Classification and evaluation strategies of auto-segmentation approaches for PET: Report of AAPM task group No. 211

The effect of harmonization on the variability of PET radiomic features extracted using various segmentation methods

PURPOSE

This study aimed to examine the robustness of positron emission tomography (PET) radiomic features extracted via different segmentation methods before and after ComBat harmonization in patients with non-small cell lung cancer (NSCLC).

METHODS

We included 120 patients (positive recurrence = 46 and negative recurrence = 74) referred for PET scanning as a routine part of their care. All patients had a biopsy-proven NSCLC. Nine segmentation methods were applied to each image, including manual delineation, K-means (KM), watershed, fuzzy-C-mean, region-growing, local active contour (LAC), and iterative thresholding (IT) with 40, 45, and 50% thresholds. Diverse image discretizations, both without a filter and with different wavelet decompositions, were applied to PET images. Overall, 6741 radiomic features were extracted from each image (749 radiomic features from each segmented area). Non-parametric empirical Bayes (NPEB) ComBat harmonization was used to harmonize the features. Linear Support Vector Classifier (LinearSVC) with L1 regularization For feature selection and Support Vector Machine classifier (SVM) with fivefold nested cross-validation was performed using StratifiedKFold with 'n_splits' set to 5 to predict recurrence in NSCLC patients and assess the impact of ComBat harmonization on the outcome.

RESULTS

From 749 extracted radiomic features, 206 (27%) and 389 (51%) features showed excellent reliability (ICC ≥ 0.90) against segmentation method variation before and after NPEB ComBat harmonization, respectively. Among all, 39 features demonstrated poor reliability, which declined to 10 after ComBat harmonization. The 64 fixed bin widths (without any filter) and wavelets (LLL)-based radiomic features set achieved the best performance in terms of robustness against diverse segmentation techniques before and after ComBat harmonization. The first-order and GLRLM and also first-order and NGTDM feature families showed the largest number of robust features before and after ComBat harmonization, respectively. In terms of predicting recurrence in NSCLC, our findings indicate that using ComBat harmonization can significantly enhance machine learning outcomes, particularly improving the accuracy of watershed segmentation, which initially had fewer reliable features than manual contouring. Following the application of ComBat harmonization, the majority of cases saw substantial increase in sensitivity and specificity.

CONCLUSION

Radiomic features are vulnerable to different segmentation methods. ComBat harmonization might be considered a solution to overcome the poor reliability of radiomic features.

Does consensus contour improve robustness and accuracy in 18F-FDG PET radiomic features?

PURPOSE

The purpose of our study is to validate the robustness and accuracy of consensus contour in 2-deoxy-2-[18 F]fluoro-D-glucose (18 F-FDG) PET radiomic features.

METHODS

225 nasopharyngeal carcinoma (NPC) and 13 extended cardio-torso (XCAT) simulated data were enrolled. All segmentation were performed with four segmentation methods under two different initial masks, respectively. Consensus contour (ConSeg) was then developed using the majority vote rule. 107 radiomic features were extracted by Pyradiomics based on segmentation and the intraclass correlation coefficient (ICC) was calculated for each feature between masks or among segmentation, respectively. In XCAT ICC between segmentation and simulated ground truth were also calculated to access the accuracy.

RESULTS

ICC varied with the dataset, segmentation method, initial mask and feature type. ConSeg presented higher ICC for radiomic features in robustness tests and similar ICC in accuracy tests, compared with the average of four segmentation results. Higher ICC were also generally observed in irregular initial masks compared with rectangular masks in both robustness and accuracy tests. Furthermore, 19 features (17.76%) had ICC ≥ 0.75 in both robustness and accuracy tests for any of the segmentation methods or initial masks. The dataset was observed to have a large impact on the correlation relationships between radiomic features, but not the segmentation method or initial mask.

CONCLUSIONS

The consensus contour combined with irregular initial mask could improve the robustness and accuracy in radiomic analysis to some extent. The correlation relationships between radiomic features and feature clusters largely depended on the dataset, but not segmentation method or initial mask.

Comparison of semi-automatic and manual segmentation methods for tumor delineation on head and neck squamous cell carcinoma (HNSCC) positron emission tomography (PET) images

Objective. Accurate and reproducible tumor delineation on positron emission tomography (PET) images is required to validate predictive and prognostic models based on PET radiomic features. Manual segmentation of tumors is time-consuming whereas semi-automatic methods are easily implementable and inexpensive. This study assessed the reliability of semi-automatic segmentation methods over manual segmentation for tumor delineation in head and neck squamous cell carcinoma (HNSCC) PET images.Approach. We employed manual and six semi-automatic segmentation methods (just enough interaction (JEI), watershed, grow from seeds (GfS), flood filling (FF), 30% SUVmax and 40%SUVmax threshold) using 3D slicer software to extract 128 radiomic features from FDG-PET images of 100 HNSCC patients independently by three operators. We assessed the distributional properties of all features and considered 92 log-transformed features for subsequent analysis. For each paired comparison of a feature, we fitted a separate linear mixed effect model using the method (two levels; manual versus one semi-automatic method) as a fixed effect and the subject and the operator as the random effects. We estimated different statistics-the intraclass correlation coefficient agreement (aICC), limits of agreement (LoA), total deviation index (TDI), coverage probability (CP) and coefficient of individual agreement (CIA)-to evaluate the agreement between the manual and semi-automatic methods.Main results. Accounting for all statistics across 92 features, the JEI method consistently demonstrated acceptable agreement with the manual method, with median values of aICC = 0.86, TDI = 0.94, CP = 0.66, and CIA = 0.91.Significance. This study demonstrated that JEI method is a reliable semi-automatic method for tumor delineation on HNSCC PET images.

Semi-supervised learning towards automated segmentation of PET images with limited annotations: application to lymphoma patients

Manual segmentation poses a time-consuming challenge for disease quantification, therapy evaluation, treatment planning, and outcome prediction. Convolutional neural networks (CNNs) hold promise in accurately identifying tumor locations and boundaries in PET scans. However, a major hurdle is the extensive amount of supervised and annotated data necessary for training. To overcome this limitation, this study explores semi-supervised approaches utilizing unlabeled data, specifically focusing on PET images of diffuse large B-cell lymphoma (DLBCL) and primary mediastinal large B-cell lymphoma (PMBCL) obtained from two centers. We considered 2-[18F]FDG PET images of 292 patients PMBCL (n = 104) and DLBCL (n = 188) (n = 232 for training and validation, and n = 60 for external testing). We harnessed classical wisdom embedded in traditional segmentation methods, such as the fuzzy clustering loss function (FCM), to tailor the training strategy for a 3D U-Net model, incorporating both supervised and unsupervised learning approaches. Various supervision levels were explored, including fully supervised methods with labeled FCM and unified focal/Dice loss, unsupervised methods with robust FCM (RFCM) and Mumford-Shah (MS) loss, and semi-supervised methods combining FCM with supervised Dice loss (MS + Dice) or labeled FCM (RFCM + FCM). The unified loss function yielded higher Dice scores (0.73 ± 0.11; 95% CI 0.67-0.8) than Dice loss (p value < 0.01). Among the semi-supervised approaches, RFCM + αFCM (α = 0.3) showed the best performance, with Dice score of 0.68 ± 0.10 (95% CI 0.45-0.77), outperforming MS + αDice for any supervision level (any α) (p < 0.01). Another semi-supervised approach with MS + αDice (α = 0.2) achieved Dice score of 0.59 ± 0.09 (95% CI 0.44-0.76) surpassing other supervision levels (p < 0.01). Given the time-consuming nature of manual delineations and the inconsistencies they may introduce, semi-supervised approaches hold promise for automating medical imaging segmentation workflows.

Improving Generalizability of PET DL Algorithms: List-Mode Reconstructions Improve DOTATATE PET Hepatic Lesion Detection Performance

Deep learning (DL) algorithms used for DOTATATE PET lesion detection typically require large, well-annotated training datasets. These are difficult to obtain due to low incidence of gastroenteropancreatic neuroendocrine tumors (GEP-NETs) and the high cost of manual annotation. Furthermore, networks trained and tested with data acquired from site specific PET/CT instrumentation, acquisition and processing protocols have reduced performance when tested with offsite data. This lack of generalizability requires even larger, more diverse training datasets. The objective of this study is to investigate the feasibility of improving DL algorithm performance by better matching the background noise in training datasets to higher noise, out-of-domain testing datasets. 68Ga-DOTATATE PET/CT datasets were obtained from two scanners: Scanner1, a state-of-the-art digital PET/CT (GE DMI PET/CT; n = 83 subjects), and Scanner2, an older-generation analog PET/CT (GE STE; n = 123 subjects). Set1, the data set from Scanner1, was reconstructed with standard clinical parameters (5 min; Q.Clear) and list-mode reconstructions (VPFXS 2, 3, 4, and 5-min). Set2, data from Scanner2 representing out-of-domain clinical scans, used standard iterative reconstruction (5 min; OSEM). A deep neural network was trained with each dataset: Network1 for Scanner1 and Network2 for Scanner2. DL performance (Network1) was tested with out-of-domain test data (Set2). To evaluate the effect of training sample size, we tested DL model performance using a fraction (25%, 50% and 75%) of Set1 for training. Scanner1, list-mode 2-min reconstructed data demonstrated the most similar noise level compared that of Set2, resulting in the best performance (F1 = 0.713). This was not significantly different compared to the highest performance, upper-bound limit using in-domain training for Network2 (F1 = 0.755; p-value = 0.103). Regarding sample size, the F1 score significantly increased from 25% training data (F1 = 0.478) to 100% training data (F1 = 0.713; p < 0.001). List-mode data from modern PET scanners can be reconstructed to better match the noise properties of older scanners. Using existing data and their associated annotations dramatically reduces the cost and effort in generating these datasets and significantly improves the performance of existing DL algorithms. List-mode reconstructions can provide an efficient, low-cost method to improve DL algorithm generalizability.

