A Machine learning model trained on dual-energy CT radiomics significantly improves immunotherapy response prediction for patients with stage IV melanoma

Radiomic analysis of patient and interorgan heterogeneity in response to immunotherapies and BRAF-targeted therapy in metastatic melanoma

Variability in treatment response may be attributable to organ-level heterogeneity in tumor lesions. Radiomic analysis of medical images can elucidate non-invasive biomarkers of clinical outcome. Organ-specific radiomic comparison across immunotherapies and targeted therapies has not been previously reported. We queried the UPMC Hillman Cancer Center registry for patients with metastatic melanoma (MEL) treated with immune checkpoint inhibitors (ICI) (anti-programmed cell death protein-1 (PD-1)/cytotoxic T-lymphocyte associated protein 4 (CTLA-4) (ipilimumab+nivolumab; I+N) or anti-PD-1 monotherapy) or BRAF-targeted therapy. The best overall response was measured using Response Evaluation Criteria in Solid Tumors V.1.1. Lesions were segmented into discrete volume-of-interest with 400 radiomics features extracted. Overall and organ-specific machine-learning models were constructed to predict disease control (DC) versus progressive disease (PD) using XGBoost. 291 patients with MEL were identified, including 242 ICI (91 I+N, 151 PD-1) and 49 BRAF. 667 metastases were analyzed, including 541 ICI (236 I+N, 305 PD-1) and 126 BRAF. Across cohorts, baseline demographics included 39-47% women, 24%-29% M1C, 24-46% M1D, and 61-80% with elevated lactate dehydrogenase. Among ICI patients experiencing DC, the organs with the greatest reduction were liver (-66%±8%; mean±SEM) and lung (-63%±5%). For patients with multiple same-organ target lesions, the highest interlesion heterogeneity was observed in brain among patients who received ICI while no intraorgan heterogeneity was observed in BRAF. 221 ICI patients were included for radiomic modeling, consisting of 86 I+N and 135 PD-1. Models consisting of optimized radiomic signatures classified DC/PD across I+N (area under curve (AUC)=0.85) and PD-1 (0.71) and within individual organ sites (AUC=0.72~0.94). Integration of clinical variables improved the models' performance. Comparison of models between treatments and across organ sites suggested mostly non-overlapping DC or PD features. Skewness, kurtosis, and informational measure of correlation (IMC) were among the radiomic features shared between overall response models. Kurtosis and IMC were also used by multiple organ-site models. In conclusion, differential organ-specific response was observed across BRAF and ICI with within organ heterogeneity observed for ICI but not for BRAF. Radiomic features of organ-specific response demonstrated little overlap. Integrating clinical factors with radiomics improves the prediction of disease course outcome and prediction of tumor heterogeneity.

Convergence of evolving artificial intelligence and machine learning techniques in precision oncology

The confluence of new technologies with artificial intelligence (AI) and machine learning (ML) analytical techniques is rapidly advancing the field of precision oncology, promising to improve diagnostic approaches and therapeutic strategies for patients with cancer. By analyzing multi-dimensional, multiomic, spatial pathology, and radiomic data, these technologies enable a deeper understanding of the intricate molecular pathways, aiding in the identification of critical nodes within the tumor's biology to optimize treatment selection. The applications of AI/ML in precision oncology are extensive and include the generation of synthetic data, e.g., digital twins, in order to provide the necessary information to design or expedite the conduct of clinical trials. Currently, many operational and technical challenges exist related to data technology, engineering, and storage; algorithm development and structures; quality and quantity of the data and the analytical pipeline; data sharing and generalizability; and the incorporation of these technologies into the current clinical workflow and reimbursement models.

Quantitative multi-energy CT in oncology: State of the art and future directions

Multi-energy computed tomography (CT) involves acquisition of two or more CT measurements with distinct energy spectra. Using the differential attenuation of tissues and materials at different X-ray energies, multi-energy CT allows distinction of tissues and materials. Multi-energy technology encompasses different types of CT systems, such as dual-energy CT and photon-counting CT, that can use information from the energy and type of material present in acquired images to create multiple datasets. These scanners have overcome many of the limitations of conventional CT, making it possible to improve the diagnostic performance of CT and expand its use to new applications through better tissue characterization and multiple quantitative parameters. Quantitative imaging biomarkers based on multi-energy CT have enormous potential in oncologic imaging, from the diagnosis and characterization of tumor phenotypes to the evaluation of the response to treatment. Nevertheless, implementing these techniques in clinical practice remains challenging. This article reviews the basic principles underlying multi-energy CT and the most recent technical developments in these systems together with their advantages and limitations to establish the value of quantitative imaging derived from multi-energy CT in the field of oncology.

