Radiomics in PET: principles and applications

The prognostic value of [18F]FDG PET/CT texture analysis prior to transplantation for unresectable colorectal liver metastases

INTRODUCTION

To determine whether heterogeneity in colorectal liver metastases (CRLM) 18F fluorodeoxyglucose [18F]FDG distribution is predictive of disease-free survival (DFS) and overall survival (OS) following liver transplantation (LT) for unresectable CRLM.

METHODS

The preoperative [18F]FDG positron emission tomography/computed tomography examinations of all patients in the secondary cancer 1 and 2 studies were retrospectively assessed. Maximum standardized uptake value (SUVmax), metabolic tumour volume (MTV), and six texture heterogeneity parameters (joint entropyGLCM, dissimilarityGLCM, grey level varianceSZM, size zone varianceSZM, and zone percentageSZM, and morphological feature convex deficiency) were obtained. DFS and OS for patients over and under the median value for each of these parameters were compared by using the Kaplan Meier method and log rank test.

RESULTS

Twenty-eight out of 40 patients who underwent LT for unresectable CRLM had liver metastases with uptake above liver background and were eligible for inclusion. Low MTV (p < 0.001) and dissimilarityGLCM (p = 0.016) were correlated to longer DFS. Low MTV (p < 0.001) and low values of the texture parameters dissimilarityGLCM (p = 0.038), joint entropyGLCM (p = 0.015) and zone percentageSZM (p = 0.037) were significantly correlated to longer OS. SUVmax was not correlated to DFS and OS.

CONCLUSION

Although some texture parameters were able to significantly predict DFS and OS, MTV seems to be superior to predict both DFS and OS following LT for unresectable CRLM.

A Machine Learning-Based Clustering Using Radiomics of F-18 Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography for the Prediction of Prognosis in Patients with Intrahepatic Cholangiocarcinoma

BACKGROUND

Intrahepatic cholangiocarcinoma (IHCC) is highly aggressive primary hepatic malignancy with an increasing incidence.

OBJECTIVE

This study aimed to develop machine learning-based radiomic clustering using F-18 fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) for predicting recurrence-free survival (RFS) and overall survival (OS) in IHCC.

METHODS

We retrospectively reviewed pretreatment F-18 FDG PET/CT scans of 60 IHCC patients who underwent surgery without neoadjuvant treatment between January 2008 and July 2020. Radiomic features such as first order, shape, and gray level were extracted from the scans of 52 patients and analyzed using unsupervised hierarchical clustering.

RESULTS

Of the 60 patients, 36 experienced recurrence and 31 died during follow-up. Eight patients with a negative FDG uptake were classified as Group 0. The unsupervised hierarchical clustering analysis divided the total cohort into three clusters (Group 1: n = 27; Group 2: n = 23; Group 3: n = 2). The Kaplan-Meier curves showed significant differences in RFS and OS among the clusters (p < 0.0001). Multivariate analyses showed that the PET radiomics grouping was an independent prognostic factor for RFS (hazard ratio (HR) = 3.03, p = 0.001) and OS (HR = 2.39, p = 0.030). Oxidative phosphorylation was significantly activated in Group 1, and the KRAS, P53, and WNT β-catenin pathways were enriched in Group 2.

CONCLUSIONS

This study demonstrated that machine learning-based PET radiomics clustering can preoperatively predict prognosis and provide valuable information complementing the genomic profiling of IHCC.

Future of Team-based Basic and Translational Science in Radiation Oncology

To further optimize radiotherapy, a more personalized treatment towards individual patient's risk profiles, dissecting both patient-specific tumor and normal tissue response to multimodality treatments is needed. Novel developments in radiobiology, using in vitro patient-specific complex tissue resembling 3D models and multiomics approaches at a spatial single-cell level, may provide unprecedented insight into the radiation responses of tumors and normal tissue. Here, we describe the necessary team effort, including all disciplines in radiation oncology, to integrate such data into clinical prediction models and link the relatively "big data" from the clinical practice, allowing accurate patient stratification for personalized treatment approaches.

Development and Validation of Prognostic Models Using Radiomic Features from Pre-Treatment Positron Emission Tomography (PET) Images in Head and Neck Squamous Cell Carcinoma (HNSCC) Patients

High-dimensional radiomics features derived from pre-treatment positron emission tomography (PET) images offer prognostic insights for patients with head and neck squamous cell carcinoma (HNSCC). Using 124 PET radiomics features and clinical variables (age, sex, stage of cancer, site of cancer) from a cohort of 232 patients, we evaluated four survival models-penalized Cox model, random forest, gradient boosted model and support vector machine-to predict all-cause mortality (ACM), locoregional recurrence/residual disease (LR) and distant metastasis (DM) probability during 36, 24 and 24 months of follow-up, respectively. We developed models with five-fold cross-validation, selected the best-performing model for each outcome based on the concordance index (C-statistic) and the integrated Brier score (IBS) and validated them in an independent cohort of 102 patients. The penalized Cox model demonstrated better performance for ACM (C-statistic = 0.70, IBS = 0.12) and DM (C-statistic = 0.70, IBS = 0.08) while the random forest model displayed better performance for LR (C-statistic = 0.76, IBS = 0.07). We conclude that the ML-based prognostic model can aid clinicians in quantifying prognosis and determining effective treatment strategies, thereby improving favorable outcomes in HNSCC patients.

2-[18F]FDG PET-based quantification of lymph node metabolic heterogeneity for predicting lymph node metastasis in patients with colorectal cancer

BACKGROUND AND PURPOSE

The pre-surgical estimation of lymph node (LN) metastasis in colorectal cancer (CRC) poses a significant diagnostic predicament. The associations between LN morphology, density, and metabolic heterogeneity and LN metastasis status in CRCs have been seldomly examined through the lens of radiomics. This research aimed to assess 2-[18F]FDG PET-based quantification of intratumoral metabolic heterogeneity for predicting lymph node metastasis in patients with colorectal cancer.

MATERIALS AND METHODS

The construction of the model utilized data from 264 CRC patients, all of whom underwent preoperative 2-[18F]FDG PET/CT. Radiomic features were extracted from PET and CT images of LNs. Least absolute shrinkage and selection operator (LASSO) regression was implemented for selecting pertinent imaging features with a tenfold cross-validation. The predictive accuracy for LN metastasis status was juxtaposed against traditional methodologies (comprising CT-reported LN status and PET/CT-reported LN status) by deploying the receiver operating characteristic (ROC) curve analysis. The radiomics signature was evaluated based on discrimination, calibration, and clinical utility parameters. The model was further subjected to validation using an independent cohort of 132 patients from the period of January 2012 to June 2020.

RESULTS

The radiomics model was composed of eight significant radiomic features (five from PET and three from CT), encapsulating metabolic and density heterogeneity. The radiomics signature (area under the curve (AUC), 0.908) showcased a significantly superior performance compared to CT-reported LN status (AUC, 0.563, P < 0.001) and PET/CT-reported LN status (AUC, 0.64, P < 0.001) for predicting LN-positive or LN-negative status. The radiomics signature (AUC, 0.885) also showcased a significantly superior performance compared to CT-reported LN status (AUC, 0.587, P < 0.001) and PET/CT-reported LN status (AUC, 0.621, P < 0.001) to identify N1 and N2. This signature maintained its independence from clinical risk factors and exhibited robustness in the validation test set. Decision curve analysis attested to the clinical utility of the radiomics signature.

CONCLUSIONS

The radiomics signature based on 2-[18F]FDG PET/CT, which derived image features directly from LNs irrespective of clinical risk factors, displayed enhanced diagnostic performance compared to conventional CT or PET/CT-reported LN status. This allows for the identification of pre-surgical LN metastasis status and facilitates a patient-specific prediction of LN metastasis status in CRC patients.

Refining mutanome-based individualised immunotherapy of melanoma using artificial intelligence

Using the particular nature of melanoma mutanomes to develop medicines that activate the immune system against specific mutations is a game changer in immunotherapy individualisation. It offers a viable solution to the recent rise in resistance to accessible immunotherapy alternatives, with some patients demonstrating innate resistance to these drugs despite past sensitisation to these agents. However, various obstacles stand in the way of this method, most notably the practicality of sequencing each patient's mutanome, selecting immunotherapy targets, and manufacturing specific medications on a large scale. With the robustness and advancement in research techniques, artificial intelligence (AI) is a potential tool that can help refine the mutanome-based immunotherapy for melanoma. Mutanome-based techniques are being employed in the development of immune-stimulating vaccines, improving current options such as adoptive cell treatment, and simplifying immunotherapy responses. Although the use of AI in these approaches is limited by data paucity, cost implications, flaws in AI inference capabilities, and the incapacity of AI to apply data to a broad population, its potential for improving immunotherapy is limitless. Thus, in-depth research on how AI might help the individualisation of immunotherapy utilising knowledge of mutanomes is critical, and this should be at the forefront of melanoma management.