Learning Without Real Data Annotations to Detect Hepatic Lesions in PET Images

OBJECTIVE

Deep neural networks have been recently applied to lesion identification in fluorodeoxyglucose (FDG) positron emission tomography (PET) images, but they typically rely on a large amount of well-annotated data for model training. This is extremely difficult to achieve for neuroendocrine tumors (NETs), because of low incidence of NETs and expensive lesion annotation in PET images. The objective of this study is to design a novel, adaptable deep learning method, which uses no real lesion annotations but instead low-cost, list mode-simulated data, for hepatic lesion detection in real-world clinical NET PET images.

METHODS

We first propose a region-guided generative adversarial network (RG-GAN) for lesion-preserved image-to-image translation. Then, we design a specific data augmentation module for our list-mode simulated data and incorporate this module into the RG-GAN to improve model training. Finally, we combine the RG-GAN, the data augmentation module and a lesion detection neural network into a unified framework for joint-task learning to adaptatively identify lesions in real-world PET data.

RESULTS

The proposed method outperforms recent state-of-the-art lesion detection methods in real clinical 68Ga-DOTATATE PET images, and produces very competitive performance with the target model that is trained with real lesion annotations.

CONCLUSION

With RG-GAN modeling and specific data augmentation, we can obtain good lesion detection performance without using any real data annotations.

SIGNIFICANCE

This study introduces an adaptable deep learning method for hepatic lesion identification in NETs, which can significantly reduce human effort for data annotation and improve model generalizability for lesion detection with PET imaging.

The Challenge of Single-Photon Emission Computed Tomography Image Segmentation in the Internal Dosimetry of 177Lu Molecular Therapies

The aim of this article is to review the single photon emission computed tomography (SPECT) segmentation methods used in patient-specific dosimetry of 177Lu molecular therapy. Notably, 177Lu-labelled radiopharmaceuticals are currently used in molecular therapy of metastatic neuroendocrine tumours (ligands for somatostatin receptors) and metastatic prostate adenocarcinomas (PSMA ligands). The proper segmentation of the organs at risk and tumours in targeted radionuclide therapy is an important part of the optimisation process of internal patient dosimetry in this kind of therapy. Because this is the first step in dosimetry assessments, on which further dose calculations are based, it is important to know the level of uncertainty that is associated with this part of the analysis. However, the robust quantification of SPECT images, which would ensure accurate dosimetry assessments, is very hard to achieve due to the intrinsic features of this device. In this article, papers on this topic were collected and reviewed to weigh up the advantages and disadvantages of the segmentation methods used in clinical practice. Degrading factors of SPECT images were also studied to assess their impact on the quantification of 177Lu therapy images. Our review of the recent literature gives an insight into this important topic. However, based on the PubMed and IEEE databases, only a few papers investigating segmentation methods in 177Lumolecular therapy were found. Although segmentation is an important step in internal dose calculations, this subject has been relatively lightly investigated for SPECT systems. This is mostly due to the inner features of SPECT. What is more, even when studies are conducted, they usually utilise the diagnostic radionuclide 99mTc and not a therapeutic one like 177Lu, which could be of concern regarding SPECT camera performance and its overall outcome on dosimetry.

Comparison of Multiple Segmentation Methods for Volumetric Delineation of Primary Prostate Cancer with Prostate-Specific Membrane Antigen-Targeted 18F-DCFPyL PET/CT

This study aimed to assess the accuracy of intraprostatic tumor volume measurements on prostate-specific membrane antigen-targeted 18F-DCFPyL PET/CT made with various segmentation methods. An accurate understanding of tumor volumes versus segmentation techniques is critical for therapy planning, such as radiation dose volume determination and response assessment. Methods: Twenty-five men with clinically localized, high-risk prostate cancer were imaged with 18F-DCFPyL PET/CT before radical prostatectomy. The tumor volumes and tumor-to-prostate ratios (TPRs) of dominant intraprostatic foci of uptake were determined using semiautomatic segmentation (applying SUVmax percentage [SUV%] thresholds of SUV30%-SUV70%), adaptive segmentation (using adaptive segmentation percentage [A%] thresholds of A30%-A70%), and manual contouring. The histopathologic tumor volume (TV-Histo) served as the reference standard. The significance of differences between TV-Histo and PET-based tumor volume were assessed using the paired-sample Wilcoxon signed-rank test. The Spearman correlation coefficient was used to establish the strength of the association between TV-Histo and PET-derived tumor volume. Results: Median TV-Histo was 2.03 cm3 (interquartile ratio [IQR], 1.16-3.36 cm3), and median TPR was 10.16%. The adaptive method with an A40% threshold most closely determined the tumor volume, with a median difference of +0.19 (IQR, -0.71 to +2.01) and a median relative difference of +7.6%. The paired-sample Wilcoxon test showed no significant difference in PET-derived tumor volume and TV-Histo using A40%, A50%, SUV40%, and SUV50% threshold segmentation algorithms (P > 0.05). For both threshold-based segmentation methods, use of higher thresholds (e.g., SUV60% or SUV70% and A50%-A70%) resulted in underestimation of tumor volumes, and use of lower thresholds (e.g., SUV30% or SUV40% and A30%) resulted in overestimation of tumor volumes relative to TV-Histo and TPR. Manual segmentation overestimated the tumor volume, with a median difference of +2.49 (IQR, 0.42-4.11) and a median relative difference of +130%. Conclusion: Segmentation of intraprostatic tumor volume and TPR with an adaptive segmentation approach most closely approximates TV-Histo. This information might be used to guide the primary treatment of men with clinically localized, high-risk prostate cancer.

Information fusion for fully automated segmentation of head and neck tumors from PET and CT images

BACKGROUND

PET/CT images combining anatomic and metabolic data provide complementary information that can improve clinical task performance. PET image segmentation algorithms exploiting the multi-modal information available are still lacking.

PURPOSE

Our study aimed to assess the performance of PET and CT image fusion for gross tumor volume (GTV) segmentations of head and neck cancers (HNCs) utilizing conventional, deep learning (DL), and output-level voting-based fusions.

METHODS

The current study is based on a total of 328 histologically confirmed HNCs from six different centers. The images were automatically cropped to a 200 × 200 head and neck region box, and CT and PET images were normalized for further processing. Eighteen conventional image-level fusions were implemented. In addition, a modified U2-Net architecture as DL fusion model baseline was used. Three different input, layer, and decision-level information fusions were used. Simultaneous truth and performance level estimation (STAPLE) and majority voting to merge different segmentation outputs (from PET and image-level and network-level fusions), that is, output-level information fusion (voting-based fusions) were employed. Different networks were trained in a 2D manner with a batch size of 64. Twenty percent of the dataset with stratification concerning the centers (20% in each center) were used for final result reporting. Different standard segmentation metrics and conventional PET metrics, such as SUV, were calculated.

RESULTS

In single modalities, PET had a reasonable performance with a Dice score of 0.77 ± 0.09, while CT did not perform acceptably and reached a Dice score of only 0.38 ± 0.22. Conventional fusion algorithms obtained a Dice score range of [0.76-0.81] with guided-filter-based context enhancement (GFCE) at the low-end, and anisotropic diffusion and Karhunen-Loeve transform fusion (ADF), multi-resolution singular value decomposition (MSVD), and multi-level image decomposition based on latent low-rank representation (MDLatLRR) at the high-end. All DL fusion models achieved Dice scores of 0.80. Output-level voting-based models outperformed all other models, achieving superior results with a Dice score of 0.84 for Majority_ImgFus, Majority_All, and Majority_Fast. A mean error of almost zero was achieved for all fusions using SUVpeak , SUVmean and SUVmedian .

CONCLUSION

PET/CT information fusion adds significant value to segmentation tasks, considerably outperforming PET-only and CT-only methods. In addition, both conventional image-level and DL fusions achieve competitive results. Meanwhile, output-level voting-based fusion using majority voting of several algorithms results in statistically significant improvements in the segmentation of HNC.

