current clinical exploration of PET-based classical machine learning and deep learning methods, including disease diagnosis, the prediction of histological subtype, gene mutation status, tumor metastasis, tumor relapse, therapeutic side effects, therapeutic intervention and evaluation of prognosis. The applications of AI in oncology will be mainly discussed. The image-guided biopsy or surgery assisted by PET-based AI will be introduced as well. This paper aims to present the applications and methods of AI for PET imaging, which may offer important details for further clinical studies. Relevant precautions are put forward and future research directions are suggested.

Radiomic Model for Distinguishing Dissecting Aneurysms from Complicated Saccular Aneurysms on high-Resolution Magnetic Resonance Imaging

OBJECTIVE

To build radiomic model in differentiating dissecting aneurysm (DA) from complicated saccular aneurysm (SA) based on high-resolution magnetic resonance imaging (HR-MRI) through machine-learning algorithm.

METHODS

Overall, 851 radiomic features from 77 cases were retrospectively analyzed, and the ElasticNet algorithm was used to build the radiomic model. A clinico-radiological model using clinical features and conventional MRI findings was also built. An integrated model was then built by incorporating the radiomic model and clinico-radiological model. The diagnostic abilities of these models were evaluated using leave one out cross validation and quantified using the receiver operating characteristic (ROC) analysis. The diagnostic performance of radiologists was also evaluated for comparison.

RESULTS

Five features were used to form the radiomic model, which yielded an area under the ROC curve (AUC) of 0.912 (95 % CI 0.846-0.976), sensitivity of 0.852, and specificity of 0.861. The radiomic model achieved a better diagnostic performance than the clinico-radiological model (AUC=0.743, 95 % CI 0.623-0.862), integrated model (AUC=0.888, 95 % CI 0.811-0.965), and even many radiologists.

CONCLUSION

Radiomic features derived from HR-MRI can reliably be used to build a radiomic model for effectively differentiating between DA and complicated SA, and it can provide an objective basis for the selection of clinical treatment plan.

Applications of PET in Diagnosis and Prognosis of Leukemia

As a malignant hematopoietic stem cell disease, leukemia remains life-threatening due to its increasing incidence rate and mortality rate. Therefore, its early diagnosis and treatment play a very important role. In the present work, we systematically reviewed the current applications and future directions of positron emission tomography (PET) in patients with leukemia, especially ¹⁸F-FDG PET/CT. As a useful imaging approach, PET significantly contributes to the diagnosis and treatment of different types of leukemia, especially in the evaluation of extramedullary infiltration, monitoring of leukemia relapse, detection of Richter’s transformation (RT), and assessment of the inflammatory activity associated with acute graft versus host disease. Future investigations should be focused on the potential of PET/CT in the prediction of clinical outcomes in patients with leukemia and the utility of novel radiotracers.

Histologic subtype classification of non-small cell lung cancer using PET/CT images

PURPOSES

To evaluate the capability of PET/CT images for differentiating the histologic subtypes of non-small cell lung cancer (NSCLC) and to identify the optimal model from radiomics-based machine learning/deep learning algorithms.

METHODS

In this study, 867 patients with adenocarcinoma (ADC) and 552 patients with squamous cell carcinoma (SCC) were retrospectively analysed. A stratified random sample of 283 patients (20%) was used as the testing set (173 ADC and 110 SCC); the remaining data were used as the training set. A total of 688 features were extracted from each outlined tumour region. Ten feature selection techniques, ten machine learning (ML) models and the VGG16 deep learning (DL) algorithm were evaluated to construct an optimal classification model for the differential diagnosis of ADC and SCC. Tenfold cross-validation and grid search technique were employed to evaluate and optimize the model hyperparameters on the training dataset. The area under the receiver operating characteristic curve (AUROC), accuracy, precision, sensitivity and specificity was used to evaluate the performance of the models on the test dataset.

RESULTS

Fifty top-ranked subset features were selected by each feature selection technique for classification. The linear discriminant analysis (LDA) (AUROC, 0.863; accuracy, 0.794) and support vector machine (SVM) (AUROC, 0.863; accuracy, 0.792) classifiers, both of which coupled with the ℓ_2,1NR feature selection method, achieved optimal performance. The random forest (RF) classifier (AUROC, 0.824; accuracy, 0.775) and ℓ_2,1NR feature selection method (AUROC, 0.815; accuracy, 0.764) showed excellent average performance among the classifiers and feature selection methods employed in our study, respectively. Furthermore, the VGG16 DL algorithm (AUROC, 0.903; accuracy, 0.841) outperformed all conventional machine learning methods in combination with radiomics.

CONCLUSION

Employing radiomic machine learning/deep learning algorithms could help radiologists to differentiate the histologic subtypes of NSCLC via PET/CT images.

Improved cardiovascular risk prediction in patients with end-stage renal disease on hemodialysis using machine learning modeling and circulating microribonucleic acids

Rationale: To test whether novel biomarkers, such as microribonucleic acids (miRNAs), and nonstandard predictive models, such as decision tree learning, provide useful information for medical decision-making in patients on hemodialysis (HD).

Methods: Samples from patients with end-stage renal disease receiving HD included in the AURORA trial were investigated (n=810). The study included two independent phases: phase I (matched cases and controls, n=410) and phase II (unmatched cases and controls, n=400). The composite endpoint was cardiovascular death, nonfatal myocardial infarction or nonfatal stroke. miRNA quantification was performed using miRNA sequencing and RT-qPCR. The CART algorithm was used to construct regression tree models. A bagging-based procedure was used for validation.