Deep learning on CT scans to predict checkpoint inhibitor treatment outcomes in advanced melanoma

Immune checkpoint inhibitor (ICI) treatment has proven successful for advanced melanoma, but is associated with potentially severe toxicity and high costs. Accurate biomarkers for response are lacking. The present work is the first to investigate the value of deep learning on CT imaging of metastatic lesions for predicting ICI treatment outcomes in advanced melanoma. Adult patients that were treated with ICI for advanced melanoma were retrospectively identified from ten participating centers. A deep learning model (DLM) was trained on volumes of lesions on baseline CT to predict clinical benefit. The DLM was compared to and combined with a model of known clinical predictors (presence of liver and brain metastasis, level of lactate dehydrogenase, performance status and number of affected organs). A total of 730 eligible patients with 2722 lesions were included. The DLM reached an area under the receiver operating characteristic (AUROC) of 0.607 [95%CI 0.565-0.648]. In comparison, a model of clinical predictors reached an AUROC of 0.635 [95%CI 0.59 -0.678]. The combination model reached an AUROC of 0.635 [95% CI 0.595-0.676]. Differences in AUROC were not statistically significant. The output of the DLM was significantly correlated with four of the five input variables of the clinical model. The DLM reached a statistically significant discriminative value, but was unable to improve over known clinical predictors. The present work shows that the assessment over known clinical predictors is an essential step for imaging-based prediction and brings important nuance to the almost exclusively positive findings in this field.

A multi-task deep learning model based on comprehensive feature integration and self-attention mechanism for predicting response to anti-PD1/PD-L1

BACKGROUND

Immune checkpoint inhibitor (ICI) has been widely used in the treatment of advanced cancers, but predicting their efficacy remains challenging. Traditional biomarkers are numerous but exhibit heterogeneity within populations. For comprehensively utilizing the ICI-related biomarkers, we aim to conduct multidimensional feature selection and deep learning model construction.

METHODS

We used statistical and machine learning methods to map features of different levels to next-generation sequencing gene expression. We integrated genes from different sources into the feature input of a deep learning model, by means of self-attention mechanism.

RESULTS

We performed feature selection at the single-cell sequencing level, PD-L1 (CD274) analysis level, tumor mutational burden (TMB)/mismatch repair (MMR) level, and somatic copy number alteration (SCNA) level, obtaining 96 feature genes. Based on the pan-cancer dataset, we trained a multi-task deep learning model. We tested the model in the bladder urothelial carcinoma testing set 1 (AUC = 0.62, n = 298), bladder urothelial carcinoma testing set 2 (AUC = 0.66, n = 89), non-small cell lung cancer testing set (AUC = 0.85, n = 27), and skin cutaneous melanoma testing set (AUC = 0.71, n = 27).

CONCLUSION

Our study demonstrates the potential of the deep learning model for integrating multidimensional features in predicting the outcome of ICI. Our study also provides a potential methodological case for medical scenarios requiring the integration of multiple levels of features.

Development and validation of a deep learning model for predicting gastric cancer recurrence based on CT imaging: a multicenter study

INTRODUCTION

The postoperative recurrence of gastric cancer (GC) has a significant impact on the overall prognosis of patients. Therefore, accurately predicting the postoperative recurrence of GC is crucial.

METHODS

This retrospective study gathered data from 2813 GC patients who underwent radical surgery between 2011 and 2017 at two medical centers. Follow-up was extended until May 2023, and cases were categorized as recurrent or nonrecurrent based on postoperative outcomes. Clinical pathological information and imaging data were collected for all patients. A new deep learning signature (DLS) was generated using pretreatment computed tomography images, based on a pretrained baseline (a customized Resnet50), for predicting postoperative recurrence. The deep learning fusion signature (DLFS) was created by combining the score of DLS with the weighted values of identified clinical features. The predictive performance of the model was evaluated based on discrimination, calibration, and clinical usefulness. Survival curves were plotted to investigate the differences between DLFS and prognosis.

RESULTS

In this study, 2813 patients with GC were recruited and allocated into training, internal validation, and external validation cohorts. The DLFS was developed and assessed for its capability in predicting the risk of postoperative recurrence. The DLFS exhibited excellent performance with AUCs of 0.833 (95% CI: 0.809-0.858) in the training set, 0.831 (95% CI: 0.792-0.871) in the internal validation set, and 0.859 (95% CI: 0.806-0.912) in the external validation set, along with satisfactory calibration across all cohorts ( P >0.05). Furthermore, the DLFS model significantly outperformed both the clinical model and DLS ( P <0.05). High-risk recurrent patients exhibit a significantly poorer prognosis compared to low-risk recurrent patients ( P <0.05).

CONCLUSIONS

The integrated model developed in this study, focusing on GC patients undergoing radical surgery, accurately identifies cases at high-risk of postoperative recurrence and highlights the potential of DLFS as a prognostic factor for GC patients.

Impact of multi-parameter images obtained from dual-energy CT on radiomics to predict pathological grading of bladder urothelial carcinoma

OBJECTIVE

To investigate the effect of radiomics models obtained from dual-energy CT (DECT) material decomposition images and virtual monoenergetic images (VMIs) in predicting the pathological grading of bladder urothelial carcinoma (BUC).