18F-FDG PET/CT-based deep learning radiomics predicts 5-years disease-free survival after failure to achieve pathologic complete response to neoadjuvant chemotherapy in breast cancer

BACKGROUND

This study aimed to assess whether a combined model incorporating radiomic and depth features extracted from PET/CT can predict disease-free survival (DFS) in patients who failed to achieve pathologic complete response (pCR) after neoadjuvant chemotherapy.

RESULTS

This study retrospectively included one hundred and five non-pCR patients. After a median follow-up of 71 months, 15 and 7 patients experienced recurrence and death, respectively. The primary tumor volume underwent feature extraction, yielding a total of 3644 radiomic features and 4096 depth features. The modeling procedure employed Cox regression for feature selection and utilized Cox proportional-hazards models to make predictions on DFS. Time-dependent receiver operating characteristic (ROC) curves and the area under the ROC curve (AUC) were utilized to evaluate and compare the predictive performance of different models. 2 clinical features (RCB, cT), 4 radiomic features, and 7 depth features were significant predictors of DFS and were included to develop models. The integrated model incorporating RCB, cT, and radiomic and depth features extracted from PET/CT images exhibited the highest accuracy for predicting 5-year DFS in the training (AUC 0.943) and the validation cohort (AUC 0.938).

CONCLUSION

The integrated model combining radiomic and depth features extracted from PET/CT images can accurately predict 5-year DFS in non-pCR patients. It can help identify patients with a high risk of recurrence and strengthen adjuvant therapy to improve survival.

Role of radiomic analysis of [18F]fluoromethylcholine PET/CT in predicting biochemical recurrence in a cohort of intermediate and high risk prostate cancer patients at initial staging

AIM

To study the feasibility of radiomic analysis of baseline [18F]fluoromethylcholine positron emission tomography/computed tomography (PET/CT) for the prediction of biochemical recurrence (BCR) in a cohort of intermediate and high-risk prostate cancer (PCa) patients.

MATERIAL AND METHODS

Seventy-four patients were prospectively collected. We analyzed three prostate gland (PG) segmentations (i.e., PGwhole: whole PG; PG41%: prostate having standardized uptake value - SUV > 0.41*SUVmax; PG2.5: prostate having SUV > 2.5) together with three SUV discretization steps (i.e., 0.2, 0.4, and 0.6). For each segmentation/discretization step, we trained a logistic regression model to predict BCR using radiomic and/or clinical features.

RESULTS

The median baseline prostate-specific antigen was 11 ng/mL, the Gleason score was > 7 for 54% of patients, and the clinical stage was T1/T2 for 89% and T3 for 9% of patients. The baseline clinical model achieved an area under the receiver operating characteristic curve (AUC) of 0.73. Performances improved when clinical data were combined with radiomic features, in particular for PG2.5 and 0.4 discretization, for which the median test AUC was 0.78.

CONCLUSION

Radiomics reinforces clinical parameters in predicting BCR in intermediate and high-risk PCa patients. These first data strongly encourage further investigations on the use of radiomic analysis to identify patients at risk of BCR.

CLINICAL RELEVANCE STATEMENT

The application of AI combined with radiomic analysis of [18F]fluoromethylcholine PET/CT images has proven to be a promising tool to stratify patients with intermediate or high-risk PCa in order to predict biochemical recurrence and tailor the best treatment options.

KEY POINTS

• Stratification of patients with intermediate and high-risk prostate cancer at risk of biochemical recurrence before initial treatment would help determine the optimal curative strategy. • Artificial intelligence combined with radiomic analysis of [18F]fluorocholine PET/CT images allows prediction of biochemical recurrence, especially when radiomic features are complemented with patients' clinical information (highest median AUC of 0.78). • Radiomics reinforces the information of conventional clinical parameters (i.e., Gleason score and initial prostate-specific antigen level) in predicting biochemical recurrence.

Radiomics predictive modeling from dual-time-point FDG PET Ki parametric maps: application to chemotherapy response in lymphoma

BACKGROUND

To investigate the use of dynamic radiomics features derived from dual-time-point (DTP-feature) [¹⁸F]FDG PET metabolic uptake rate K_i parametric maps to develop a predictive model for response to chemotherapy in lymphoma patients.

METHODS

We analyzed 126 lesions from 45 lymphoma patients (responding n = 75 and non-responding n = 51) treated with chemotherapy from two different centers. Static and DTP radiomics features were extracted from baseline static PET images and DTP K_i parametric maps. Spearman's rank correlations were calculated between static and DTP features to identify features with potential additional information. We first employed univariate analysis to determine correlations between individual features, and subsequently utilized multivariate analysis to derive predictive models utilizing DTP and static radiomics features before and after ComBat harmonization. For multivariate modeling, we utilized both the minimum redundancy maximum relevance feature selection technique and the XGBoost classifier. To evaluate our model, we partitioned the patient datasets into training/validation and testing sets using an 80/20% split. Different metrics for classification including area under the curve (AUC), sensitivity (SEN), specificity (SPE), and accuracy (ACC) were reported in test sets.

RESULTS

Via Spearman's rank correlations, there was negligible to moderate correlation between 32 out of 65 DTP features and some static features (ρ < 0.7); all the other 33 features showed high correlations (ρ ≥ 0.7). In univariate modeling, no significant difference between AUC of DTP and static features was observed. GLRLM_RLNU from static features demonstrated a strong correlation (AUC = 0.75, p value = 0.0001, q value = 0.0007) with therapy response. The most predictive DTP features were GLCM_Energy, GLCM_Entropy, and Uniformity, each with AUC = 0.73, p value = 0.0001, and q value < 0.0005. In multivariate analysis, the mean ranges of AUCs increased following harmonization. Use of harmonization plus combining DTP and static features was shown to provide significantly improved predictions (AUC = 0.97 ± 0.02, accuracy = 0.89 ± 0.05, sensitivity = 0.92 ± 0.09, and specificity = 0.88 ± 0.05). All models depicted significant performance in terms of AUC, ACC, SEN, and SPE (p < 0.05, Mann-Whitney test).

CONCLUSIONS

Our results demonstrate significant value in harmonization of radiomics features as well as combining DTP and static radiomics models for predicting response to chemotherapy in lymphoma patients.

Phenomic Studies on Diseases: Potential and Challenges

The rapid development of such research field as multi-omics and artificial intelligence (AI) has made it possible to acquire and analyze the multi-dimensional big data of human phenomes. Increasing evidence has indicated that phenomics can provide a revolutionary strategy and approach for discovering new risk factors, diagnostic biomarkers and precision therapies of diseases, which holds profound advantages over conventional approaches for realizing precision medicine: first, the big data of patients' phenomes can provide remarkably richer information than that of the genomes; second, phenomic studies on diseases may expose the correlations among cross-scale and multi-dimensional phenomic parameters as well as the mechanisms underlying the correlations; and third, phenomics-based studies are big data-driven studies, which can significantly enhance the possibility and efficiency for generating novel discoveries. However, phenomic studies on human diseases are still in early developmental stage, which are facing multiple major challenges and tasks: first, there is significant deficiency in analytical and modeling approaches for analyzing the multi-dimensional data of human phenomes; second, it is crucial to establish universal standards for acquirement and management of phenomic data of patients; third, new methods and devices for acquirement of phenomic data of patients under clinical settings should be developed; fourth, it is of significance to establish the regulatory and ethical guidelines for phenomic studies on diseases; and fifth, it is important to develop effective international cooperation. It is expected that phenomic studies on diseases would profoundly and comprehensively enhance our capacity in prevention, diagnosis and treatment of diseases.

Radiomics-based sialadenitis staging in contrast-enhanced computed tomography and ultrasonography: a preliminary rat model study

OBJECTIVE

The aim of this study was to measure the ability of radiomics analysis to diagnose different stages of sialadenitis, compare the diagnostic accuracy of computed tomography (CT) and ultrasonography (US), and suggest radiomics features selected through 3 machine learning algorithms that would be helpful in discriminating between stages of sialadenitis with both imaging systems.

STUDY DESIGN

Wistar rats were treated to induce acute and chronic sialadenitis in the left and right submandibular glands, respectively. Contrast-enhanced CT and US of the glands were performed, followed by extirpation and histopathologic confirmation. Radiomics feature values of the glands were obtained from all images. Based on 3 feature selection methods, an optimal feature set was defined after a comparison of the receiver operating characteristic area under the curve (AUC) of each combination of 3 deep learning algorithms and 3 classification models.

RESULTS

The attribute features for the CT model were 2 gray-level run length matrices and 2 gray-level zone length matrices. In the US model, there were 2 gray-level co-occurrence matrices and 2 gray-level zone length matrices. The most accurate diagnostic models of CT and US yielded outstanding (AUC = 1.000) and excellent (AUC = 0.879) discrimination, respectively.