Results: In phase I, miRNA sequencing in a subset of samples (n=20) revealed miR-632 as a candidate (fold change=2.9). miR-632 was associated with the endpoint, even after adjusting for confounding factors (HR from 1.43 to 1.53). These findings were not reproduced in phase II. Regression tree models identified eight patient subgroups with specific risk patterns. miR-186-5p and miR-632 entered the tree by redefining two risk groups: patients older than 64 years and with hsCRP<0.827 mg/L and diabetic patients younger than 64 years. miRNAs improved the discrimination accuracy at the beginning of the follow-up (24 months) compared to the models without miRNAs (integrated AUC [iAUC]=0.71).

Conclusions: The circulating miRNA profile complements conventional risk factors to identify specific cardiovascular risk patterns among patients receiving maintenance HD.

Machine learning in haematological malignancies

Machine learning is a branch of computer science and statistics that generates predictive or descriptive models by learning from training data rather than by being rigidly programmed. It has attracted substantial attention for its many applications in medicine, both as a catalyst for research and as a means of improving clinical care across the cycle of diagnosis, prognosis, and treatment of disease. These applications include the management of haematological malignancy, in which machine learning has created inroads in pathology, radiology, genomics, and the analysis of electronic health record data. As computational power becomes cheaper and the tools for implementing machine learning become increasingly democratised, it is likely to become increasingly integrated into the research and practice landscape of haematology. As such, machine learning merits understanding and attention from researchers and clinicians alike. This narrative Review describes important concepts in machine learning for unfamiliar readers, details machine learning's current applications in haematological malignancy, and summarises important concepts for clinicians to be aware of when appraising research that uses machine learning.

Association of MRI-derived radiomic biomarker with disease-free survival in patients with early-stage cervical cancer

Pre-treatment survival prediction plays a key role in many diseases. We aimed to determine the prognostic value of pre-treatment Magnetic Resonance Imaging (MRI) based radiomic score for disease-free survival (DFS) in patients with early-stage (IB-IIA) cervical cancer.

Methods: A total of 248 patients with early-stage cervical cancer underwent radical hysterectomy were included from two institutions between January 1, 2011 and December 31, 2017, whose MR imaging data, clinicopathological data and DFS data were collected. Patients data were randomly divided into the training cohort (n = 166) and the validation cohort (n=82). Radiomic features were extracted from the pre-treatment T2-weighted (T2w) and contrast-enhanced T1-weighted (CET1w) MR imagings for each patient. Least absolute shrinkage and selection operator (LASSO) regression and Cox proportional hazard model were applied to construct radiomic score (Rad-score). According to the cutoff of Rad-score, patients were divided into low- and high- risk groups. Pearson's correlation and Kaplan-Meier analysis were used to evaluate the association of Rad-score with DFS. A combined model incorporating Rad-score, lymph node metastasis (LNM) and lymphovascular space invasion (LVI) by multivariate Cox proportional hazard model was constructed to estimate DFS individually.

Results: Higher Rad-scores were significantly associated with worse DFS in the training and validation cohorts (P<0.001 and P=0.011, respectively). The Rad-score demonstrated better prognostic performance in estimating DFS (C-index, 0.753; 95% CI: 0.696-0.805) than the clinicopathological features (C-index, 0.632; 95% CI: 0.567-0.700). However, the combined model showed no significant improvement (C-index, 0.714; 95%CI: 0.642-0.784).

Conclusion: The results demonstrated that MRI-derived Rad-score can be used as a prognostic biomarker for patients with early-stage (IB-IIA) cervical cancer, which can facilitate clinical decision-making.

Machine learning classification models for fetal skeletal development performance prediction using maternal bone metabolic proteins in goats

Background

In developing countries, maternal undernutrition is the major intrauterine environmental factor contributing to fetal development and adverse pregnancy outcomes. Maternal nutrition restriction (MNR) in gestation has proven to impact overall growth, bone development, and proliferation and metabolism of mesenchymal stem cells in offspring. However, the efficient method for elucidation of fetal bone development performance through maternal bone metabolic biochemical markers remains elusive.

Methods

We adapted goats to elucidate fetal bone development state with maternal serum bone metabolic proteins under malnutrition conditions in mid- and late-gestation stages. We used the experimental data to create 72 datasets by mixing different input features such as one-hot encoding of experimental conditions, metabolic original data, experimental-centered features and experimental condition probabilities. Seven Machine Learning methods have been used to predict six fetal bone parameters (weight, length, and diameter of femur/humerus).

Results

The results indicated that MNR influences fetal bone development (femur and humerus) and fetal bone metabolic protein levels (C-terminal telopeptides of collagen I, CTx, in middle-gestation and N-terminal telopeptides of collagen I, NTx, in late-gestation), and maternal bone metabolites (low bone alkaline phosphatase, BALP, in middle-gestation and high BALP in late-gestation). The results show the importance of experimental conditions (ECs) encoding by mixing the information with the serum metabolic data. The best classification models obtained for femur weight (Fw) and length (FI), and humerus weight (Hw) are Support Vector Machines classifiers with the leave-one-out cross-validation accuracy of 1. The rest of the accuracies are 0.98, 0.946 and 0.696 for the diameter of femur (Fd), diameter and length of humerus (Hd, Hl), respectively. With the feature importance analysis, the moving averages mixed ECs are generally more important for the majority of the models. The moving average of parathyroid hormone (PTH) within nutritional conditions (MA-PTH-experim) is important for Fd, Hd and Hl prediction models but its removal for enhancing the Fw, Fl and Hw model performance. Further, using one feature models, it is possible to obtain even more accurate models compared with the feature importance analysis models. In conclusion, the machine learning is an efficient method to confirm the important role of PTH and BALP mixed with nutritional conditions for fetal bone growth performance of goats. All the Python scripts including results and comments are available into an open repository at https://gitlab.com/muntisa/goat-bones-machine-learning.