MATERIALS AND METHODS

A retrospective analysis of preoperative DECT examination was conducted on 112 patients diagnosed with BUC. This cohort included 76 cases of high-grade urothelial carcinoma and 36 cases of low-grade urothelial carcinoma. DECT can provide material decomposition images of venous phase Iodine maps and Water maps based on the differences in attenuation of substances, as well as VMIs at 40 to 140 keV (interval 10 keV). A total of 13 image sets were obtained, and radiomics features were extracted and analyzed from each set to achieve preoperative prediction of BUC. The best features related to BUC were identified by recursive feature elimination (RFE), the Minimum Redundancy Maximum Relevance (mRMR), and the Least Absolute Shrinkage and Selection Operator (LASSO) in order. A five-fold cross-validation method was used to divide the samples into training and testing sets, and models for pathological prediction of BUC grading were constructed by a random forest (RF) classifier. Receiver operating curves (ROC) were plotted to evaluate the performance of 13 models obtained from each image set.

RESULTS

Despite the notable differences in the best radiomics features chosen from each image set, all the features selected from 40 to 100 keV VMIs included the Dependence Variance of the GLDM feature set. There were no statistically significant differences in the area under the curve (AUC) between the training set and the testing set for all 13 models. In the testing set, the AUCs of the models established through 40 keV to 140 keV (interval of 10 keV) image sets were 0.895, 0.874, 0.855, 0.889, 0.841, 0.868, 0.852, 0.847, 0.889, 0.887 and 0.863 respectively. The AUCs for the models established using the Iodine maps and Water maps image sets were 0.873 and 0.852, respectively.

CONCLUSION

Despite the differences in the selected radiomic features from DECT multi-parameter images, the performance of radiomics models in predicting the pathological grading of BUC was not affected by the variations in the types of images used for model training.

A dual-energy computed tomography-based radiomics nomogram for predicting time since stroke onset: a multicenter study

OBJECTIVES

We aimed to develop and validate a radiomics nomogram based on dual-energy computed tomography (DECT) images and clinical features to classify the time since stroke (TSS), which could facilitate stroke decision-making.

MATERIALS AND METHODS

This retrospective three-center study consecutively included 488 stroke patients who underwent DECT between August 2016 and August 2022. The eligible patients were divided into training, test, and validation cohorts according to the center. The patients were classified into two groups based on an estimated TSS threshold of ≤ 4.5 h. Virtual images optimized the visibility of early ischemic lesions with more CT attenuation. A total of 535 radiomics features were extracted from polyenergetic, iodine concentration, virtual monoenergetic, and non-contrast images reconstructed using DECT. Demographic factors were assessed to build a clinical model. A radiomics nomogram was a tool that the Rad score and clinical factors to classify the TSS using multivariate logistic regression analysis. Predictive performance was evaluated using receiver operating characteristic (ROC) analysis, and decision curve analysis (DCA) was used to compare the clinical utility and benefits of different models.

RESULTS

Twelve features were used to build the radiomics model. The nomogram incorporating both clinical and radiomics features showed favorable predictive value for TSS. In the validation cohort, the nomogram showed a higher AUC than the radiomics-only and clinical-only models (AUC: 0.936 vs 0.905 vs 0.824). DCA demonstrated the clinical utility of the radiomics nomogram model.

CONCLUSIONS

The DECT-based radiomics nomogram provides a promising approach to predicting the TSS of patients.

CLINICAL RELEVANCE STATEMENT

The findings support the potential clinical use of DECT-based radiomics nomograms for predicting the TSS.

KEY POINTS

Accurately determining the TSS onset is crucial in deciding a treatment approach. The radiomics-clinical nomogram showed the best performance for predicting the TSS. Using the developed model to identify patients at different times since stroke can facilitate individualized management.

Tumor therapeutics in the era of "RECIST": past, current insights, and future prospects

In recent years, advancements in medical treatment and imaging technologies have revolutionized the assessment of tumor response. However, the Response Evaluation Criteria in Solid Tumors (RECIST) has long been established as the gold standard for evaluating tumor treatment. As treatment modalities evolve, the need for continuous refinement and adaptation of RECIST becomes increasingly apparent. This review explores the historical evolution, current applications, limitations, and future directions of RECIST. It discusses the challenges of distinguishing true progression from pseudo-progression in ICIs (immune checkpoint inhibitors), the integration of advanced imaging tools, and the necessity for RECIST criteria tailored to specific therapies like neoadjuvant treatments. The review highlights the ongoing efforts to enhance RECIST's accuracy and reliability in clinical decision-making and the potential for developing new standards to better evaluate treatment efficacy in the rapidly evolving landscape of oncology.

What to Expect (and What Not) from Dual-Energy CT Imaging Now and in the Future?

Dual-energy CT (DECT) imaging has broadened the potential of CT imaging by offering multiple postprocessing datasets with a single acquisition at more than one energy level. DECT shows profound capabilities to improve diagnosis based on its superior material differentiation and its quantitative value. However, the potential of dual-energy imaging remains relatively untapped, possibly due to its intricate workflow and the intrinsic technical limitations of DECT. Knowing the clinical advantages of dual-energy imaging and recognizing its limitations and pitfalls is necessary for an appropriate clinical use. The aims of this paper are to review the physical and technical bases of DECT acquisition and analysis, to discuss the advantages and limitations of DECT in different clinical scenarios, to review the technical constraints in material labeling and quantification, and to evaluate the cutting-edge applications of DECT imaging, including artificial intelligence, qualitative and quantitative imaging biomarkers, and DECT-derived radiomics and radiogenomics.