Diagnosis and prognostic predictive value of delineation methods from 18F-FDG PET/CT and PET/MR in pancreatic lesion

The aim was to utilize three segmentation methods on 18F-FDG PET/CT and PET/MR images of pancreatic neoplasm patients, and further compare the effectiveness in differentiating benign from malignant, TNM-stage and prognosis. We conducted a retrospective analysis of 51 patients with pancreatic neoplasm who had undergone 18F-FDG PET/CT and PET/MR before treatment. The patients were categorized into malignant and benign groups. For each patient, the lesion was segmented by 3 thresholds and we recorded TNM-stage, treatment strategy, time to death, and the performance status of survivors. We used receiver operating characteristic (ROC) analysis to compare the diagnostic performance of different threshold delineations between benign and malignant, as well as TNM-stage of adenocarcinoma patients. The optimal model of prognostic value was also assessed by Cox proportional hazards regression analysis and Kaplan-Meier survival analysis. For both PET/CT and PET/MR, SUVmax had the best diagnostic efficacy in identifying malignant tumors. The background method of PET/MR exhibited the outstanding performance in M-stage (sensitivity/specificity, 92.90%/88.20%), with the weighted factor being whole-body total lesion glycolysis (WBTLG). In multivariate analysis, WBTLG (Exp [B] = 1.009; P = 0.009), and surgery (Exp [B] = 15.542; P = 0.008) were independent predictive factors associated with prognosis. This study found that SUVmax from PET/CT had the best diagnostic efficacy in identifying malignancy, while PET/MR showed higher specificity and accuracy for M-stage. The treatment strategy and WBTLG were independent prognostic factors in pancreatic neoplasm patients. PET/MR using the background method was identified as the optimal predictive model for prognosis.

Lesion Dosimetry for [177Lu]Lu-PSMA-617 Radiopharmaceutical Therapy Combined with Stereotactic Body Radiotherapy in Patients with Oligometastatic Castration-Sensitive Prostate Cancer

A single-institution prospective pilot clinical trial was performed to demonstrate the feasibility of combining [177Lu]Lu-PSMA-617 radiopharmaceutical therapy (RPT) with stereotactic body radiotherapy (SBRT) for the treatment of oligometastatic castration-sensitive prostate cancer. Methods: Six patients with 9 prostate-specific membrane antigen (PSMA)-positive oligometastases received 2 cycles of [177Lu]Lu-PSMA-617 RPT followed by SBRT. After the first intravenous infusion of [177Lu]Lu-PSMA-617 (7.46 ± 0.15 GBq), patients underwent SPECT/CT at 3.2 ± 0.5, 23.9 ± 0.4, and 87.4 ± 12.0 h. Voxel-based dosimetry was performed with calibration factors (11.7 counts per second/MBq) and recovery coefficients derived from in-house phantom experiments. Lesions were segmented on baseline PSMA PET/CT (50% SUVmax). After a second cycle of [177Lu]Lu-PSMA-617 (44 ± 3 d; 7.50 ± 0.10 GBq) and an interim PSMA PET/CT scan, SBRT (27 Gy in 3 fractions) was delivered to all PSMA-avid oligometastatic sites, followed by post-PSMA PET/CT. RPT and SBRT voxelwise dose maps were scaled (α/β = 3 Gy; repair half-time, 1.5 h) to calculate the biologically effective dose (BED). Results: All patients completed the combination therapy without complications. No grade 3+ toxicities were noted. The median of the lesion SUVmax as measured on PSMA PET was 16.8 (interquartile range [IQR], 11.6) (baseline), 6.2 (IQR, 2.7) (interim), and 2.9 (IQR, 1.4) (post). PET-derived lesion volumes were 0.4-1.7 cm3 The median lesion-absorbed dose (AD) from the first cycle of [177Lu]Lu-PSMA-617 RPT (ADRPT) was 27.7 Gy (range, 8.3-58.2 Gy; corresponding to 3.7 Gy/GBq, range, 1.1-7.7 Gy/GBq), whereas the median lesion AD from SBRT was 28.1 Gy (range, 26.7-28.8 Gy). Spearman rank correlation, ρ, was 0.90 between the baseline lesion PET SUVmax and SPECT SUVmax (P = 0.005), 0.74 (P = 0.046) between the baseline PET SUVmax and the lesion ADRPT, and -0.81 (P = 0.022) between the lesion ADRPT and the percent change in PET SUVmax (baseline to interim). The median for the lesion BED from RPT and SBRT was 159 Gy (range, 124-219 Gy). ρ between the BED from RPT and SBRT and the percent change in PET SUVmax (baseline to post) was -0.88 (P = 0.007). Two cycles of [177Lu]Lu-PSMA-617 RPT contributed approximately 40% to the maximum BED from RPT and SBRT. Conclusion: Lesional dosimetry in patients with oligometastatic castration-sensitive prostate cancer undergoing [177Lu]Lu-PSMA-617 RPT followed by SBRT is feasible. Combined RPT and SBRT may provide an efficient method to maximize the delivery of meaningful doses to oligometastatic disease while addressing potential microscopic disease reservoirs and limiting the dose exposure to normal tissues.

Whole-body tumor segmentation from PET/CT images using a two-stage cascaded neural network with camouflaged object detection mechanisms

BACKGROUND

Whole-body Metabolic Tumor Volume (MTVwb) is an independent prognostic factor for overall survival in lung cancer patients. Automatic segmentation methods have been proposed for MTV calculation. Nevertheless, most of existing methods for patients with lung cancer only segment tumors in the thoracic region.

PURPOSE

In this paper, we present a Two-Stage cascaded neural network integrated with Camouflaged Object Detection mEchanisms (TS-Code-Net) for automatic segmenting tumors from whole-body PET/CT images.

METHODS

Firstly, tumors are detected from the Maximum Intensity Projection (MIP) images of PET/CT scans, and tumors' approximate localizations along z-axis are identified. Secondly, the segmentations are performed on PET/CT slices that contain tumors identified by the first step. Camouflaged object detection mechanisms are utilized to distinguish the tumors from their surrounding regions that have similar Standard Uptake Values (SUV) and texture appearance. Finally, the TS-Code-Net is trained by minimizing the total loss that incorporates the segmentation accuracy loss and the class imbalance loss.

RESULTS

The performance of the TS-Code-Net is tested on a whole-body PET/CT image data-set including 480 Non-Small Cell Lung Cancer (NSCLC) patients with five-fold cross-validation using image segmentation metrics. Our method achieves 0.70, 0.76, and 0.70, for Dice, Sensitivity and Precision, respectively, which demonstrates the superiority of the TS-Code-Net over several existing methods related to metastatic lung cancer segmentation from whole-body PET/CT images.

CONCLUSIONS

The proposed TS-Code-Net is effective for whole-body tumor segmentation of PET/CT images. Codes for TS-Code-Net are available at: https://github.com/zyj19/TS-Code-Net.

Fusion-based tensor radiomics using reproducible features: Application to survival prediction in head and neck cancer

BACKGROUND

Numerous features are commonly generated in radiomics applications as applied to medical imaging, and identification of robust radiomics features (RFs) can be an important step to derivation of reliable, reproducible solutions. In this work, we utilize a tensor radiomics (TR) framework, where numerous fusions are explored, to generate different flavours of RFs, and we aimed to identify RFs that are robust to fusion techniques in head and neck cancer. Overall, we aimed to predict progression-free survival (PFS) using Hybrid Machine Learning Systems (HMLS) and reproducible RFs.

METHODS

The study was performed on 408 patients with head and neck cancer from The Cancer Imaging Archive. After image preprocessing, 15 fusion techniques were employed to combine Positron Emission Tomography (PET) and Computed Tomography (CT) images. Subsequently, 215 RFs were extracted through a standardized radiomics software, with 17 'flavours' generated using PET-only, CT-only, and 15 fused PET&CT images. The variability of RFs across flavours was studied using the Intraclass Correlation Coefficient (ICC). Furthermore, the features were categorized into seven reliability groups, 106 reproducible RFs with ICC>0.75 were selected, highly correlated flavours were removed, Principal Component Analysis was used to convert 17 flavours to 1 attribute, the polynomial function was utilized to increase RFs, and Analysis of variance (ANOVA) was used to select the relevant attributes. Finally, 3 classifiers including Random Forest (RFC), Logistic regression (LR), and Multi-layer perceptron were applied to the preselected relevant attributes to predict binary PFS. In 5-fold cross-validation, 80% of 4 divisions were utilized to train the model, and the remaining 20% was utilized to evaluate the model. Further, the remaining fold was used for external nested testing.

RESULTS

Reliability analysis indicated that most morphological features belong to the high-reliability category. By contrast, local intensity and statistical features extracted from images belong to the low-reliability category. In the tensor framework, the highest 5-fold cross-validation accuracy of 76.7%±3.3% with an external nested testing of 70.6%±6.7% resulted from the reproducible TR+polynomial function+ANOVA+LR algorithm while the accuracy of 70.0%±4.2% with the external nested testing of 67.7%±4.9% was achieved through the PCA fusion+RFC (non-tensor paradigm).

CONCLUSIONS

This study demonstrated that using reproducible RFs as utilized within a tensor fusion radiomics framework, linked with ANOVA and LR, added value to prediction of progression-free survival outcome in head and neck cancer patients.