CONCLUSIONS

The radiomics diagnostic model using gray-level zone length matrices-based features conferred clinically outstanding discriminating ability among stages of sialadenitis using CT and excellent discrimination with US in almost all combinations of machine learning feature selections and classification models.

Predictive and prognostic potential of pretreatment 68Ga-PSMA PET tumor heterogeneity index in patients with metastatic castration-resistant prostate cancer treated with 177Lu-PSMA

INTRODUCTION

This study was conducted to evaluate the predictive values of volumetric parameters and radiomic features (RFs) extracted from pretreatment 68Ga-PSMA PET and baseline clinical parameters in response to 177Lu-PSMA therapy.

MATERIALS AND METHODS

In this retrospective multicenter study, mCRPC patients undergoing 177Lu-PSMA therapy were enrolled. According to the outcome of therapy, the patients were classified into two groups including positive biochemical response (BCR) (≥ 50% reduction in the serum PSA value) and negative BCR (< 50%). Sixty-five RFs, eight volumetric parameters, and also seventeen clinical parameters were evaluated for the prediction of BCR. In addition, the impact of such parameters on overall survival (OS) was evaluated.

RESULTS

33 prostate cancer patients with a median age of 69 years (range: 49-89) were enrolled. BCR was observed in 22 cases (66%), and 16 cases (48.5%) died during the follow-up time. The results of Spearman correlation test indicated a significant relationship between BCR and treatment cycle, administered dose, HISTO energy, GLCM entropy, and GLZLM LZLGE (p<0.05). In addition, according to the Mann-Whitney U test, age, cycle, dose, GLCM entropy, and GLZLM LZLGE were significantly different between BCR and non BCR patients (p<0.05). According to the ROC curve analysis for feature selection for prediction of BCR, GLCM entropy, age, treatment cycle, and administered dose showed acceptable results (p<0.05). According to SVM for assessing the best model for prediction of response to therapy, GLCM entropy alone showed the highest predictive performance in treatment planning. For the entire cohort, the Kaplan-Meier test revealed a median OS of 21 months (95% CI: 12.12-29.88). The median OS was estimated at 26 months (95% CI: 17.43-34.56) for BCR patients and 13 months (95% CI: 9.18-16.81) for non BCR patients. Among all variables included in the Kaplan Meier, the only response to therapy was statistically significant (p=0.01).

CONCLUSION

This exploratory study showed that the heterogeneity parameter of pretreatment 68Ga-PSMA PET images might be a potential predictive value for response to 177Lu-PSMA therapy in mCRPC; however, further prospective studies need to be carried out to verify these findings.

Exploratory analysis of radiomic as prognostic biomarkers in 18F-FDG PET/CT scan in uterine cervical cancer

Objective

To evaluate the performance of 18F-fluorodeoxyglucose positron emission tomography (¹⁸F-FDG PET/CT) radiomic features to predict overall survival (OS) in patients with locally advanced uterine cervical carcinoma.

Methods

Longitudinal and retrospective study that evaluated 50 patients with cervical epidermoid carcinoma (clinical stage IB2 to IVA according to FIGO). Segmentation of the 18F-FDG PET/CT tumors was performed using the LIFEx software, generating the radiomic features. We used the Mann-Whitney test to select radiomic features associated with the clinical outcome (death), excluding the features highly correlated with each other with Spearman correlation. Subsequently, ROC curves and a Kaplan-Meier analysis were performed. A p-value < 0.05 were considered significant.

Results

The median follow-up was 23.5 months and longer than 24 months in all surviving patients. Independent predictors for OS were found-SUVpeak with an AUC of 0.74, sensitivity of 77.8%, and specificity of 72.7% (p = 0.006); and the textural feature gray-level run-length matrix GLRLM_LRLGE, with AUC of 0.74, sensitivity of 72.2%, and specificity of 81.8% (p = 0.005). When we used the derived cut-off points from these ROC curves (12.76 for SUVpeak and 0.001 for GLRLM_LRLGE) in a Kaplan-Meier analysis, we can see two different groups (one with an overall survival probability of approximately 90% and the other with 30%). These biomarkers are independent of FIGO staging.

Conclusion

By radiomic ¹⁸F-FDG PET/CT data analysis, SUVpeak and GLRLM_LRLGE textural feature presented the best performance to predict OS in patients with cervical cancer undergoing chemo-radiotherapy and brachytherapy.

68Ga-PSMA-11 PET/CT Features Extracted from Different Radiomic Zones Predict Response to Androgen Deprivation Therapy in Patients with Advanced Prostate Cancer

Purpose: Prediction of treatment response to androgen deprivation therapy (ADT) prior to treatment initiation remains difficult. This study was undertaken to investigate whether 68Ga-PSMA-11 PET/CT features extracted from different radiomic zones within the prostate gland might predict response to ADT in patients with advanced prostate cancer (PCa). Methods: A total of 35 patients with prostate adenocarcinoma underwent two 68Ga-PSMA-11 PET/CT scans—termed PET-1 and PET-2—before and after 3 months of ADT, respectively. The prostate was divided into three radiomic zones, with zone-1 being the metabolic tumor zone, zone-2 the proximal peripheral tumor zone, and zone-3 the extended peripheral tumor zone. Patients in the response group were those who showed a reduction ratio > 30% for PET-derived parameters measured at PET-1 and PET-2. The remaining patients were classified as non-responders. Results: Seven features (glcm_idmn, glcm_idn, glcm_imc1, ngtdm_Contrast, glrlm_rln, gldm_dn, and shape_MeshVolume) from zone-1, two features (gldm_sdlgle and shape_MinorAxisLength) from zone-2, and two features (diagnostics_Mask-interpolated_Minimum and shape_Sphericity) from zone-3 successfully distinguished responders from non-responders to ADT. One predictive feature (shape_SurfaceVolumeRatio) was consistently identified in all of the three zones. Conclusions: this study demonstrates the potential usefulness of radiomic features extracted from different prostatic zones in distinguishing responders from non-responders prior to ADT initiation.

Could 18-FDG PET-CT Radiomic Features Predict the Locoregional Progression-Free Survival in Inoperable or Unresectable Oesophageal Cancer?

Background: We evaluated the value of pre-treatment positron-emission tomography−computed tomography (PET-CT)-based radiomic features in predicting the locoregional progression-free survival (LR-PFS) of patients with inoperable or unresectable oesophageal cancer. Material and Methods: Forty-six patients were included and 230 radiomic parameters were extracted. After a principal component analysis (PCA), we identified the more robust radiomic parameters, and we used them to develop a heatmap. Finally, we correlated these radiomic features with LR-PFS. Results: The median follow-up time was 17 months. The two-year LR-PFS and PFS rates were 35.9% (95% CI: 18.9−53.3) and 21.6% (95%CI: 10.0−36.2), respectively. After the correlation analysis, we identified 55 radiomic parameters that were included in the heatmap. According to the results of the hierarchical clustering, we identified two groups of patients presenting statistically different median LR-PFSs (22.8 months vs. 9.9 months; HR = 2.64; 95% CI 0.97−7.15; p = 0.0573). We also identified two radiomic features (“F_rlm_rl_entr_per” and “F_rlm_2_5D_rl_entr”) significantly associated with LR-PFS. Patients expressing a “F_rlm_2_5D_rl_entr” of <3.3 had a better median LR- PFS (29.4 months vs. 8.2 months; p = 0.0343). Patients presenting a “F_rlm_rl_entr_per” of <4.7 had a better median LR-PFS (50.4 months vs. 9.9 months; p = 0.0132). Conclusion: We identified two radiomic signatures associated with a lower risk of locoregional relapse after CRT.

Synergistic impact of motion and acquisition/reconstruction parameters on 18 F-FDG PET radiomic features in non-small cell lung cancer: phantom and clinical studies

Objectives

This study is aimed at examining the synergistic impact of motion and acquisition/reconstruction parameters on ¹⁸F‐FDG PET image radiomic features in non‐small cell lung cancer (NSCLC) patients, and investigating the robustness of features performance in differentiating NSCLC histopathology subtypes.

Methods

An in‐house developed thoracic phantom incorporating lesions with different sizes was used with different reconstruction settings, including various reconstruction algorithms, number of subsets and iterations, full‐width at half‐maximum of post‐reconstruction smoothing filter and acquisition parameters, including injected activity and test–retest with and without motion simulation. To simulate motion, a special motor was manufactured to simulate respiratory motion based on a normal patient in two directions. The lesions were delineated semi‐automatically to extract 174 radiomic features. All radiomic features were categorized according to the coefficient of variation (COV) to select robust features. A cohort consisting of 40 NSCLC patients with adenocarcinoma (n = 20) and squamous cell carcinoma (n = 20) was retrospectively analyzed. Statistical analysis was performed to discriminate robust features in differentiating histopathology subtypes of NSCLC lesions.