Radiomic analysis of patient and inter-organ heterogeneity in response to immunotherapies and BRAF targeted therapy in metastatic melanoma

Background

Variability in treatment response may be attributable to organ-level heterogeneity in tumor lesions. Radiomic analysis of medical images can elucidate non-invasive biomarkers of clinical outcome. Organ-specific radiomic comparison across immunotherapies and targeted therapies has not been previously reported.

Methods

We queried UPMC Hillman Cancer Center registry for patients with metastatic melanoma (MEL) treated with immune checkpoint inhibitors (ICI) (anti-PD1/CTLA4 [ipilimumab+nivolumab; I+N] or anti-PD1 monotherapy) or BRAF targeted therapy. Best overall response was measured using RECIST v1.1. Lesions were segmented into discrete volume-of-interest with 400 radiomics features extracted. Overall and organ-specific machine-learning models were constructed to predict disease control (DC) versus progressive disease (PD) using XGBoost.

Results

291 MEL patients were identified, including 242 ICI (91 I+N, 151 PD1) and 49 BRAF. 667 metastases were analyzed, including 541 ICI (236 I+N, 305 PD1) and 126 BRAF. Across cohorts, baseline demographics included 39-47% female, 24-29% M1C, 24-46% M1D, and 61-80% with elevated LDH. Among patients experiencing DC, the organs with the greatest reduction were liver (-88%±12%, I+N; mean±S.E.M.) and lung (-72%±8%, I+N). For patients with multiple same-organ target lesions, the highest inter-lesion heterogeneity was observed in brain among patients who received ICI while no intra-organ heterogeneity was observed in BRAF. 267 patients were kept for radiomic modeling, including 221 ICI (86 I+N, 135 PD1) and 46 BRAF. Models consisting of optimized radiomic signatures classified DC/PD across I+N (AUC=0.85) and PD1 (0.71) and within individual organ sites (AUC=0.72∼0.94). Integration of clinical variables improved the models' performance. Comparison of models between treatments and across organ sites suggested mostly non-overlapping DC or PD features. Skewness, kurtosis, and informational measure of correlation (IMC) were among the radiomic features shared between overall response models. Kurtosis and IMC were also utilized by multiple organ-site models.

Conclusions

Differential organ-specific response was observed across BRAF and ICI with within organ heterogeneity observed for ICI but not for BRAF. Radiomic features of organ-specific response demonstrated little overlap. Integrating clinical factors with radiomics improves the prediction of disease course outcome and prediction of tumor heterogeneity.

The role of artificial intelligence and convolutional neural networks in the management of melanoma: a clinical, pathological, and radiological perspective

Clinical dermatoscopy and pathological slide assessment are essential in the diagnosis and management of patients with cutaneous melanoma. For those presenting with stage IIC disease and beyond, radiological investigations are often considered. The dermatoscopic, whole slide and radiological images used during clinical care are often stored digitally, enabling artificial intelligence (AI) and convolutional neural networks (CNN) to learn, analyse and contribute to the clinical decision-making. A keyword search of the Medline database was performed to assess the progression, capabilities and limitations of AI and CNN and its use in diagnosis and management of cutaneous melanoma. Full-text articles were reviewed if they related to dermatoscopy, pathological slide assessment or radiology. Through analysis of 95 studies, we demonstrate that diagnostic accuracy of AI/CNN can be superior (or at least equal) to clinicians. However, variability in image acquisition, pre-processing, segmentation, and feature extraction remains challenging. With current technological abilities, AI/CNN and clinicians synergistically working together are better than one another in all subspecialty domains relating to cutaneous melanoma. AI has the potential to enhance the diagnostic capabilities of junior dermatology trainees, primary care skin cancer clinicians and general practitioners. For experienced clinicians, AI provides a cost-efficient second opinion. From a pathological and radiological perspective, CNN has the potential to improve workflow efficiency, allowing clinicians to achieve more in a finite amount of time. Until the challenges of AI/CNN are reliably met, however, they can only remain an adjunct to clinical decision-making.

Artificial intelligence for predictive biomarker discovery in immuno-oncology: a systematic review

BACKGROUND

The widespread use of immune checkpoint inhibitors (ICIs) has revolutionised treatment of multiple cancer types. However, selecting patients who may benefit from ICI remains challenging. Artificial intelligence (AI) approaches allow exploitation of high-dimension oncological data in research and development of precision immuno-oncology.

MATERIALS AND METHODS

We conducted a systematic literature review of peer-reviewed original articles studying the ICI efficacy prediction in cancer patients across five data modalities: genomics (including genomics, transcriptomics, and epigenomics), radiomics, digital pathology (pathomics), and real-world and multimodality data.

RESULTS

A total of 90 studies were included in this systematic review, with 80% published in 2021-2022. Among them, 37 studies included genomic, 20 radiomic, 8 pathomic, 20 real-world, and 5 multimodal data. Standard machine learning (ML) methods were used in 72% of studies, deep learning (DL) methods in 22%, and both in 6%. The most frequently studied cancer type was non-small-cell lung cancer (36%), followed by melanoma (16%), while 25% included pan-cancer studies. No prospective study design incorporated AI-based methodologies from the outset; rather, all implemented AI as a post hoc analysis. Novel biomarkers for ICI in radiomics and pathomics were identified using AI approaches, and molecular biomarkers have expanded past genomics into transcriptomics and epigenomics. Finally, complex algorithms and new types of AI-based markers, such as meta-biomarkers, are emerging by integrating multimodal/multi-omics data.