From coarse to fine: a deep 3D probability volume contours framework for tumour segmentation and dose painting in PET images

With the increasing integration of functional imaging techniques like Positron Emission Tomography (PET) into radiotherapy (RT) practices, a paradigm shift in cancer treatment methodologies is underway. A fundamental step in RT planning is the accurate segmentation of tumours based on clinical diagnosis. Furthermore, novel tumour control methods, such as intensity modulated radiation therapy (IMRT) dose painting, demand the precise delineation of multiple intensity value contours to ensure optimal tumour dose distribution. Recently, convolutional neural networks (CNNs) have made significant strides in 3D image segmentation tasks, most of which present the output map at a voxel-wise level. However, because of information loss in subsequent downsampling layers, they frequently fail to precisely identify precise object boundaries. Moreover, in the context of dose painting strategies, there is an imperative need for reliable and precise image segmentation techniques to delineate high recurrence-risk contours. To address these challenges, we introduce a 3D coarse-to-fine framework, integrating a CNN with a kernel smoothing-based probability volume contour approach (KsPC). This integrated approach generates contour-based segmentation volumes, mimicking expert-level precision and providing accurate probability contours crucial for optimizing dose painting/IMRT strategies. Our final model, named KsPC-Net, leverages a CNN backbone to automatically learn parameters in the kernel smoothing process, thereby obviating the need for user-supplied tuning parameters. The 3D KsPC-Net exploits the strength of KsPC to simultaneously identify object boundaries and generate corresponding probability volume contours, which can be trained within an end-to-end framework. The proposed model has demonstrated promising performance, surpassing state-of-the-art models when tested against the MICCAI 2021 challenge dataset (HECKTOR).

Application of a smart 18F-FDG-PET adaptive threshold segmentation algorithm for the biological target volume delineation in head and neck cancer

BACKGROUND

The aim of the present study is to evaluate the reliability of a 18F-fluorodeoxyglucose (18F-FDG) PET adaptive threshold segmentation (ATS) algorithm, previously validated in a preclinical setting on several scanners, for the biological target volume (BTV) delineation of head and neck radiotherapy planning.

METHODS

[18F]FDG PET ATS algorithm was studied in treatment plans of head and neck squamous cell carcinoma on a dedicated workstation (iTaRT, Tecnologie Avanzate, Turin, Italy). BTVs segmented by the present ATS algorithm (BTVATS) were compared with those manually segmented for the original radiotherapy treatment planning (BTVVIS). We performed a qualitative and quantitative volumetric analysis with a comparison tool within the ImSimQA TM software package (Oncology Systems Limited, Shrewsbury, UK). We reported figures of merit (FOMs) to convey complementary information: Dice Similarity Coefficient, Sensitivity Index, and Inclusiveness Index.

RESULTS

The study was conducted on 32 treatment plans. Median BTVATS was 11 cm3 while median BTVVIS was 14 cm3. The median Dice Similarity Coefficient, Sensitivity Index, Inclusiveness Index were 0.72, 63%, 88%, respectively. Interestingly, the median volume and the median distance of the voxels that are over contoured by ATS were respectively 1 cm3 and 1 mm.

CONCLUSIONS

ATS algorithm could be a smart and an independent operator tool when implemented for 18F-FDG-PET-based tumor volume delineation. Furthermore, it might be relevant in case of BTV-based dose painting.

Activity recovery for differently shaped objects in quantitative SPECT

Objective.The aim was to theoretically and experimentally investigate recovery in SPECT images with objects of different shapes. Furthermore, the accuracy of volume estimation by thresholding was studied for those shapes.Approach.Nine spheres, nine oblate spheroids, and nine prolate spheroids phantom inserts were used, of which the six smaller spheres were part of the NEMA IEC body phantom and the rest of the inserts were 3D-printed. The inserts were filled with99mTc and177Lu. When filled with99mTc, SPECT images were acquired in a Siemens Symbia Intevo Bold gamma camera and when filled with177Lu in a General Electric NM/CT 870 DR gamma camera. The signal rate per activity (SRPA) was determined for all inserts and represented as a function of the volume-to-surface ratio and of the volume-equivalent radius using VOIs defined according to the sphere dimensions and VOIs defined using thresholding. Experimental values were compared with theoretical curves obtained analytically (spheres) or numerically (spheroids), starting from the convolution of a source distribution with a point-spread function. Validation of the activity estimation strategy was performed using four 3D-printed ellipsoids. Lastly, the threshold values necessary to determine the volume of each insert were obtained.Main results.Results showed that SRPA values for the oblate spheroids diverted from the other inserts, when SRPA were represented as a function of the volume-equivalent radius. However, SRPA values for all inserts followed a similar behaviour when represented as a function of the volume-to-surface ratio. Results for ellipsoids were in agreement with those results. For the three types of inserts the volume could be accurately estimated using a threshold method for volumes larger than 25 ml.Significance.Determination of SRPA independently of lesion or organ shape should decrease uncertainties in estimated activities and thereby, in the long term, be beneficial to patient care.

CNN-based automatic segmentations and radiomics feature reliability on contrast-enhanced ultrasound images for renal tumors

Objective

To investigate the feasibility and efficiency of automatic segmentation of contrast-enhanced ultrasound (CEUS) images in renal tumors by convolutional neural network (CNN) based models and their further application in radiomic analysis.

Materials and methods

From 94 pathologically confirmed renal tumor cases, 3355 CEUS images were extracted and randomly divided into training set (3020 images) and test set (335 images). According to the histological subtypes of renal cell carcinoma, the test set was further split into clear cell renal cell carcinoma (ccRCC) set (225 images), renal angiomyolipoma (AML) set (77 images) and set of other subtypes (33 images). Manual segmentation was the gold standard and serves as ground truth. Seven CNN-based models including DeepLabV3+, UNet, UNet++, UNet3+, SegNet, MultilResUNet and Attention UNet were used for automatic segmentation. Python 3.7.0 and Pyradiomics package 3.0.1 were used for radiomic feature extraction. Performance of all approaches was evaluated by the metrics of mean intersection over union (mIOU), dice similarity coefficient (DSC), precision, and recall. Reliability and reproducibility of radiomics features were evaluated by the Pearson coefficient and the intraclass correlation coefficient (ICC).

Results

All seven CNN-based models achieved good performance with the mIOU, DSC, precision and recall ranging between 81.97%-93.04%, 78.67%-92.70%, 93.92%-97.56%, and 85.29%-95.17%, respectively. The average Pearson coefficients ranged from 0.81 to 0.95, and the average ICCs ranged from 0.77 to 0.92. The UNet++ model showed the best performance with the mIOU, DSC, precision and recall of 93.04%, 92.70%, 97.43% and 95.17%, respectively. For ccRCC, AML and other subtypes, the reliability and reproducibility of radiomic analysis derived from automatically segmented CEUS images were excellent, with the average Pearson coefficients of 0.95, 0.96 and 0.96, and the average ICCs for different subtypes were 0.91, 0.93 and 0.94, respectively.

Conclusion

This retrospective single-center study showed that the CNN-based models had good performance on automatic segmentation of CEUS images for renal tumors, especially the UNet++ model. The radiomics features extracted from automatically segmented CEUS images were feasible and reliable, and further validation by multi-center research is necessary.

Practical and technical key challenges in head and neck adaptive radiotherapy: The GORTEC point of view

Anatomical variations occur during head and neck (H&N) radiotherapy (RT) treatment. These variations may result in underdosage to the target volume or overdosage to the organ at risk. Replanning during the treatment course can be triggered to overcome this issue. Due to technological, methodological and clinical evolutions, tools for adaptive RT (ART) are becoming increasingly sophisticated. The aim of this paper is to give an overview of the key steps of an H&N ART workflow and tools from the point of view of a group of French-speaking medical physicists and physicians (from GORTEC). Focuses are made on image registration, segmentation, estimation of the delivered dose of the day, workflow and quality assurance for an implementation of H&N offline and online ART. Practical recommendations are given to assist physicians and medical physicists in a clinical workflow.

Overview of the HECKTOR Challenge at MICCAI 2022: Automatic Head and Neck Tumor Segmentation and Outcome Prediction in PET/CT

This paper presents an overview of the third edition of the HEad and neCK TumOR segmentation and outcome prediction (HECKTOR) challenge, organized as a satellite event of the 25th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2022. The challenge comprises two tasks related to the automatic analysis of FDG-PET/CT images for patients with Head and Neck cancer (H&N), focusing on the oropharynx region. Task 1 is the fully automatic segmentation of H&N primary Gross Tumor Volume (GTVp) and metastatic lymph nodes (GTVn) from FDG-PET/CT images. Task 2 is the fully automatic prediction of Recurrence-Free Survival (RFS) from the same FDG-PET/CT and clinical data. The data were collected from nine centers for a total of 883 cases consisting of FDG-PET/CT images and clinical information, split into 524 training and 359 test cases. The best methods obtained an aggregated Dice Similarity Coefficient (DSCagg) of 0.788 in Task 1, and a Concordance index (C-index) of 0.682 in Task 2.

Does consensus contours improve robustness and accuracy on [Formula: see text]F-FDG PET imaging tumor delineation?

PURPOSE

The aim of this study is to explore the robustness and accuracy of consensus contours with 225 nasopharyngeal carcinoma (NPC) clinical cases and 13 extended cardio-torso simulated lung tumors (XCAT) based on 2-deoxy-2-[[Formula: see text]F]fluoro-D-glucose ([Formula: see text]F-FDG) PET imaging.