Results

Overall, 29% of radiomic features showed a COV ≤5% against motion. Forty‐five percent and 76% of the features showed a COV ≤ 5% against the test–retest with and without motion in large lesions, respectively. Thirty‐three percent and 45% of the features showed a COV ≤ 5% against different reconstruction parameters with and without motion, respectively. For NSCLC histopathological subtype differentiation, statistical analysis showed that 31 features were significant (p‐value < 0.05). Two out of the 31 significant features, namely, the joint entropy of GLCM (AUC = 0.71, COV = 0.019) and median absolute deviation of intensity histogram (AUC = 0.7, COV = 0.046), were robust against the motion (same reconstruction setting).

Conclusions

Motion, acquisition, and reconstruction parameters significantly impact radiomic features, just as their synergies. Radiomic features with high predictive performance (statistically significant) in differentiating histopathological subtype of NSCLC may be eliminated due to non‐reproducibility.

Computer-Aided Diagnosis of Spinal Tuberculosis From CT Images Based on Deep Learning With Multimodal Feature Fusion

Background

Spinal tuberculosis (TB) has the highest incidence in remote plateau areas, particularly in Tibet, China, due to inadequate local healthcare services, which not only facilitates the transmission of TB bacteria but also increases the burden on grassroots hospitals. Computer-aided diagnosis (CAD) is urgently required to improve the efficiency of clinical diagnosis of TB using computed tomography (CT) images. However, classical machine learning with handcrafted features generally has low accuracy, and deep learning with self-extracting features relies heavily on the size of medical datasets. Therefore, CAD, which effectively fuses multimodal features, is an alternative solution for spinal TB detection.

Methods

A new deep learning method is proposed that fuses four elaborate image features, specifically three handcrafted features and one convolutional neural network (CNN) feature. Spinal TB CT images were collected from 197 patients with spinal TB, from 2013 to 2020, in the People’s Hospital of Tibet Autonomous Region, China; 3,000 effective lumbar spine CT images were randomly screened to our dataset, from which two sets of 1,500 images each were classified as tuberculosis (positive) and health (negative). In addition, virtual data augmentation is proposed to enlarge the handcrafted features of the TB dataset. Essentially, the proposed multimodal feature fusion CNN consists of four main sections: matching network, backbone (ResNet-18/50, VGG-11/16, DenseNet-121/161), fallen network, and gated information fusion network. Detailed performance analyses were conducted based on the multimodal features, proposed augmentation, model stability, and model-focused heatmap.

Results

Experimental results showed that the proposed model with VGG-11 and virtual data augmentation exhibited optimal performance in terms of accuracy, specificity, sensitivity, and area under curve. In addition, an inverse relationship existed between the model size and test accuracy. The model-focused heatmap also shifted from the irrelevant region to the bone destruction caused by TB.

Conclusion

The proposed augmentation effectively simulated the real data distribution in the feature space. More importantly, all the evaluation metrics and analyses demonstrated that the proposed deep learning model exhibits efficient feature fusion for multimodal features. Our study provides a profound insight into the preliminary auxiliary diagnosis of spinal TB from CT images applicable to the Tibetan area.

Robustness and reproducibility of radiomics in T2 weighted images from magnetic resonance image guided linear accelerator in a phantom study

PURPOSE

Quantitative radiomics features extracted from medical images have been shown to provide value in predicting clinical outcomes. The study for robustness and reproducibility of radiomics features obtained with magnetic resonance image guided linear accelerator (MR-Linac) is insufficient. The objective of this work was to investigate the stability of radiomics features extracted from T2-weighted images of MR-Linac for five common effect factors.

MATERIALS AND METHOD

In this work, ten jellies, five fruits/vegetables, and a dynamic phantom were used to evaluate the impact of test-retest, intraobserver, varied thicknesses, radiation, and motion. These phantoms were scanned on a 1.5 T MRI system of MR-Linac. For test-retest data, the phantoms were scanned twice with repositioning within 15 min. To assess for intraobserver comparison, the segmentation of MR images was repeated by one observer in a double-blind manner. Three slice thicknesses (1.2 mm, 2.4 mm, and 4.8 mm) were used to select robust features that were insensitive to different thicknesses. The effect of radiation on features was studied by acquiring images when the beam was on. Common movement images of patients during radiotherapy were simulated by a dynamic phantom with five motion states to study the motion effect. A total of 1409 radiomics features, including shape features, first-order features, and texture features, were extracted from the original, wavelet, square, logarithmic, exponential and gradient images. The robustness and reproducibility features were evaluated using the concordance correlation coefficient (CCC).

RESULT

The intraobserver group had the most robust features (936/1079, 86.7%), while the group of motion effects had the lowest robustness (56/936, 6.0%), followed by the group of different thickness cohorts (374/936, 40.0%). The stability of features in the test-retest and radiation groups was 1072 of 1312 (81.7%) and 810 of 936 (86.5%), respectively. Overall, 25 of 1409 (2.4%) radiomics features remained robust in all five tests, mostly focusing on the image type of the wavelet. The number of stable features extracted from when the beam was on was less than that extracted when the beam was off. Shape features were the most robust of all of the features in all of the groups, excluding the motion group.

CONCLUSION

Compared with other factors fewer features remained robust to the effect of motion. This result emphasizes the need to consider the effect of respiration motion. The study for T2-weighted images from MR-Linac under different conditions will help us to build a robust predictive model applicable for radiotherapy.

Early prediction of tumour-response to radiotherapy in NSCLC patients

Objective. In this study we developed an automatic method to predict tumour volume and shape in weeks 3 and 4 of radiotherapy (RT), using cone-beam computed tomography (CBCT) scans acquired up to week 2, allowing identification of large tumour changes.Approach. 240 non-small cell lung cancer (NSCLC) patients, treated with 55 Gy in 20 fractions, were collected. CBCTs were rigidly registered to the planning CT. Intensity values were extracted in each voxel of the planning target volume across all CBCT images from days 1, 2, 3, 7 and 14. For each patient and in each voxel, four regression models were fitted to voxel intensity; applying linear, Gaussian, quadratic and cubic methods. These models predicted the intensity value for each voxel in weeks 3 and 4, and the tumour volume found by thresholding. Each model was evaluated by computing the root mean square error in pixel value and structural similarity index metric (SSIM) for all patients. Finally, the sensitivity and specificity to predict a 30% change in volume were calculated for each model.Main results. The linear, Gaussian, quadratic and cubic models achieved a comparable similarity score, the average SSIM for all patients was 0.94, 0.94, 0.90, 0.83 in week 3, respectively. At week 3, a sensitivity of 84%, 53%, 90% and 88%, and specificity of 99%, 100%, 91% and 42% were observed for the linear, Gaussian, quadratic and cubic models respectively. Overall, the linear model performed best at predicting those patients that will benefit from RT adaptation. The linear model identified 21% and 23% of patients in our cohort with more than 30% tumour volume reduction to benefit from treatment adaptation in weeks 3 and 4 respectively.Significance. We have shown that it is feasible to predict the shape and volume of NSCLC tumours from routine CBCTs and effectively identify patients who will respond to treatment early.

FDG-PET/CT in Lymphoma: Where Do We Go Now?

18F-fluorodeoxyglucose positron emission tomography combined with computed tomography (FDG-PET/CT) is an essential part of the management of patients with lymphoma at staging and response evaluation. Efforts to standardize PET acquisition and reporting, including the 5-point Deauville scale, have enabled PET to become a surrogate for treatment success or failure in common lymphoma subtypes. This review summarizes the key clinical-trial evidence that supports PET-directed personalized approaches in lymphoma but also points out the potential place of innovative PET/CT metrics or new radiopharmaceuticals in the future.

Transpathology: molecular imaging-based pathology

Pathology is the medical specialty concerned with the study of the disease nature and causes, playing a key role in bridging basic researches and clinical medicine. In the course of development, pathology has significantly expanded our understanding of disease, and exerted enormous impact on the management of patients. However, challenges facing pathology, the inherent invasiveness of pathological practice and the persistent concerns on the sample representativeness, constitute its limitations. Molecular imaging is a noninvasive technique to visualize, characterize, and measure biological processes at the molecular level in living subjects. With the continuous development of equipment and probes, molecular imaging has enabled an increasingly precise evaluation of pathophysiological changes. A new pathophysiology visualization system based on molecular imaging is forming and shows the great potential to reform the pathological practice. Several improvements in "trans-," including trans-scale, transparency, and translation, would be driven by this new kind of pathological practice. Pathological changes could be evaluated in a trans-scale imaging mode; tissues could be transparentized to better present the underlying pathophysiological information; and the translational processes of basic research to the clinical practice would be better facilitated. Thus, transpathology would greatly facilitate in deciphering the pathophysiological events in a multiscale perspective, and supporting the precision medicine in the future.