CONCLUSION

AI-based methods have expanded the horizon for biomarker discovery, demonstrating the power of integrating multimodal data from existing datasets to discover new meta-biomarkers. While most of the included studies showed promise for AI-based prediction of benefit from immunotherapy, none provided high-level evidence for immediate practice change. A priori planned prospective trial designs are needed to cover all lifecycle steps of these software biomarkers, from development and validation to integration into clinical practice.

Artificial Intelligence and Radiomics: Clinical Applications for Patients with Advanced Melanoma Treated with Immunotherapy

Immunotherapy has greatly improved the outcomes of patients with metastatic melanoma. However, it has also led to new patterns of response and progression, creating an unmet need for better biomarkers to identify patients likely to achieve a lasting clinical benefit or experience immune-related adverse events. In this study, we performed a focused literature survey covering the application of artificial intelligence (AI; in the form of radiomics, machine learning, and deep learning) to patients diagnosed with melanoma and treated with immunotherapy, reviewing 12 studies relevant to the topic published up to early 2022. The most commonly investigated imaging modality was CT imaging in isolation (n = 9, 75.0%), while patient cohorts were most frequently recruited retrospectively and from single institutions (n = 7, 58.3%). Most studies concerned the development of AI tools to assist in prognostication (n = 5, 41.7%) or the prediction of treatment response (n = 6, 50.0%). Validation methods were disparate, with two studies (16.7%) performing no validation and equal numbers using cross-validation (n = 3, 25%), a validation set (n = 3, 25%), or a test set (n = 3, 25%). Only one study used both validation and test sets (n = 1, 8.3%). Overall, promising results have been observed for the application of AI to immunotherapy-treated melanoma. Further improvement and eventual integration into clinical practice may be achieved through the implementation of rigorous validation using heterogeneous, prospective patient cohorts.

Virtual biopsy using CT radiomics for evaluation of disagreement in pathology between endoscopic biopsy and postoperative specimens in patients with gastric cancer: a dual-energy CT generalizability study

PURPOSE

To develop a noninvasive radiomics-based nomogram for identification of disagreement in pathology between endoscopic biopsy and postoperative specimens in gastric cancer (GC).

MATERIALS AND METHODS

This observational study recruited 181 GC patients who underwent pre-treatment computed tomography (CT) and divided them into a training set (n = 112, single-energy CT, SECT), a test set (n = 29, single-energy CT, SECT) and a validation cohort (n = 40, dual-energy CT, DECT). Radiomics signatures (RS) based on five machine learning algorithms were constructed from the venous-phase CT images. AUC and DeLong test were used to evaluate and compare the performance of the RS. We assessed the dual-energy generalization ability of the best RS. An individualized nomogram combined the best RS and clinical variables was developed, and its discrimination, calibration, and clinical usefulness were determined.

RESULTS

RS obtained with support vector machine (SVM) showed promising predictive capability with AUC of 0.91 and 0.83 in the training and test sets, respectively. The AUC of the best RS in the DECT validation cohort (AUC, 0.71) was significantly lower than that of the training set (Delong test, p = 0.035). The clinical-radiomic nomogram accurately predicted pathologic disagreement in the training and test sets, fitting well in the calibration curves. Decision curve analysis confirmed the clinical usefulness of the nomogram.

CONCLUSION

CT-based radiomics nomogram showed potential as a clinical aid for predicting pathologic disagreement status between biopsy samples and resected specimens in GC. When practicability and stability are considered, the SECT-based radiomics model is not recommended for DECT generalization.

CRITICAL RELEVANCE STATEMENT

Radiomics can identify disagreement in pathology between endoscopic biopsy and postoperative specimen.

Deep learning radio-clinical signatures for predicting neoadjuvant chemotherapy response and prognosis from pretreatment CT images of locally advanced gastric cancer patients

BACKGROUND

Early noninvasive screening of patients who would benefit from neoadjuvant chemotherapy (NCT) is essential for personalized treatment of locally advanced gastric cancer (LAGC). The aim of this study was to identify radio-clinical signatures from pretreatment oversampled computed tomography (CT) images to predict the response to NCT and prognosis of LAGC patients.

METHODS

LAGC patients were retrospectively recruited from six hospitals from January 2008 to December 2021. An SE-ResNet50-based chemotherapy response prediction system was developed from pretreatment CT images preprocessed with an imaging oversampling method (i.e. DeepSMOTE). Then, the deep learning (DL) signature and clinic-based features were fed into the deep learning radio-clinical signature (DLCS). The predictive performance of the model was evaluated based on discrimination, calibration, and clinical usefulness. An additional model was built to predict overall survival (OS) and explore the survival benefit of the proposed DL signature and clinicopathological characteristics.