METHODS

Primary tumor segmentation was performed with two different initial masks on 225 NPC [Formula: see text]F-FDG PET datasets and 13 XCAT simulations using methods of automatic segmentation with active contour, affinity propagation (AP), contrast-oriented thresholding (ST), and 41% maximum tumor value (41MAX), respectively. Consensus contours (ConSeg) were subsequently generated based on the majority vote rule. The metabolically active tumor volume (MATV), relative volume error (RE), Dice similarity coefficient (DSC) and their respective test-retest (TRT) metrics between different masks were adopted to analyze the results quantitatively. The nonparametric Friedman and post hoc Wilcoxon tests with Bonferroni adjustment for multiple comparisons were performed with [Formula: see text] 0.05 considered to be significant.

RESULTS

AP presented the highest variability for MATV in different masks, and ConSeg presented much better TRT performances in MATV compared with AP, and slightly poorer TRT in MATV compared with ST or 41MAXin most cases. Similar trends were also found in RE and DSC with the simulated data. The average of four segmentation results (AveSeg) showed better or comparable results in accuracy for most cases with respect to ConSeg. AP, AveSeg and ConSeg presented better RE and DSC in irregular masks as compared with rectangle masks. Additionally, all methods underestimated the tumour boundaries in relation to the ground truth for XCAT including respiratory motion.

CONCLUSIONS

The consensus method could be a robust approach to alleviate segmentation variabilities, but did not seem to improve the accuracy of segmentation results on average. Irregular initial masks might be at least in some cases attributable to mitigate the segmentation variability as well.

Clinical use of positron emission tomography for radiotherapy planning - Medical physics considerations

PET/CT imaging plays an increasing role in radiotherapy treatment planning. The aim of this article was to identify the major use cases and technical as well as medical physics challenges during integration of these data into treatment planning. Dedicated aspects, such as (i) PET/CT-based radiotherapy simulation, (ii) PET-based target volume delineation, (iii) functional avoidance to optimized organ-at-risk sparing and (iv) functionally adapted individualized radiotherapy are discussed in this article. Furthermore, medical physics aspects to be taken into account are summarized and presented in form of check-lists.

Prognostic Role of Biologically Active Volume of Disease in Patients With Metastatic Lung Adenocarcinoma

BACKGROUND

Number of metastatic sites can identify patient populations with non-small cell lung cancer (NSCLC) that benefit from aggressive therapy. Total volume of disease is also relevant. We evaluated the prognostic impact of biologically active volume of disease (BaVD) on patients with metastatic lung adenocarcinoma.

MATERIALS AND METHODS

Positron emission tomography/computerized tomography (PET/CT) scans from patients with newly diagnosed lung adenocarcinoma prior to starting any therapy were identified. SUV thresholds of 3 and 4 were used to auto-contour all FDG avid areas. Kaplan-Meier analysis and Cox regression were performed to examine influence on OS.

RESULTS

One hundred forty-eight patients were included in the analysis. The median BaVD when using an SUV threshold of 3 was 122.8 mL. The median BaVD when using an SUV threshold of 4 was 46.2 mL When stratified by median BaVD using an SUV of 3, median OS was higher for patients with <=122.8 mL (2.12 years) compared to patients with >122.8 mL (1.46 years) (log-rank P = .001). Similarly, when stratified by median BaVD using an SUV of 4, median OS was higher for patients with <=46.2 mL (1.91 years; 95% CI: 1.65-3.22 years) compared to patients with >46.2 mL (1.48 years; 95% CI: 1.07-1.80 years) (log-rank P = .007). On multivariable analysis, BaVD was significantly associated with OS when using an SUV threshold of 3 (HR: 20.169, P < .001) and 4 (HR: 4.117, P < .001).

CONCLUSION

BaVD is an important prognostic factor in metastatic lung adenocarcinoma and may aid identification of patients with limited disease who may be candidates for more aggressive therapies.

DMFF-Net: A dual encoding multiscale feature fusion network for ovarian tumor segmentation

Ovarian cancer is a serious threat to the female reproductive system. Precise segmentation of the tumor area helps the doctors to further diagnose the disease. Automatic segmentation techniques for abstracting high-quality features from images through autonomous learning of model have become a hot research topic nowadays. However, the existing methods still have the problem of poor segmentation of ovarian tumor details. To cope with this problem, a dual encoding based multiscale feature fusion network (DMFF-Net) is proposed for ovarian tumor segmentation. Firstly, a dual encoding method is proposed to extract diverse features. These two encoding paths are composed of residual blocks and single dense aggregation blocks, respectively. Secondly, a multiscale feature fusion block is proposed to generate more advanced features. This block constructs feature fusion between two encoding paths to alleviate the feature loss during deep extraction and further increase the information content of the features. Finally, coordinate attention is added to the decoding stage after the feature concatenation, which enables the decoding stage to capture the valid information accurately. The test results show that the proposed method outperforms existing medical image segmentation algorithms for segmenting lesion details. Moreover, the proposed method also performs well in two other segmentation tasks.

The role of artificial intelligence based on PET/CT radiomics in NSCLC: Disease management, opportunities, and challenges

Objectives

Lung cancer has been widely characterized through radiomics and artificial intelligence (AI). This review aims to summarize the published studies of AI based on positron emission tomography/computed tomography (PET/CT) radiomics in non-small-cell lung cancer (NSCLC).

Materials and methods

A comprehensive search of literature published between 2012 and 2022 was conducted on the PubMed database. There were no language or publication status restrictions on the search. About 127 articles in the search results were screened and gradually excluded according to the exclusion criteria. Finally, this review included 39 articles for analysis.

Results

Classification is conducted according to purposes and several studies were identified at each stage of disease:1) Cancer detection (n=8), 2) histology and stage of cancer (n=11), 3) metastases (n=6), 4) genotype (n=6), 5) treatment outcome and survival (n=8). There is a wide range of heterogeneity among studies due to differences in patient sources, evaluation criteria and workflow of radiomics. On the whole, most models show diagnostic performance comparable to or even better than experts, and the common problems are repeatability and clinical transformability.

Conclusion

AI-based PET/CT Radiomics play potential roles in NSCLC clinical management. However, there is still a long way to go before being translated into clinical application. Large-scale, multi-center, prospective research is the direction of future efforts, while we need to face the risk of repeatability of radiomics features and the limitation of access to large databases.

Joint EANM/SNMMI guideline on radiomics in nuclear medicine : Jointly supported by the EANM Physics Committee and the SNMMI Physics, Instrumentation and Data Sciences Council

PURPOSE

The purpose of this guideline is to provide comprehensive information on best practices for robust radiomics analyses for both hand-crafted and deep learning-based approaches.

METHODS

In a cooperative effort between the EANM and SNMMI, we agreed upon current best practices and recommendations for relevant aspects of radiomics analyses, including study design, quality assurance, data collection, impact of acquisition and reconstruction, detection and segmentation, feature standardization and implementation, as well as appropriate modelling schemes, model evaluation, and interpretation. We also offer an outlook for future perspectives.

CONCLUSION

Radiomics is a very quickly evolving field of research. The present guideline focused on established findings as well as recommendations based on the state of the art. Though this guideline recognizes both hand-crafted and deep learning-based radiomics approaches, it primarily focuses on the former as this field is more mature. This guideline will be updated once more studies and results have contributed to improved consensus regarding the application of deep learning methods for radiomics. Although methodological recommendations in the present document are valid for most medical image modalities, we focus here on nuclear medicine, and specific recommendations when necessary are made for PET/CT, PET/MR, and quantitative SPECT.

Intra- and inter-operator variability in MRI-based manual segmentation of HCC lesions and its impact on dosimetry

PURPOSE

The aim was to quantify inter- and intra-observer variability in manually delineated hepatocellular carcinoma (HCC) lesion contours and the resulting impact on radioembolization (RE) dosimetry.

METHODS

Ten patients with HCC lesions treated with Y-90 RE and imaged with post-therapy Y-90 PET/CT were selected for retrospective analysis. Three radiologists contoured 20 lesions manually on baseline multiphase contrast-enhanced MRIs, and two of the radiologists re-contoured at two additional sessions. Contours were transferred to co-registered PET/CT-based Y-90 dose maps. Volume-dependent recovery coefficients were applied for partial volume correction (PVC) when reporting mean absorbed dose. To understand how uncertainty varies with tumor size, we fit power models regressing relative uncertainty in volume and in mean absorbed dose on contour volume. Finally, we determined effects of segmentation uncertainty on tumor control probability (TCP), as calculated using logistic models developed in a previous RE study.

RESULTS

The average lesion volume ranged from 1.8 to 194.5 mL, and the mean absorbed dose ranged from 23.4 to 1629.0 Gy. The mean inter-observer Dice coefficient for lesion contours was significantly less than the mean intra-observer Dice coefficient (0.79 vs. 0.85, p < 0.001). Uncertainty in segmented volume, as measured by the Coefficient of Variation (CV), ranged from 4.2 to 34.7% with an average of 17.2%. The CV in mean absorbed dose had an average value of 5.4% (range 1.2-13.1%) without PVC while it was 15.1% (range 1.5-55.2%) with PVC. Using the fitted models for uncertainty as a function of volume on our prior data, the mean change in TCP due to segmentation uncertainty alone was estimated as 16.2% (maximum 48.5%).