Supervised machine learning enables non-invasive lesion characterization in primary prostate cancer with [68Ga]Ga-PSMA-11 PET/MRI

PURPOSE

Risk classification of primary prostate cancer in clinical routine is mainly based on prostate-specific antigen (PSA) levels, Gleason scores from biopsy samples, and tumor-nodes-metastasis (TNM) staging. This study aimed to investigate the diagnostic performance of positron emission tomography/magnetic resonance imaging (PET/MRI) in vivo models for predicting low-vs-high lesion risk (LH) as well as biochemical recurrence (BCR) and overall patient risk (OPR) with machine learning.

METHODS

Fifty-two patients who underwent multi-parametric dual-tracer [18F]FMC and [68Ga]Ga-PSMA-11 PET/MRI as well as radical prostatectomy between 2014 and 2015 were included as part of a single-center pilot to a randomized prospective trial (NCT02659527). Radiomics in combination with ensemble machine learning was applied including the [68Ga]Ga-PSMA-11 PET, the apparent diffusion coefficient, and the transverse relaxation time-weighted MRI scans of each patient to establish a low-vs-high risk lesion prediction model (MLH). Furthermore, MBCR and MOPR predictive model schemes were built by combining MLH, PSA, and clinical stage values of patients. Performance evaluation of the established models was performed with 1000-fold Monte Carlo (MC) cross-validation. Results were additionally compared to conventional [68Ga]Ga-PSMA-11 standardized uptake value (SUV) analyses.

RESULTS

The area under the receiver operator characteristic curve (AUC) of the MLH model (0.86) was higher than the AUC of the [68Ga]Ga-PSMA-11 SUVmax analysis (0.80). MC cross-validation revealed 89% and 91% accuracies with 0.90 and 0.94 AUCs for the MBCR and MOPR models respectively, while standard routine analysis based on PSA, biopsy Gleason score, and TNM staging resulted in 69% and 70% accuracies to predict BCR and OPR respectively.

CONCLUSION

Our results demonstrate the potential to enhance risk classification in primary prostate cancer patients built on PET/MRI radiomics and machine learning without biopsy sampling.

Differentiation between non-small cell lung cancer and radiation pneumonitis after carbon-ion radiotherapy by 18F-FDG PET/CT texture analysis

The differentiation of non-small cell lung cancer (NSCLC) and radiation pneumonitis (RP) is critically essential for selecting optimal clinical therapeutic strategies to manage post carbon-ion radiotherapy (CIRT) in patients with NSCLC. The aim of this study was to assess the ability of ¹⁸F-FDG PET/CT metabolic parameters and its textural image features to differentiate NSCLC from RP after CIRT to develop a differential diagnosis of malignancy and benign lesion. We retrospectively analyzed ¹⁸F-FDG PET/CT image data from 32 patients with histopathologically proven NSCLC who were scheduled to undergo CIRT and 31 patients diagnosed with RP after CIRT. The SUV parameters, metabolic tumor volume (MTV), total lesion glycolysis (TLG) as well as fifty-six texture parameters derived from seven matrices were determined using PETSTAT image-analysis software. Data were statistically compared between NSCLC and RP using Wilcoxon rank-sum tests. Diagnostic accuracy was assessed using receiver operating characteristics (ROC) curves. Several texture parameters significantly differed between NSCLC and RP (p < 0.05). The parameters that were high in areas under the ROC curves (AUC) were as follows: SUV_max, 0.64; GLRLM run percentage, 0.83 and NGTDM coarseness, 0.82. Diagnostic accuracy was improved using GLRLM run percentage or NGTDM coarseness compared with SUV_max (p < 0.01). The texture parameters of ¹⁸F-FDG uptake yielded excellent outcomes for differentiating NSCLC from radiation pneumonitis after CIRT, which outperformed SUV-based evaluation. In particular, GLRLM run percentage and NGTDM coarseness of ¹⁸F-FDG PET/CT images would be appropriate parameters that can offer high diagnostic accuracy.

Radiomics in immuno-oncology

With the ongoing advances in imaging techniques, increasing volumes of anatomical and functional data are being generated as part of the routine clinical workflow. This surge of available imaging data coincides with increasing research in quantitative imaging, particularly in the domain of imaging features. An important and novel approach is radiomics, where high-dimensional image properties are extracted from routine medical images. The fundamental principle of radiomics is the hypothesis that biomedical images contain predictive information, not discernible to the human eye, that can be mined through quantitative image analysis. In this review, a general outline of radiomics and artificial intelligence (AI) will be provided, along with prominent use cases in immunotherapy (e.g. response and adverse event prediction) and targeted therapy (i.e. radiogenomics). While the increased use and development of radiomics and AI in immuno-oncology is highly promising, the technology is still in its early stages, and different challenges still need to be overcome. Nevertheless, novel AI algorithms are being constructed with an ever-increasing scope of applications.

Radiomics and Dosiomics for Predicting Local Control after Carbon-Ion Radiotherapy in Skull-Base Chordoma

Skull-base chordoma (SBC) can be treated with carbon ion radiotherapy (CIRT) to improve local control (LC). The study aimed to explore the role of multi-parametric radiomic, dosiomic and clinical features as prognostic factors for LC in SBC patients undergoing CIRT. Before CIRT, 57 patients underwent MR and CT imaging, from which tumour contours and dose maps were obtained. MRI and CT-based radiomic, and dosiomic features were selected and fed to two survival models, singularly or by combining them with clinical factors. Adverse LC was given by in-field recurrence or tumour progression. The dataset was split in development and test sets and the models' performance evaluated using the concordance index (C-index). Patients were then assigned a low- or high-risk score. Survival curves were estimated, and risk groups compared through log-rank tests (after Bonferroni correction α = 0.0083). The best performing models were built on features describing tumour shape and dosiomic heterogeneity (median/interquartile range validation C-index: 0.80/024 and 0.79/0.26), followed by combined (0.73/0.30 and 0.75/0.27) and CT-based models (0.77/0.24 and 0.64/0.28). Dosiomic and combined models could consistently stratify patients in two significantly different groups. Dosiomic and multi-parametric radiomic features showed to be promising prognostic factors for LC in SBC treated with CIRT.

Computed tomography-based skeletal segmentation for quantitative PET metrics of bone involvement in multiple myeloma

Purpose

Quantifications in nuclear medicine are occasionally limited by the lack of standardization for defining volumes of interest (VOIs) on functional images. In the present article, we propose the use of computed tomography (CT)–based skeletal segmentation to determine anatomically the VOI in order to calculate quantitative parameters of fluorine 18 fluorodeoxyglucose (¹⁸F-FDG) PET/CT images from patients with multiple myeloma.

Methods

We evaluated 101 whole-body ¹⁸F-FDG PET/CTs of 58 patients with multiple myeloma. An initial subjective visual analysis of the PET images was used to classify the bone involvement as negative/mild, moderate, or marked. Then, a fully automated CT–based segmentation of the skeleton was performed on PET images. The maximum, mean, and SD of the standardized uptake values (SUV_max, SUV_mean, and SD_SUV) were calculated for bone tissue and compared with the visual analysis.

Results

Forty-five (44.5%), 32 (31.7%), and 24 (23.8%) PET images were, respectively, classified as negative/mild, moderate, or marked bone involvement. All quantitative parameters were significantly related to the visual assessment of bone involvement. This association was stronger for the SUV_mean [odds ratio (OR): 10.52 (95% confidence interval (CI), 5.68–19.48); P < 0.0001] and for the SD_SUV [OR: 5.58 (95% CI, 3.31–9.42); P < 0.001) than for the SUV_max [OR: 1.01 (95% CI, 1.003–1.022); P = 0.003].

Conclusion

CT–based skeletal segmentation allows for automated and therefore reproducible calculation of PET quantitative parameters of bone involvement in patients with multiple myeloma. Using this method, the SUV_mean and its respective SD correlated better with the visual analysis of ¹⁸F-FDG PET images than SUVmax. Its value in staging and evaluating therapy response needs to be evaluated.

Prediction of lymph node metastases using pre-treatment PET radiomics of the primary tumour in esophageal adenocarcinoma: an external validation study

OBJECTIVES

To improve clinical lymph node staging (cN-stage) in oesophageal adenocarcinoma by developing and externally validating three prediction models; one with clinical variables only, one with positron emission tomography (PET) radiomics only, and a combined clinical and radiomics model.

METHODS

Consecutive patients with fluorodeoxyglucose (FDG) avid tumours treated with neoadjuvant therapy between 2010 and 2016 in two international centres (n = 130 and n = 60, respectively) were included. Four clinical variables (age, gender, clinical T-stage and tumour regression grade) and PET radiomics from the primary tumour were used for model development. Diagnostic accuracy, area under curve (AUC), discrimination and calibration were calculated for each model. The prognostic significance was also assessed.

RESULTS

The incidence of lymph node metastases was 58% in both cohorts. The areas under the curve of the clinical, radiomics and combined models were 0.79, 0.69 and 0.82 in the developmental cohort, and 0.65, 0.63 and 0.69 in the external validation cohort, with good calibration demonstrated. The area under the curve of current cN-stage in development and validation cohorts was 0.60 and 0.66, respectively. For overall survival, the combined clinical and radiomics model achieved the best discrimination performance in the external validation cohort (X² = 6.08, df = 1, p = 0.01).