RESULTS

A total of 1060 LAGC patients were recruited from six hospitals; the training cohort (TC) and internal validation cohort (IVC) patients were randomly selected from center I. An external validation cohort (EVC) of 265 patients from five other centers was also included. The DLCS exhibited excellent performance in predicting the response to NCT in the IVC [area under the curve (AUC), 0.86] and EVC (AUC, 0.82), with good calibration in all cohorts ( P >0.05). Moreover, the DLCS model outperformed the clinical model ( P <0.05). Additionally, we found that the DL signature could serve as an independent factor for prognosis [hazard ratio (HR), 0.828, P =0.004]. The concordance index (C-index), integrated area under the time-dependent ROC curve (iAUC), and integrated Brier score (IBS) for the OS model were 0.64, 1.24, and 0.71 in the test set.

CONCLUSION

The authors proposed a DLCS model that combined imaging features with clinical risk factors to accurately predict tumor response and identify the risk of OS in LAGC patients prior to NCT, which can then be used to guide personalized treatment plans with the help of computerized tumor-level characterization.

Computed tomography angiography-based radiomics model for predicting carotid atherosclerotic plaque vulnerability

Vulnerable carotid atherosclerotic plaque (CAP) significantly contributes to ischemic stroke. Neovascularization within plaques is an emerging biomarker linked to plaque vulnerability that can be detected using contrast-enhanced ultrasound (CEUS). Computed tomography angiography (CTA) is a common method used in clinical cerebrovascular assessments that can be employed to evaluate the vulnerability of CAPs. Radiomics is a technique that automatically extracts radiomic features from images. This study aimed to identify radiomic features associated with the neovascularization of CAP and construct a prediction model for CAP vulnerability based on radiomic features. CTA data and clinical data of patients with CAPs who underwent CTA and CEUS between January 2018 and December 2021 in Beijing Hospital were retrospectively collected. The data were divided into a training cohort and a testing cohort using a 7:3 split. According to the examination of CEUS, CAPs were dichotomized into vulnerable and stable groups. 3D Slicer software was used to delineate the region of interest in CTA images, and the Pyradiomics package was used to extract radiomic features in Python. Machine learning algorithms containing logistic regression (LR), support vector machine (SVM), random forest (RF), light gradient boosting machine (LGBM), adaptive boosting (AdaBoost), extreme gradient boosting (XGBoost), and multi-layer perception (MLP) were used to construct the models. The confusion matrix, receiver operating characteristic (ROC) curve, accuracy, precision, recall, and f-1 score were used to evaluate the performance of the models. A total of 74 patients with 110 CAPs were included. In all, 1,316 radiomic features were extracted, and 10 radiomic features were selected for machine-learning model construction. After evaluating several models on the testing cohorts, it was discovered that model_RF outperformed the others, achieving an AUC value of 0.93 (95% CI: 0.88-0.99). The accuracy, precision, recall, and f-1 score of model_RF in the testing cohort were 0.85, 0.87, 0.85, and 0.85, respectively. Radiomic features associated with the neovascularization of CAP were obtained. Our study highlights the potential of radiomics-based models for improving the accuracy and efficiency of diagnosing vulnerable CAP. In particular, the model_RF, utilizing radiomic features extracted from CTA, provides a noninvasive and efficient method for accurately predicting the vulnerability status of CAP. This model shows great potential for offering clinical guidance for early detection and improving patient outcomes.

Artificial intelligence-assisted selection and efficacy prediction of antineoplastic strategies for precision cancer therapy

The rapid development of artificial intelligence (AI) technologies in the context of the vast amount of collectable data obtained from high-throughput sequencing has led to an unprecedented understanding of cancer and accelerated the advent of a new era of clinical oncology with a tone of precision treatment and personalized medicine. However, the gains achieved by a variety of AI models in clinical oncology practice are far from what one would expect, and in particular, there are still many uncertainties in the selection of clinical treatment options that pose significant challenges to the application of AI in clinical oncology. In this review, we summarize emerging approaches, relevant datasets and open-source software of AI and show how to integrate them to address problems from clinical oncology and cancer research. We focus on the principles and procedures for identifying different antitumor strategies with the assistance of AI, including targeted cancer therapy, conventional cancer therapy, and cancer immunotherapy. In addition, we also highlight the current challenges and directions of AI in clinical oncology translation. Overall, we hope this article will provide researchers and clinicians with a deeper understanding of the role and implications of AI in precision cancer therapy, and help AI move more quickly into accepted cancer guidelines.

CT radiomics compared to a clinical model for predicting checkpoint inhibitor treatment outcomes in patients with advanced melanoma

INTRODUCTION

Predicting checkpoint inhibitors treatment outcomes in melanoma is a relevant task, due to the unpredictable and potentially fatal toxicity and high costs for society. However, accurate biomarkers for treatment outcomes are lacking. Radiomics are a technique to quantitatively capture tumour characteristics on readily available computed tomography (CT) imaging. The purpose of this study was to investigate the added value of radiomics for predicting clinical benefit from checkpoint inhibitors in melanoma in a large, multicenter cohort.