CONCLUSIONS

Though we find relatively high inter- and intra-observer reliability overall, uncertainty in tumor contouring propagates into non-negligible uncertainty in dose metrics and outcome prediction for individual cases that should be considered in dosimetry-guided treatment.

Cross modality fusion for modality-specific lung tumor segmentation in PET-CT images

Although positron emission tomography-computed tomography (PET-CT) images have been widely used, it is still challenging to accurately segment the lung tumor. The respiration, movement and imaging modality lead to large modality discrepancy of the lung tumors between PET images and CT images. To overcome these difficulties, a novel network is designed to simultaneously obtain the corresponding lung tumors of PET images and CT images. The proposed network can fuse the complementary information and preserve modality-specific features of PET images and CT images. Due to the complementarity between PET images and CT images, the two modality images should be fused for automatic lung tumor segmentation. Therefore, cross modality decoding blocks are designed to extract modality-specific features of PET images and CT images with the constraints of the other modality. The edge consistency loss is also designed to solve the problem of blurred boundaries of PET images and CT images. The proposed method is tested on 126 PET-CT images with non-small cell lung cancer, and Dice similarity coefficient scores of lung tumor segmentation reach 75.66 ± 19.42 in CT images and 79.85 ± 16.76 in PET images, respectively. Extensive comparisons with state-of-the-art lung tumor segmentation methods have also been performed to demonstrate the superiority of the proposed network.

Standardised lesion segmentation for imaging biomarker quantitation: a consensus recommendation from ESR and EORTC

BACKGROUND

Lesion/tissue segmentation on digital medical images enables biomarker extraction, image-guided therapy delivery, treatment response measurement, and training/validation for developing artificial intelligence algorithms and workflows. To ensure data reproducibility, criteria for standardised segmentation are critical but currently unavailable.

METHODS

A modified Delphi process initiated by the European Imaging Biomarker Alliance (EIBALL) of the European Society of Radiology (ESR) and the European Organisation for Research and Treatment of Cancer (EORTC) Imaging Group was undertaken. Three multidisciplinary task forces addressed modality and image acquisition, segmentation methodology itself, and standards and logistics. Devised survey questions were fed via a facilitator to expert participants. The 58 respondents to Round 1 were invited to participate in Rounds 2-4. Subsequent rounds were informed by responses of previous rounds.

RESULTS/CONCLUSIONS

Items with ≥ 75% consensus are considered a recommendation. These include system performance certification, thresholds for image signal-to-noise, contrast-to-noise and tumour-to-background ratios, spatial resolution, and artefact levels. Direct, iterative, and machine or deep learning reconstruction methods, use of a mixture of CE marked and verified research tools were agreed and use of specified reference standards and validation processes considered essential. Operator training and refreshment were considered mandatory for clinical trials and clinical research. Items with a 60-74% agreement require reporting (site-specific accreditation for clinical research, minimal pixel number within lesion segmented, use of post-reconstruction algorithms, operator training refreshment for clinical practice). Items with ≤ 60% agreement are outside current recommendations for segmentation (frequency of system performance tests, use of only CE-marked tools, board certification of operators, frequency of operator refresher training). Recommendations by anatomical area are also specified.

Functional lung imaging in thoracic tumor radiotherapy: Application and progress

Radiotherapy plays an irreplaceable and unique role in treating thoracic tumors, but the occurrence of radiation-induced lung injury has limited the increase in tumor target doses and has influenced patients’ quality of life. However, the introduction of functional lung imaging has been incorporating functional lungs into radiotherapy planning. The design of the functional lung protection plan, while meeting the target dose requirements and dose limitations of the organs at risk (OARs), minimizes the radiation dose to the functional lung, thus reducing the occurrence of radiation-induced lung injury. In this manuscript, we mainly reviewed the lung ventilation or/and perfusion functional imaging modalities, application, and progress, as well as the results based on the functional lung protection planning in thoracic tumors. In addition, we also discussed the problems that should be explored and further studied in the practical application based on functional lung radiotherapy planning.

18F-FDG PET/CT-Based Prognostic Survival Model After Surgery for Head and Neck Cancer

The aims of this multicenter study were to identify clinical and preoperative PET/CT parameters predicting overall survival (OS) and distant metastasis-free survival (DMFS) in a cohort of head and neck squamous cell carcinoma patients treated with surgery, to generate a prognostic model of OS and DMFS, and to validate this prognostic model with an independent cohort. Methods: A total of 382 consecutive patients with head and neck squamous cell carcinoma, divided into training (n = 318) and validation (n = 64) cohorts, were retrospectively included. The following PET/CT parameters were analyzed: clinical parameters, SUV_max, SUV_mean, metabolic tumor volume (MTV), total lesion glycolysis, and distance parameters for the primary tumor and lymph nodes defined by 2 segmentation methods (relative SUV_max threshold and absolute SUV threshold). Cox analyses were performed for OS and DMFS in the training cohort. The concordance index (c-index) was used to identify highly prognostic parameters. These prognostic parameters were externally tested in the validation cohort. Results: In multivariable analysis, the significant parameters for OS were T stage and nodal MTV, with a c-index of 0.64 (P < 0.001). For DMFS, the significant parameters were T stage, nodal MTV, and maximal tumor-node distance, with a c-index of 0.76 (P < 0.001). These combinations of parameters were externally validated, with c-indices of 0.63 (P < 0.001) and 0.71 (P < 0.001) for OS and DMFS, respectively. Conclusion: The nodal MTV associated with the maximal tumor-node distance was significantly correlated with the risk of DMFS. Moreover, this parameter, in addition to clinical parameters, was associated with a higher risk of death. These prognostic factors may be used to tailor individualized treatment.

Decentralized Distributed Multi-institutional PET Image Segmentation Using a Federated Deep Learning Framework

PURPOSE

The generalizability and trustworthiness of deep learning (DL)-based algorithms depend on the size and heterogeneity of training datasets. However, because of patient privacy concerns and ethical and legal issues, sharing medical images between different centers is restricted. Our objective is to build a federated DL-based framework for PET image segmentation utilizing a multicentric dataset and to compare its performance with the centralized DL approach.

METHODS

PET images from 405 head and neck cancer patients from 9 different centers formed the basis of this study. All tumors were segmented manually. PET images converted to SUV maps were resampled to isotropic voxels (3 × 3 × 3 mm3) and then normalized. PET image subvolumes (12 × 12 × 12 cm3) consisting of whole tumors and background were analyzed. Data from each center were divided into train/validation (80% of patients) and test sets (20% of patients). The modified R2U-Net was used as core DL model. A parallel federated DL model was developed and compared with the centralized approach where the data sets are pooled to one server. Segmentation metrics, including Dice similarity and Jaccard coefficients, percent relative errors (RE%) of SUVpeak, SUVmean, SUVmedian, SUVmax, metabolic tumor volume, and total lesion glycolysis were computed and compared with manual delineations.

RESULTS

The performance of the centralized versus federated DL methods was nearly identical for segmentation metrics: Dice (0.84 ± 0.06 vs 0.84 ± 0.05) and Jaccard (0.73 ± 0.08 vs 0.73 ± 0.07). For quantitative PET parameters, we obtained comparable RE% for SUVmean (6.43% ± 4.72% vs 6.61% ± 5.42%), metabolic tumor volume (12.2% ± 16.2% vs 12.1% ± 15.89%), and total lesion glycolysis (6.93% ± 9.6% vs 7.07% ± 9.85%) and negligible RE% for SUVmax and SUVpeak. No significant differences in performance (P > 0.05) between the 2 frameworks (centralized vs federated) were observed.

CONCLUSION

The developed federated DL model achieved comparable quantitative performance with respect to the centralized DL model. Federated DL models could provide robust and generalizable segmentation, while addressing patient privacy and legal and ethical issues in clinical data sharing.

Impact of acquisition time and misregistration with CT on data-driven gated PET

Objective. Data-driven gating (DDG) can address patient motion issues and enhance PET quantification but suffers from increased image noise from utilization of <100% of PET data. Misregistration between DDG-PET and CT may also occur, altering the potential benefits of gating. Here, the effects of PET acquisition time and CT misregistration were assessed with a combined DDG-PET/DDG-CT technique.Approach. In the primary PET bed with lesions of interest and likely respiratory motion effects, PET acquisition time was extended to 12 min and a low-dose cine CT was acquired to enable DDG-CT. Retrospective reconstructions were created for both non-gated (NG) and DDG-PET using 30 s to 12 min of PET data. Both the standard helical CT and DDG-CT were used for attenuation correction of DDG-PET data. SUVmax, SUVpeak, and CNR were compared for 45 lesions in the liver and lung from 27 cases.Main results. For both NG-PET (p= 0.0041) and DDG-PET (p= 0.0028), only the 30 s acquisition time showed clear SUVmaxbias relative to the 3 min clinical standard. SUVpeakshowed no bias at any change in acquisition time. DDG-PET alone increased SUVmaxby 15 ± 20% (p< 0.0001), then was increased further by an additional 15 ± 29% (p= 0.0007) with DDG-PET/CT. Both 3 min and 6 min DDG-PET had lesion CNR statistically equivalent to 3 min NG-PET, but then increased at 12 min by 28 ± 48% (p= 0.0022). DDG-PET/CT at 6 min had comparable counts to 3 min NG-PET, but significantly increased CNR by 39 ± 46% (p< 0.0001).Significance. 50% counts DDG-PET did not lead to inaccurate or biased SUV-increased SUV resulted from gating. Improved registration from DDG-CT was equally as important as motion correction with DDG-PET for increasing SUV in DDG-PET/CT. Lesion detectability could be significantly improved when DDG-PET used equivalent counts to NG-PET, but only when combined with DDG-CT in DDG-PET/CT.