CONCLUSION

Accurate diagnosis of lymph node metastases is crucial for prognosis and guiding treatment decisions. Despite finding improved predictive performance in the development cohort, the models using PET radiomics derived from the primary tumour were not fully replicated in an external validation cohort.

ADVANCES IN KNOWLEDGE

This international study attempted to externally validate a new prediction model for lymph node metastases using PET radiomics. A model combining clinical variables and PET radiomics improved discrimination of lymph node metastases, but these results were not externally replicated.

Prognostic value of FDG-PET radiomics with machine learning in pancreatic cancer

Patients with pancreatic cancer have a poor prognosis, therefore identifying particular tumor characteristics associated with prognosis is important. This study aims to investigate the utility of radiomics with machine learning using 18F-fluorodeoxyglucose (FDG)-PET in patients with pancreatic cancer. We enrolled 161 patients with pancreatic cancer underwent pretreatment FDG-PET/CT. The area of the primary tumor was semi-automatically contoured with a threshold of 40% of the maximum standardized uptake value, and 42 PET features were extracted. To identify relevant PET parameters for predicting 1-year survival, Gini index was measured using random forest (RF) classifier. Twenty-three patients were censored within 1 year of follow-up, and the remaining 138 patients were used for the analysis. Among the PET parameters, 10 features showed statistical significance for predicting overall survival. Multivariate analysis using Cox HR regression revealed gray-level zone length matrix (GLZLM) gray-level non-uniformity (GLNU) as the only PET parameter showing statistical significance. In RF model, GLZLM GLNU was the most relevant factor for predicting 1-year survival, followed by total lesion glycolysis (TLG). The combination of GLZLM GLNU and TLG stratified patients into three groups according to risk of poor prognosis. Radiomics with machine learning using FDG-PET in patients with pancreatic cancer provided useful prognostic information.

Dual tracer 68Ga-DOTATOC and 18F-FDG PET/computed tomography radiomics in pancreatic neuroendocrine neoplasms: an endearing tool for preoperative risk assessment

AIM

To explore the potentiality of radiomics analysis, performed on Ga-DOTATOC and fluorine-18-fluorodeoxyglucose (F-FDG) PET/computed tomography (CT) images, in predicting tumour aggressiveness and outcome in patients candidate to surgery for pancreatic neuroendocrine neoplasms (PanNENs).

PATIENTS AND METHODS

Retrospective study including 61 patients who underwent Ga-DOTATOC and F-FDG PET/CT before surgery for PanNEN. Semiquantitative variables [SUVmax and somatostatin receptor density (SRD) for Ga-DOTATOC PET; SUVmax and MTV for F-FDG PET] and texture features [intensity variability, size zone variability (SZV), zone percentage, entropy; homogeneity, dissimilarity and coefficient of variation (Co-V)] have been analysed to evaluate their possible role in predicting tumour characteristics. Principal component analysis (PCA) was firstly performed and then multiple regression analyses were performed by using the extracted principal components.

RESULTS

Regarding Ga-DOTATOC PET, SZV, entropy, intensity variability and SRD were predictive for tumour dimension. Regarding F-FDG PET, intensity variability, SZV, homogeneity, SUVmax and MTV were predictive for tumour dimension. Four principal components were extracted from PCA: PC1 correlated with all F-FDG variables, while PC2, PC3 and PC4 with Ga-DOTATOC variables. PC1 was the only significantly predicting angioinvasion (P = 0.0222); PC4 was the only one significantly predicting lymph nodal involvement (P = 0.0151). All principal components except PC4 significantly predicted tumour dimension (P  <0.0001 for PC1, P = 0.0016 for PC2 and P < 0.0001 for PC3). Co-V from Ga-DOTATOC PET/CT was predictive of the outcome.

CONCLUSION

Specific texture features derived from preoperative Ga-DOTATOC and F-FDG PET/CT could noninvasively predict specific tumour characteristics and patients' outcome, delineating the potential role of dual tracer technique and texture analysis in the risk assessment of patients with PanNENs.

Radiomics for precision medicine: Current challenges, future prospects, and the proposal of a new framework

The advancement of artificial intelligence concurrent with the development of medical imaging techniques provided a unique opportunity to turn medical imaging from mostly qualitative, to further quantitative and mineable data that can be explored for the development of clinical decision support systems (cDSS). Radiomics, a method for the high throughput extraction of hand-crafted features from medical images, and deep learning -the data driven modeling techniques based on the principles of simplified brain neuron interactions, are the most researched quantitative imaging techniques. Many studies reported on the potential of such techniques in the context of cDSS. Such techniques could be highly appealing due to the reuse of existing data, automation of clinical workflows, minimal invasiveness, three-dimensional volumetric characterization, and the promise of high accuracy and reproducibility of results and cost-effectiveness. Nevertheless, there are several challenges that quantitative imaging techniques face, and need to be addressed before the translation to clinical use. These challenges include, but are not limited to, the explainability of the models, the reproducibility of the quantitative imaging features, and their sensitivity to variations in image acquisition and reconstruction parameters. In this narrative review, we report on the status of quantitative medical image analysis using radiomics and deep learning, the challenges the field is facing, propose a framework for robust radiomics analysis, and discuss future prospects.

2-[18F]FDG PET/CT radiomics in lung cancer: An overview of the technical aspect and its emerging role in management of the disease

Lung cancer is the most common cancer, worldwide, and a major health issue with a remarkable mortality rate. 2-[¹⁸F]fluoro-2-deoxy-D-glucose positron emission tomography/computed tomography (2-[¹⁸F]FDG PET/CT) plays an indispensable role in the management of lung cancer patients. Long-established quantitative parameters such as size, density, and metabolic activity have been and are being employed in the current practice to enhance interpretation and improve diagnostic and prognostic value. The introduction of radiomics analysis revolutionized the quantitative evaluation of medical imaging, revealing data within images beyond visual interpretation. The "big data" are extracted from high-quality images and are converted into information that correlates to relevant genetic, pathologic, clinical, or prognostic features. Technically advanced, diverse methods have been implemented in different studies. The standardization of image acquisition, segmentation and features analysis is still a debated issue. Importantly, a body of features has been extracted and employed for diagnosis, staging, risk stratification, prognostication, and therapeutic response. 2-[¹⁸F]FDG PET/CT-derived features show promising value in non-invasively diagnosing the malignant nature of pulmonary nodules, differentiating lung cancer subtypes, and predicting response to different therapies as well as survival. In this review article, we aimed to provide an overview of the technical aspects used in radiomics analysis in non-small cell lung cancer (NSCLC) and elucidate the role of 2-[¹⁸F]FDG PET/CT-derived radiomics in the diagnosis, prognostication, and therapeutic response.

Transcriptomics in cancer revealed by Positron Emission Tomography radiomics

Metabolic images from Positron Emission Tomography (PET) are used routinely for diagnosis, follow-up or treatment planning purposes of cancer patients. In this study we aimed at determining if radiomic features extracted from 18F-Fluoro Deoxy Glucose (FDG) PET images could mirror tumor transcriptomics. In this study we analyzed 45 patients with locally advanced head and neck cancer (H&N) that underwent FDG-PET scans at the time of diagnosis and transcriptome analysis using RNAs from both cancer and healthy tissues on microarrays. Association between PET radiomics and transcriptomics was carried out with the Genomica software and a functional annotation was used to associate PET radiomics, gene expression and altered biological pathways. We identified relationships between PET radiomics and genes involved in cell-cycle, disease, DNA repair, extracellular matrix organization, immune system, metabolism or signal transduction pathways, according to the Reactome classification. Our results suggest that these FDG PET radiomic features could be used to infer tissue gene expression and cellular pathway activity in H&N cancers. These observations strengthen the value of radiomics as a promising approach to personalize treatments through targeting tumor-specific molecular processes.

What can artificial intelligence teach us about the molecular mechanisms underlying disease?

While molecular imaging with positron emission tomography or single-photon emission computed tomography already reports on tumour molecular mechanisms on a macroscopic scale, there is increasing evidence that there are multiple additional features within medical images that can further improve tumour characterization, treatment prediction and prognostication. Early reports have already revealed the power of radiomics to personalize and improve patient management and outcomes. What remains unclear is how these additional metrics relate to underlying molecular mechanisms of disease. Furthermore, the ability to deal with increasingly large amounts of data from medical images and beyond in a rapid, reproducible and transparent manner is essential for future clinical practice. Here, artificial intelligence (AI) may have an impact. AI encompasses a broad range of 'intelligent' functions performed by computers, including language processing, knowledge representation, problem solving and planning. While rule-based algorithms, e.g. computer-aided diagnosis, have been in use for medical imaging since the 1990s, the resurgent interest in AI is related to improvements in computing power and advances in machine learning (ML). In this review we consider why molecular and cellular processes are of interest and which processes have already been exposed to AI and ML methods as reported in the literature. Non-small-cell lung cancer is used as an exemplar and the focus of this review as the most common tumour type in which AI and ML approaches have been tested and to illustrate some of the concepts.