METHODS

Patients who received first-line anti-PD1±anti-CTLA4 treatment for advanced cutaneous melanoma were retrospectively identified from nine participating hospitals. For every patient, up to five representative lesions were segmented on baseline CT, and radiomics features were extracted. A machine learning pipeline was trained on the radiomics features to predict clinical benefit, defined as stable disease for more than 6 months or response per RECIST 1.1 criteria. This approach was evaluated using a leave-one-centre-out cross validation and compared to a model based on previously discovered clinical predictors. Lastly, a combination model was built on the radiomics and clinical model.

RESULTS

A total of 620 patients were included, of which 59.2% experienced clinical benefit. The radiomics model achieved an area under the receiver operator characteristic curve (AUROC) of 0.607 [95% CI, 0.562-0.652], lower than that of the clinical model (AUROC=0.646 [95% CI, 0.600-0.692]). The combination model yielded no improvement over the clinical model in terms of discrimination (AUROC=0.636 [95% CI, 0.592-0.680]) or calibration. The output of the radiomics model was significantly correlated with three out of five input variables of the clinical model (p < 0.001).

DISCUSSION

The radiomics model achieved a moderate predictive value of clinical benefit, which was statistically significant. However, a radiomics approach was unable to add value to a simpler clinical model, most likely due to the overlap in predictive information learned by both models. Future research should focus on the application of deep learning, spectral CT-derived radiomics, and a multimodal approach for accurately predicting benefit to checkpoint inhibitor treatment in advanced melanoma.

A deep belief network-based clinical decision system for patients with osteosarcoma

Osteosarcoma was the most frequent type of malignant primary bone tumor with a poor survival rate mainly occurring in children and adolescents. For precision treatment, an accurate individualized prognosis for Osteosarcoma patients is highly desired. In recent years, many machine learning-based approaches have been used to predict distant metastasis and overall survival based on available individual information. In this study, we compared the performance of the deep belief networks (DBN) algorithm with six other machine learning algorithms, including Random Forest, XGBoost, Decision Tree, Gradient Boosting Machine, Logistic Regression, and Naive Bayes Classifier, to predict lung metastasis for Osteosarcoma patients. Therefore the DBN-based lung metastasis prediction model was integrated as a parameter into the Cox proportional hazards model to predict the overall survival of Osteosarcoma patients. The accuracy, precision, recall, and F1 score of the DBN algorithm were 0.917/0.888, 0.896/0.643, 0.956/0.900, and 0.925/0.750 in the training/validation sets, respectively, which were better than the other six machine-learning algorithms. For the performance of the DBN survival Cox model, the areas under the curve (AUCs) for the 1-, 3- and 5-year survival in the training set were 0.851, 0.806 and 0.793, respectively, indicating good discrimination, and the calibration curves showed good agreement between the prediction and actual observations. The DBN survival Cox model also demonstrated promising performance in the validation set. In addition, a nomogram integrating the DBN output was designed as a tool to aid clinical decision-making.

Artificial intelligence in oncologic imaging

Radiology is integral to cancer care. Compared to molecular assays, imaging has its advantages. Imaging as a noninvasive tool can assess the entirety of tumor unbiased by sampling error and is routinely acquired at multiple time points in oncological practice. Imaging data can be digitally post-processed for quantitative assessment. The ever-increasing application of Artificial intelligence (AI) to clinical imaging is challenging radiology to become a discipline with competence in data science, which plays an important role in modern oncology. Beyond streamlining certain clinical tasks, the power of AI lies in its ability to reveal previously undetected or even imperceptible radiographic patterns that may be difficult to ascertain by the human sensory system. Here, we provide a narrative review of the emerging AI applications relevant to the oncological imaging spectrum and elaborate on emerging paradigms and opportunities. We envision that these technical advances will change radiology in the coming years, leading to the optimization of imaging acquisition and discovery of clinically relevant biomarkers for cancer diagnosis, staging, and treatment monitoring. Together, they pave the road for future clinical translation in precision oncology.

Imaging to predict checkpoint inhibitor outcomes in cancer. A systematic review

BACKGROUND

Checkpoint inhibition has radically improved the perspective for patients with metastatic cancer, but predicting who will not respond with high certainty remains difficult. Imaging-derived biomarkers may be able to provide additional insights into the heterogeneity in tumour response between patients. In this systematic review, we aimed to summarise and qualitatively assess the current evidence on imaging biomarkers that predict response and survival in patients treated with checkpoint inhibitors in all cancer types.

METHODS

PubMed and Embase were searched from database inception to 29th November 2021. Articles eligible for inclusion described baseline imaging predictive factors, radiomics and/or imaging machine learning models for predicting response and survival in patients with any kind of malignancy treated with checkpoint inhibitors. Risk of bias was assessed using the QUIPS and PROBAST tools and data was extracted.

RESULTS

In total, 119 studies including 15,580 patients were selected. Of these studies, 73 investigated simple imaging factors. 45 studies investigated radiomic features or deep learning models. Predictors of worse survival were (i) higher tumour burden, (ii) presence of liver metastases, (iii) less subcutaneous adipose tissue, (iv) less dense muscle and (v) presence of symptomatic brain metastases. Hazard rate ratios did not exceed 2.00 for any predictor in the larger and higher quality studies. The added value of baseline fluorodeoxyglucose positron emission tomography parameters in predicting response to treatment was limited. Pilot studies of radioactive drug tracer imaging showed promising results. Reports on radiomics were almost unanimously positive, but numerous methodological concerns exist.