Nuclear Medicine and Artificial Intelligence: Best Practices for Algorithm Development

The nuclear medicine field has seen a rapid expansion of academic and commercial interest in developing artificial intelligence (AI) algorithms. Users and developers can avoid some of the pitfalls of AI by recognizing and following best practices in AI algorithm development. In this article, recommendations on technical best practices for developing AI algorithms in nuclear medicine are provided, beginning with general recommendations and then continuing with descriptions of how one might practice these principles for specific topics within nuclear medicine. This report was produced by the AI Task Force of the Society of Nuclear Medicine and Molecular Imaging.

The Accuracy and Radiomics Feature Effects of Multiple U-net-Based Automatic Segmentation Models for Transvaginal Ultrasound Images of Cervical Cancer

Ultrasound (US) imaging has been recognized and widely used as a screening and diagnostic imaging modality for cervical cancer all over the world. However, few studies have investigated the U-net-based automatic segmentation models for cervical cancer on US images and investigated the effects of automatic segmentations on radiomics features. A total of 1102 transvaginal US images from 796 cervical cancer patients were collected and randomly divided into training (800), validation (100) and test sets (202), respectively, in this study. Four U-net models (U-net, U-net with ResNet, context encoder network (CE-net), and Attention U-net) were adapted to segment the target of cervical cancer automatically on these US images. Radiomics features were extracted and evaluated from both manually and automatically segmented area. The mean Dice similarity coefficient (DSC) of U-net, Attention U-net, CE-net, and U-net with ResNet were 0.88, 0.89, 0.88, and 0.90, respectively. The average Pearson coefficients for the evaluation of the reliability of US image-based radiomics were 0.94, 0.96, 0.94, and 0.95 for U-net, U-net with ResNet, Attention U-net, and CE-net, respectively, in their comparison with manual segmentation. The reproducibility of the radiomics parameters evaluated by intraclass correlation coefficients (ICC) showed robustness of automatic segmentation with an average ICC coefficient of 0.99. In conclusion, high accuracy of U-net-based automatic segmentations was achieved in delineating the target area of cervical cancer US images. It is feasible and reliable for further radiomics studies with features extracted from automatic segmented target areas.

Comparison of Volumetric Quantitative PET Parameters Before and After a CT-Based Elastic Deformation on Dual-Time 18FDG-PET/CT Images: A Feasibility Study in a Perspective of Radiotherapy Planning in Head and Neck Cancer

Background

The use of 18FDG-PET/CT for delineating a gross tumor volume (GTV, also called MTV metabolic tumor volume) in radiotherapy (RT) planning of head neck squamous cell carcinomas (HNSCC) is not included in current recommendations, although its interest for the radiotherapist is of evidence. Because pre-RT PET scans are rarely done simultaneously with dosimetry CT, the validation of a robust image registration tool and of a reproducible MTV delineation method is still required.

Objective

Our objective was to study a CT-based elastic registration method on dual-time pre-RT 18FDG-PET/CT images to assess the feasibility of PET-based RT planning in patients with HNSCC.

Methods

Dual-time 18FDG-PET/CT [whole-body examination (wbPET) + 1 dedicated step (headPET)] were selected to simulate a 2-times scenario of pre-RT PET images deformation on dosimetry CT. ER-headPET and RR-headPET images were, respectively, reconstructed after CT-to-CT rigid (RR) and elastic (ER) registrations of the headPET on the wbPET. The MTVs delineation was performed using two methods (40%SUVmax, PET-Edge). The percentage variations of several PET parameters (SUVmax, SUVmean, SUVpeak, MTV, TLG) were calculated between wbPET, ER-headPET, and RR-headPET. Correlation between MTV values was calculated (Deming linear regression). MTVs intersections were assessed by two indices (OF, DICE) and compared together (Wilcoxon test). Additional per-volume analysis was evaluated (Mann-Whitney test). Inter- and intra-observer reproducibilities were evaluated (ICC = intra-class coefficient).

Results

36 patients (30M/6F; median age = 65 y) were retrospectively included. The changes in SUVmax, SUVmean and SUVpeak values between ER-headPET and RR-headPET images were <5%. The variations in MTV values between ER-headPET and wbPET images were −6 and −3% with 40%SUVmax and PET Edge, respectively. Their correlations were excellent whatever the delineation method (R² > 0.99). The ER-headPET MTVs had significant higher mean OF and DICE with the wbPET MTVs, for both delineation methods (p ≤ 0.002); and also when lesions had a volume > 5cc (excellent OF = 0.80 with 40%SUVmax). The inter- and intra-observer reproducibilities for MTV delineation were excellent (ICC ≥ 0.8, close to 1 with PET-Edge).

Conclusion

Our study demonstrated no significant changes in MTV after an elastic deformation of pre-RT 18FDG-PET/CT images acquired in dual-time mode. This opens possibilities for HNSCC radiotherapy planning improvement by transferring GTV-PET on dosimetry CT.

A novel semiautomated method for background activity and biological tumour volume definition to improve standardisation of 18F-FET PET imaging in glioblastoma

Background

Multicentre clinical trials evaluating the role of ¹⁸F-Fluoroethyl-l-tyrosine (¹⁸F-FET) PET as a diagnostic biomarker in glioma management have highlighted a need for standardised methods of data analysis. ¹⁸F-FET uptake normalised against background in the contralateral brain is a standard imaging technique to delineate the biological tumour volume (BTV). Quantitative analysis of ¹⁸F-FET PET images requires a consistent and robust background activity. Currently, defining background activity involves the manual selection of an arbitrary region of interest, a process that is subject to large variability. This study aims to eliminate methodological errors in background activity definition through the introduction of a semiautomated method for region of interest selection. A new method for background activity definition, involving the semiautomated generation of mirror-image (MI) reference regions, was compared with the current state-of-the-art method, involving manually drawing crescent-shape (gCS) reference regions. The MI and gCS methods were tested by measuring values of background activity and resulting BTV of ¹⁸F-FET PET scans of ten patients with recurrent glioblastoma multiforme generated from inputs provided by seven readers. To assess intra-reader variability, each scan was evaluated six times by each reader. Intra- and inter-reader variability in background activity and BTV definition was assessed by means of coefficient of variation.

Results

Compared to the gCS method, the MI method showed significantly lower intra- and inter-reader variability both in background activity and in BTV definition.

Conclusions

The proposed semiautomated MI method minimises intra- and inter-reader variability, providing a valuable approach for standardisation of ¹⁸F-FET PET quantitative parameters.

Trial registration ANZCTR, ACTRN12618001346268. Registered 9 August 2018, https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=374253

Supplementary Information

The online version contains supplementary material available at 10.1186/s40658-022-00438-2.

Joint EANM/SNMMI/ESTRO practice recommendations for the use of 2-[18F]FDG PET/CT external beam radiation treatment planning in lung cancer V1.0

Purpose

2-[¹⁸F]FDG PET/CT is of utmost importance for radiation treatment (RT) planning and response monitoring in lung cancer patients, in both non-small and small cell lung cancer (NSCLC and SCLC). This topic has been addressed in guidelines composed by experts within the field of radiation oncology. However, up to present, there is no procedural guideline on this subject, with involvement of the nuclear medicine societies.

Methods

A literature review was performed, followed by a discussion between a multidisciplinary team of experts in the different fields involved in the RT planning of lung cancer, in order to guide clinical management. The project was led by experts of the two nuclear medicine societies (EANM and SNMMI) and radiation oncology (ESTRO).

Results and conclusion

This guideline results from a joint and dynamic collaboration between the relevant disciplines for this topic. It provides a worldwide, state of the art, and multidisciplinary guide to 2-[¹⁸F]FDG PET/CT RT planning in NSCLC and SCLC. These practical recommendations describe applicable updates for existing clinical practices, highlight potential flaws, and provide solutions to overcome these as well. Finally, the recent developments considered for future application are also reviewed.

Perspective paper about the joint EANM/SNMMI/ESTRO practice recommendations for the use of 2-[18F]FDG-PET/CT external beam radiation treatment planning in lung cancer

In “Joint EANM/SNMMI/ESTRO Practice Recommendations for the Use of 2-[18F]FDG-PET/CT External Beam Radiation Treatment Planning in Lung Cancer V1.0” clinical indications for PET-CT in (non-)small cell lung cancer are highlighted and selective nodal irradiation is discussed. Additionally, concepts about target definition, target delineation and treatment evaluation are reviewed.