Prediction of Overall Survival and Progression-Free Survival by the 18F-FDG PET/CT Radiomic Features in Patients with Primary Gastric Diffuse Large B-Cell Lymphoma

Purpose. To determine whether the radiomic features of ¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) contribute to prognosis prediction in primary gastric diffuse large B-cell lymphoma (PG-DLBCL) patients. Methods. This retrospective study included 35 PG-DLBCL patients who underwent PET/CT scans at West China Hospital before curative treatment. The volume of interest (VOI) was drawn around the tumor, and radiomic analysis of the PET and CT images, within the same VOI, was conducted. The metabolic and textural features of PET and CT images were evaluated. Correlations of the extracted features with the overall survival (OS) and progression-free survival (PFS) were evaluated. Univariate and multivariate analyses were conducted to assess the prognostic value of the radiomic parameters. Results. In the univariate model, many of the textural features, including kurtosis and volume, extracted from the PET and CT datasets were significantly associated with survival (5 for OS and 7 for PFS (PET); 7 for OS and 14 for PFS (CT)). Multivariate analysis identified kurtosis (hazard ratio (HR): 28.685, 95% confidence interval (CI): 2.067–398.152, p=0.012), metabolic tumor volume (MTV) (HR: 26.152, 95% CI: 2.089–327.392, p=0.011), and gray-level nonuniformity (GLNU) (HR: 14.642, 95% CI: 2.661–80.549, p=0.002) in PET and sphericity (HR: 11.390, 95% CI: 1.360–95.371, p=0.025) and kurtosis (HR: 11.791, 95% CI: 1.583–87.808, p=0.016), gray-level nonuniformity (GLNU) (HR: 6.934, 95% CI: 1.069–44.981, p=0.042), and high gray-level zone emphasis (HGZE) (HR: 9.805, 95% CI: 1.359–70.747, p=0.024) in CT as independent prognostic factors. Conclusion. ¹⁸F-FDG PET/CT radiomic features are potentially useful for survival prediction in PG-DLBCL patients. However, studies with larger cohorts are needed to confirm the clinical prognostication of these parameters.

11C-methionine-PET for differentiating recurrent brain tumor from radiation necrosis: radiomics approach with random forest classifier

Differentiating recurrent brain tumor from radiation necrosis is often difficult. This study aims to investigate the efficacy of 11C-methionine (MET)-PET radiomics for distinguishing recurrent brain tumor from radiation necrosis, as compared with conventional tumor-to-normal cortex (T/N) ratio evaluation. We enrolled 41 patients with metastatic brain tumor or glioma treated using radiation therapy who underwent MET-PET. The area with a standardized uptake value > 1.3 times that of the normal brain cortex was contoured. Forty-two PET features were extracted and used in a random forest classifier and the diagnostic performance was evaluated using a 10-fold cross-validation scheme. Gini index was measured to identify relevant PET parameters for classification. The reference standard was surgical histopathological analysis or more than 6 months of follow-up with MRI. Forty-four lesions were used for the analysis. Thirty-three and 11 lesions were confirmed as recurrent brain tumor and radiation necrosis, respectively. Radiomics and T/N ratio evaluation showed sensitivities of 90.1% and 60.6%, and specificities of 93.9% and 72.7% with areas under the curve of 0.98 and 0.73, respectively. Gray level co-occurrence matrix dissimilarity was the most pertinent feature for diagnosis. MET-PET radiomics yielded excellent outcome for differentiating recurrent brain tumor from radiation necrosis, which outperformed T/N ratio evaluation.

Exploring Tumor Heterogeneity Using PET Imaging: The Big Picture

Personalized medicine represents a major goal in oncology. It has its underpinning in the identification of biomarkers with diagnostic, prognostic, or predictive values. Nowadays, the concept of biomarker no longer necessarily corresponds to biological characteristics measured ex vivo but includes complex physiological characteristics acquired by different technologies. Positron-emission-tomography (PET) imaging is an integral part of this approach by enabling the fine characterization of tumor heterogeneity in vivo in a non-invasive way. It can effectively be assessed by exploring the heterogeneous distribution and uptake of a tracer such as 18F-fluoro-deoxyglucose (FDG) or by using multiple radiopharmaceuticals, each providing different information. These two approaches represent two avenues of development for the research of new biomarkers in oncology. In this article, we review the existing evidence that the measurement of tumor heterogeneity with PET imaging provide essential information in clinical practice for treatment decision-making strategy, to better select patients with poor prognosis for more intensive therapy or those eligible for targeted therapy.

Radiomics features analysis of PET images in oropharyngeal and hypopharyngeal cancer

This study used radiomics image analysis to examine the differences of texture feature values extracted from oropharyngeal and hypopharyngeal cancer positron emission tomography (PET) images on various tumor segmentations, and finds the proper and stable feature groups. A total of 80 oropharyngeal and hypopharyngeal cancer cases were retrospectively recruited. Radiomics method was applied to the PET image for the 80 oropharyngeal and hypopharyngeal cancer cases to extract texture features from various defined metabolic volumes. Kruskal-Wallis one-way analysis of variance method was used to test whether feature value difference exists between groups, which were grouped by stage, response to treatment, and recurrence. If there was a significant difference, the corresponding feature cutoff value was applied to the Kaplan-Meier estimator to estimate the survival functions. For the various defined metabolic volumes, there were 16 features that had significant differences between early (T1, T2) and late tumor stages (T3, T4). Five images and 2 textural features were found to be able to predict the tumor response and recurrence, respectively, with the areas under the receiver operating characteristic curves reaching 0.7. The histogram entropy was found to be a good predictor of overall survival (OS) and primary relapse-free survival (PRFS) of oropharyngeal and hypopharyngeal cancer patients. Textural features from PET images provide predictive and prognostic information in tumor staging, tumor response, recurrence, and have the potential to be a prognosticator for OS and PRFS in oropharyngeal and hypopharyngeal cancer.

Texture analysis of computed tomography data using morphologic and metabolic delineation of esophageal cancer-relation to tumor type and neoadjuvant therapy response

The prognostic values of image-based tumor texture analysis based on computed tomography (CT) and of limiting the segmented tumor volume to metabolically active regions using fludeoxyglucose-positron emission tomography (FDG-PET) were studied in 25 patients with esophageal adenocarcinoma and 11 patients with squamous cell carcinoma. The aims of this study are to describe their CT-image-based texture characteristics before and after neoadjuvant therapy and to evaluate whether limiting the examined tumor volume to metabolically active regions detected with FDG-PET image data would further improve their value. Textural parameters (homogeneity, energy, entropy, contrast, and correlation) based on gray-level co-occurrence matrices (GLCM) were calculated for 3D volumes of segmented esophageal tumors before and after neoadjuvant chemotherapy or radiochemotherapy. Histopathological data after surgical resection and textural parameters before and after neoadjuvant treatment were compared using the Mann-Whitney U test. Significant differences in the textural parameters were observed between adenocarcinoma and squamous cell carcinoma for homogeneity, energy, inertia, and correlation. The use of contrast media during scanning resulted in significant differences in homogeneity, energy, entropy, and inertia for adenocarcinoma but not squamous cell carcinoma. There was also a significant difference in all textural parameters between pathological T status for ypT0-ypT2 and ypT3-ypT4 adenocarcinomas, but not in squamous cell carcinoma patients. No additional value was found from using PET image data to aid segmentation of CT images.

A PET Radiomics Model to Predict Refractory Mediastinal Hodgkin Lymphoma

First-order radiomic features, such as metabolic tumor volume (MTV) and total lesion glycolysis (TLG), are associated with disease progression in early-stage classical Hodgkin lymphoma (HL). We hypothesized that a model incorporating first- and second-order radiomic features would more accurately predict outcome than MTV or TLG alone. We assessed whether radiomic features extracted from baseline PET scans predicted relapsed or refractory disease status in a cohort of 251 patients with stage I-II HL who were managed at a tertiary cancer center. Models were developed and tested using a machine-learning algorithm. Features extracted from mediastinal sites were highly predictive of primary refractory disease. A model incorporating 5 of the most predictive features had an area under the curve (AUC) of 95.2% and total error rate of 1.8%. By comparison, the AUC was 78% for both MTV and TLG and was 65% for maximum standardize uptake value (SUV_max). Furthermore, among the patients with refractory mediastinal disease, our model distinguished those who were successfully salvaged from those who ultimately died of HL. We conclude that our PET radiomic model may improve upfront stratification of early-stage HL patients with mediastinal disease and thus contribute to risk-adapted, individualized management.