CONCLUSIONS

There is well-supported evidence for several imaging biomarkers that can be used in clinical decision making. Further research, however, is needed into biomarkers that can more accurately identify which patients who will not benefit from checkpoint inhibition. Radiomics and radioactive drug labelling appear to be promising approaches for this purpose.

Artificial intelligence and radiomics: fundamentals, applications, and challenges in immunotherapy

Immunotherapy offers the potential for durable clinical benefit but calls into question the association between tumor size and outcome that currently forms the basis for imaging-guided treatment. Artificial intelligence (AI) and radiomics allow for discovery of novel patterns in medical images that can increase radiology’s role in management of patients with cancer, although methodological issues in the literature limit its clinical application. Using keywords related to immunotherapy and radiomics, we performed a literature review of MEDLINE, CENTRAL, and Embase from database inception through February 2022. We removed all duplicates, non-English language reports, abstracts, reviews, editorials, perspectives, case reports, book chapters, and non-relevant studies. From the remaining articles, the following information was extracted: publication information, sample size, primary tumor site, imaging modality, primary and secondary study objectives, data collection strategy (retrospective vs prospective, single center vs multicenter), radiomic signature validation strategy, signature performance, and metrics for calculation of a Radiomics Quality Score (RQS). We identified 351 studies, of which 87 were unique reports relevant to our research question. The median (IQR) of cohort sizes was 101 (57–180). Primary stated goals for radiomics model development were prognostication (n=29, 33.3%), treatment response prediction (n=24, 27.6%), and characterization of tumor phenotype (n=14, 16.1%) or immune environment (n=13, 14.9%). Most studies were retrospective (n=75, 86.2%) and recruited patients from a single center (n=57, 65.5%). For studies with available information on model testing, most (n=54, 65.9%) used a validation set or better. Performance metrics were generally highest for radiomics signatures predicting treatment response or tumor phenotype, as opposed to immune environment and overall prognosis. Out of a possible maximum of 36 points, the median (IQR) of RQS was 12 (10–16). While a rapidly increasing number of promising results offer proof of concept that AI and radiomics could drive precision medicine approaches for a wide range of indications, standardizing the data collection as well as optimizing the methodological quality and rigor are necessary before these results can be translated into clinical practice.

Imaging approaches and radiomics: toward a new era of ultraprecision radioimmunotherapy?

Strong rationale and a growing number of preclinical and clinical studies support combining radiotherapy and immunotherapy to improve patient outcomes. However, several critical questions remain, such as the identification of patients who will benefit from immunotherapy and the identification of the best modalities of treatment to optimize patient response. Imaging biomarkers and radiomics have recently emerged as promising tools for the non-invasive assessment of the whole disease of the patient, allowing comprehensive analysis of the tumor microenvironment, the spatial heterogeneity of the disease and its temporal changes. This review presents the potential applications of medical imaging and the challenges to address, in order to help clinicians choose the optimal modalities of both radiotherapy and immunotherapy, to predict patient's outcomes and to assess response to these promising combinations.

Parameters of dual-energy CT for the differential diagnosis of thyroid nodules and the indirect prediction of lymph node metastasis in thyroid carcinoma: a retrospective diagnostic study

BACKGROUND

To further investigate the differential diagnosis of thyroid nodules using dual-energy computed tomography (DECT) and explore the relationship between DECT parameters and lymph node metastasis in thyroid carcinoma for clinical practice, especially difficult diagnosis by routine imaging examination.

METHODS

A total of 150 patients with thyroid nodules who underwent preoperative DECT and Thyroid Imaging Report and Data System (TIRADS) classification were enrolled in this study, including 96 patients with malignant tumors and 54 with benign tumors. The DECT parameters were got form regions of interest (ROI) by an experienced radiologist team and thyroid nodules and lymph node status of all patients were identified by cytology and histopathology. Statistical analyses were performed using Student's t-test, Chi-squared test, and receiver operating characteristic (ROC) curves.

RESULTS

In the differential diagnosis of benign and malignant thyroid nodules, the optimal iodine concentration (IC) and normalized iodine concentration (NIC) cut-off values were ICa (2.835 mg/mL), NIC1a (0.690), and their corresponding area under the curve (AUC) were 0.940, 0.954 respectively; meantime, the optimal computed tomography (CT) value and slope of the spectral Hounsfield unit curve (λHU) cut-off values were 70 keVa (125.05 HU) and λHU2a (1.405), and their corresponding AUC were 0.955, 0.941 respectively. For lymph node status (with or without lymph node metastasis), the optimal IC and NIC thresholds were ICa (1.715 mg/mL) and NIC2a (0.155), and their corresponding AUC were 0.717, 0.720 respectively; meanwhile, the optimal CT value and λHU thresholds were 70 keVv (89.635 HU) and λHU2v (1.185), and their corresponding AUC were 0.729, 0.641 respectively.

CONCLUSIONS

Base on our study, we think DECT is useful in differentiating malignant from benign thyroid nodules, which has potential value in the indirect prediction of lymph node metastasis in thyroid carcinoma.