Response-to-repeatability of quantitative imaging features for longitudinal response assessment

Quantitative imaging biomarkers (QIBs) are often selected and ranked based on their repeatability performance. In the context of treatment response assessment, however, one must also consider how sensitive a QIB is to measuring changes in the tumour. This work introduces response-to-repeatability ratio (R/R), which weighs the ability of a QIB to detect significant changes with respect to its measurement repeatability and applies it to the case of PET texture features. R/R is evaluated as the proportion of measurable changes from baseline to follow-up for each candidate QIB. We analyse 47 texture features extracted from lesions in bone-metastatic prostate cancer patients who received double baseline and/or baseline to treatment follow-up ¹⁸F-NaF PET/CT scans. R/R evaluates the proportion of follow-up changes outside of the 95% limits of agreement (LOA) defined by test-retest values. Intraclass correlation coefficient (ICC) and coefficient of variation (CV) are calculated for each feature. Relationship between ICC and R/R are evaluated with the Spearman's correlation coefficient. R/R varied significantly across texture features: 41/47 (87%) features demonstrated R/R  >  5%; 21/47 (45%) features demonstrated R/R  >  10%, and 11/47 (23%) features demonstrated R/R  >  20%. LOA of features ranged from [0.998, 1.001] to [0.22, 4.86]. Repeatability alone did not qualify a feature for its efficacy at detecting measurable change at follow-up, as shown by weak correlations between R/R and both CV and ICC (ρ  =  0.23 and ρ  =  0.40, respectively). Three features demonstrated excellent ICC (ICC  >  0.75) and R/R greater than that of SUV_max (R/R  =  41.8%): skewness (ICC  =  0.92, R/R  =  75.4%), kurtosis (ICC  =  0.88, R/R  =  47.0%) and diagonal moment (ICC  =  0.88, R/R  =  45.5%). R/R characterizes the sensitivity of candidate QIBs to detect measurable changes at follow-up. R/R supplements existing precision performance metrics (e.g. CV, ICC, and LOA) as an index to assess the utility of QIBs for response assessment.

A pilot study using kernelled support tensor machine for distant failure prediction in lung SBRT

We developed a kernelled support tensor machine (KSTM)-based model with tumor tensors derived from pre-treatment PET and CT imaging as input to predict distant failure in early stage non-small cell lung cancer (NSCLC) treated with stereotactic body radiation therapy (SBRT). The patient cohort included 110 early stage NSCLC patients treated with SBRT, 25 of whom experienced failure at distant sites. Three-dimensional tumor tensors were constructed and used as input for the KSTM-based classifier. A KSTM iterative algorithm with a convergent proof was developed to train the weight vectors for every mode of the tensor for the classifier. In contrast to conventional radiomics approaches that rely on handcrafted imaging features, the KSTM-based classifier uses 3D imaging as input, taking full advantage of the imaging information. The KSTM-based classifier preserves the intrinsic 3D geometry structure of the medical images and the correlation in the original images and trains the classification hyper-plane in an adaptive feature tensor space. The KSTM-based predictive algorithm was compared with three conventional machine learning models and three radiomics approaches. For PET and CT, the KSTM-based predictive method achieved the highest prediction results among the seven methods investigated in this study based on 10-fold cross validation and independent testing.

Challenges and Promises of PET Radiomics

PURPOSE

Radiomics describes the extraction of multiple, otherwise invisible, features from medical images that, with bioinformatic approaches, can be used to provide additional information that can predict underlying tumor biology and behavior.

METHODS AND MATERIALS

Radiomic signatures can be used alone or with other patient-specific data to improve tumor phenotyping, treatment response prediction, and prognosis, noninvasively. The data describing 18F-fluorodeoxyglucose positron emission tomography radiomics, often using texture or heterogeneity parameters, are increasing rapidly.

RESULTS

In relation to radiation therapy practice, early data have reported the use of radiomic approaches to better define tumor volumes and predict radiation toxicity and treatment response.

CONCLUSIONS

Although at an early stage of development, with many technical challenges remaining and a need for standardization, promise nevertheless exists that PET radiomics will contribute to personalized medicine, especially with the availability of increased computing power and the development of machine-learning approaches for imaging.

External validation of a prognostic model incorporating quantitative PET image features in oesophageal cancer

AIM

Enhanced prognostic models are required to improve risk stratification of patients with oesophageal cancer so treatment decisions can be optimised. The primary aim was to externally validate a published prognostic model incorporating PET image features. Transferability of the model was compared using only clinical variables.

METHODS

This was a Transparent Reporting of a multivariate prediction model for Individual Prognosis Or Diagnosis (TRIPOD) type 3 study. The model was validated against patients treated with neoadjuvant chemoradiotherapy according to the Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS) trial regimen using pre- and post-harmonised image features. The Kaplan-Meier method with log-rank significance tests assessed risk strata discrimination. A Cox proportional hazards model assessed model calibration. Primary outcome was overall survival (OS).

RESULTS

Between 2010 and 2015, 449 patients were included in the development (n = 302), internal validation (n = 101) and external validation (n = 46) cohorts. No statistically significant difference in OS between patient quartiles was demonstrated in prognostic models incorporating PET image features (X² = 1.42, df = 3, p = 0.70) or exclusively clinical variables (age, disease stage and treatment; X² = 1.19, df = 3, p = 0.75). The calibration slope β of both models was not significantly different from unity (p = 0.29 and 0.29, respectively). Risk groups defined using only clinical variables suggested differences in OS, although these were not statistically significant (X² = 0.71, df = 2, p = 0.70).

CONCLUSION

The prognostic model did not enable significant discrimination between the validation risk groups, but a second model with exclusively clinical variables suggested some transferable prognostic ability. PET harmonisation did not significantly change the results of model validation.

Treatment Outcome Prediction for Cancer Patients based on Radiomics and Belief Function Theory

In this study, we proposed a new radiomics-based treatment outcome prediction model for cancer patients. The prediction model is developed based on belief function theory (BFT) and sparsity learning to address the challenges of redundancy, heterogeneity, and uncertainty of radiomic features, and relatively small-sized and unbalanced training samples. The model first selects the most predictive feature subsets from relatively large amounts of radiomic features extracted from pre- and/or in-treatment positron emission tomography (PET) images and available clinical and demographic features. Then an evidential k-nearest neighbor (EK-NN) classifier is proposed to utilize the selected features for treatment outcome prediction. Twenty-five stage II-III lung, 36 esophagus, 63 stage II-III cervix, and 45 lymphoma cancer patient cases were included in this retrospective study. Performance and robustness of the proposed model were assessed with measures of feature selection stability, outcome prediction accuracy, and receiver operating characteristics (ROC) analysis. Comparison with other methods were conducted to demonstrate the feasibility and superior performance of the proposed model.

Noninvasive Imaging of Drug-Induced Liver Injury with 18F-DFA PET

Drug-induced liver failure is a significant indication for a liver transplant, and unexpected liver toxicity is a major reason that otherwise effective therapies are removed from the market. Various methods exist for monitoring liver injury but are often inadequate to predict liver failure. New diagnostic tools are needed. Methods: We evaluate in a preclinical model whether 18F-2-deoxy-2-fluoroarabinose (18F-DFA), a PET radiotracer that measures the ribose salvage pathway, can be used to monitor acetaminophen-induced liver injury and failure. Mice treated with vehicle, 100, 300, or 500 mg/kg acetaminophen for 7 or 21 h were imaged with 18F-FDG and 18F-DFA PET. Hepatic radiotracer accumulation was correlated to survival and percentage of nonnecrotic tissue in the liver. Mice treated with acetaminophen and vehicle or N-acetylcysteine were imaged with 18F-DFA PET. 18F-DFA accumulation was evaluated in human hepatocytes engrafted into the mouse liver. Results: We show that hepatic 18F-DFA accumulation is 49%-52% lower in mice treated with high-dose acetaminophen than in mice treated with low-dose acetaminophen or vehicle. Under these same conditions, hepatic 18F-FDG accumulation was unaffected. At 21 h after acetaminophen treatment, hepatic 18F-DFA accumulation can distinguish mice that will succumb to the liver injury from those that will survive it (6.2 vs. 9.7 signal to background, respectively). Hepatic 18F-DFA accumulation in this model provides a tomographic representation of hepatocyte density in the liver, with a R2 between hepatic 18F-DFA accumulation and percentage of nonnecrotic tissue of 0.70. PET imaging with 18F-DFA can be used to distinguish effective from ineffective resolution of acetaminophen-induced liver injury with N-acetylcysteine (15.6 vs. 6.2 signal to background, respectively). Human hepatocytes, in culture or engrafted into a mouse liver, have levels of ribose salvage activity similar to those of mouse hepatocytes. Conclusion: Our findings suggest that PET imaging with 18F-DFA can be used to visualize and quantify drug-induced acute liver injury and may provide information on the progression from liver injury to hepatic failure.