Head and neck tumor segmentation in PET/CT: The HECKTOR challenge

This paper relates the post-analysis of the first edition of the HEad and neCK TumOR (HECKTOR) challenge. This challenge was held as a satellite event of the 23rd International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2020, and was the first of its kind focusing on lesion segmentation in combined FDG-PET and CT image modalities. The challenge's task is the automatic segmentation of the Gross Tumor Volume (GTV) of Head and Neck (H&N) oropharyngeal primary tumors in FDG-PET/CT images. To this end, the participants were given a training set of 201 cases from four different centers and their methods were tested on a held-out set of 53 cases from a fifth center. The methods were ranked according to the Dice Score Coefficient (DSC) averaged across all test cases. An additional inter-observer agreement study was organized to assess the difficulty of the task from a human perspective. 64 teams registered to the challenge, among which 10 provided a paper detailing their approach. The best method obtained an average DSC of 0.7591, showing a large improvement over our proposed baseline method and the inter-observer agreement, associated with DSCs of 0.6610 and 0.61, respectively. The automatic methods proved to successfully leverage the wealth of metabolic and structural properties of combined PET and CT modalities, significantly outperforming human inter-observer agreement level, semi-automatic thresholding based on PET images as well as other single modality-based methods. This promising performance is one step forward towards large-scale radiomics studies in H&N cancer, obviating the need for error-prone and time-consuming manual delineation of GTVs.

Management and work-up procedures of patients with head and neck malignancies treated by radiation

Radiotherapy alone or in association with systemic treatment plays a major role in the treatment of head and neck tumours, either as a primary treatment or as a postoperative modality. The management of these tumours is multidisciplinary, requiring particular care at every treatment step. We present the update of the recommendations of the French Society of Radiation Oncology on the radiotherapy of head and neck tumours from the imaging work-up needed for optimal selection of treatment volume, to optimization of the dose distribution and delivery.

Fully Automated Gross Tumor Volume Delineation From PET in Head and Neck Cancer Using Deep Learning Algorithms

PURPOSE

The availability of automated, accurate, and robust gross tumor volume (GTV) segmentation algorithms is critical for the management of head and neck cancer (HNC) patients. In this work, we evaluated 3 state-of-the-art deep learning algorithms combined with 8 different loss functions for PET image segmentation using a comprehensive training set and evaluated its performance on an external validation set of HNC patients.

PATIENTS AND METHODS

18F-FDG PET/CT images of 470 patients presenting with HNC on which manually defined GTVs serving as standard of reference were used for training (340 patients), evaluation (30 patients), and testing (100 patients from different centers) of these algorithms. PET image intensity was converted to SUVs and normalized in the range (0-1) using the SUVmax of the whole data set. PET images were cropped to 12 × 12 × 12 cm3 subvolumes using isotropic voxel spacing of 3 × 3 × 3 mm3 containing the whole tumor and neighboring background including lymph nodes. We used different approaches for data augmentation, including rotation (-15 degrees, +15 degrees), scaling (-20%, 20%), random flipping (3 axes), and elastic deformation (sigma = 1 and proportion to deform = 0.7) to increase the number of training sets. Three state-of-the-art networks, including Dense-VNet, NN-UNet, and Res-Net, with 8 different loss functions, including Dice, generalized Wasserstein Dice loss, Dice plus XEnt loss, generalized Dice loss, cross-entropy, sensitivity-specificity, and Tversky, were used. Overall, 28 different networks were built. Standard image segmentation metrics, including Dice similarity, image-derived PET metrics, first-order, and shape radiomic features, were used for performance assessment of these algorithms.

RESULTS

The best results in terms of Dice coefficient (mean ± SD) were achieved by cross-entropy for Res-Net (0.86 ± 0.05; 95% confidence interval [CI], 0.85-0.87), Dense-VNet (0.85 ± 0.058; 95% CI, 0.84-0.86), and Dice plus XEnt for NN-UNet (0.87 ± 0.05; 95% CI, 0.86-0.88). The difference between the 3 networks was not statistically significant (P > 0.05). The percent relative error (RE%) of SUVmax quantification was less than 5% in networks with a Dice coefficient more than 0.84, whereas a lower RE% (0.41%) was achieved by Res-Net with cross-entropy loss. For maximum 3-dimensional diameter and sphericity shape features, all networks achieved a RE ≤ 5% and ≤10%, respectively, reflecting a small variability.

CONCLUSIONS

Deep learning algorithms exhibited promising performance for automated GTV delineation on HNC PET images. Different loss functions performed competitively when using different networks and cross-entropy for Res-Net, Dense-VNet, and Dice plus XEnt for NN-UNet emerged as reliable networks for GTV delineation. Caution should be exercised for clinical deployment owing to the occurrence of outliers in deep learning-based algorithms.

Simultaneous Mapping of Vasculature, Hypoxia, and Proliferation Using Dynamic Susceptibility Contrast MRI, 18F-FMISO PET, and 18F-FLT PET in Relation to Contrast Enhancement in Newly Diagnosed Glioblastoma

Conventional MRI plays a key role in the management of patients with high-grade glioma, but multiparametric MRI and PET tracers could provide further information to better characterize tumor metabolism and heterogeneity by identifying regions having a high risk of recurrence. In this study, we focused on proliferation, hypervascularization, and hypoxia, all factors considered indicative of poor prognosis. They were assessed by measuring uptake of 18F-3'-deoxy-3'-18F-fluorothymidine (18F-FLT), relative cerebral blood volume (rCBV) maps, and uptake of 18F-fluoromisonidazole (18F-FMISO), respectively. For each modality, the volumes and high-uptake subvolumes (hot spots) were semiautomatically segmented and compared with the contrast enhancement (CE) volume on T1-weighted gadolinium-enhanced (T1w-Gd) images, commonly used in the management of patients with glioblastoma. Methods: Dynamic susceptibility contrast-enhanced MRI (31 patients), 18F-FLT PET (20 patients), or 18F-FMISO PET (20 patients), for a total of 31 patients, was performed on preoperative glioblastoma patients. Volumes and hot spots were segmented on SUV maps for 18F-FLT PET (using the fuzzy locally adaptive bayesian algorithm) and 18F-FMISO PET (using a mean contralateral image + 3.3 SDs) and on rCBV maps (using a mean contralateral image + 1.96 SDs) for dynamic susceptibility contrast-enhanced MRI and overlaid on T1w-Gd images. For each modality, the percentages of the peripheral volumes and the peripheral hot spots outside the CE volume were calculated. Results: All tumors showed highly proliferated, hypervascularized, and hypoxic regions. The images also showed pronounced heterogeneity of both tracers regarding their uptake and rCBV maps, within each individual patient. Overlaid volumes on T1w-Gd images showed that some proliferative, hypervascularized, and hypoxic regions extended beyond the CE volume but with marked differences between patients. The ranges of peripheral volume outside the CE volume were 1.6%-155.5%, 1.5%-89.5%, and 3.1%-78.0% for 18F-FLT, rCBV, and 18F-FMISO, respectively. All patients had hyperproliferative hot spots outside the CE volume, whereas hypervascularized and hypoxic hot spots were detected mainly within the enhancing region. Conclusion: Spatial analysis of multiparametric maps with segmented volumes and hot spots provides valuable information to optimize the management and treatment of patients with glioblastoma.

Radiomics in PET Imaging:: A Practical Guide for Newcomers

Radiomics has undergone considerable development in recent years. In PET imaging, very promising results concerning the ability of handcrafted features to predict the biological characteristics of lesions and to assess patient prognosis or response to treatment have been reported in the literature. This article presents a checklist for designing a reliable radiomic study, gives an overview of the steps of the pipeline, and outlines approaches for data harmonization. Tips are provided for critical reading of the content of articles. The advantages and limitations of handcrafted radiomics compared with deep-learning approaches for the characterization of PET images are also discussed.

Toward High-Throughput Artificial Intelligence-Based Segmentation in Oncological PET Imaging

Artificial intelligence (AI) techniques for image-based segmentation have garnered much attention in recent years. Convolutional neural networks have shown impressive results and potential toward fully automated segmentation in medical imaging, and particularly PET imaging. To cope with the limited access to annotated data needed in supervised AI methods, given tedious and prone-to-error manual delineations, semi-supervised and unsupervised AI techniques have also been explored for segmentation of tumors or normal organs in single- and bimodality scans. This work reviews existing AI techniques for segmentation tasks and the evaluation criteria for translational AI-based segmentation efforts toward routine adoption in clinical workflows.

Artificial intelligence in radiation oncology: A review of its current status and potential application for the radiotherapy workforce

OBJECTIVE

Radiation oncology is a continually evolving speciality. With the development of new imaging modalities and advanced imaging processing techniques, there is an increasing amount of data available to practitioners. In this narrative review, Artificial Intelligence (AI) is used as a reference to machine learning, and its potential, along with current problems in the field of radiation oncology, are considered from a technical position.

KEY FINDINGS

AI has the potential to harness the availability of data for improving patient outcomes, reducing toxicity, and easing clinical burdens. However, problems including the requirement of complexity of data, undefined core outcomes and limited generalisability are apparent.

CONCLUSION

This original review highlights considerations for the radiotherapy workforce, particularly therapeutic radiographers, as there will be an increasing requirement for their familiarity with AI due to their unique position as the interface between imaging technology and patients.

IMPLICATIONS FOR PRACTICE

Collaboration between AI experts and the radiotherapy workforce are required to overcome current issues before clinical adoption. The development of educational resources and standardised reporting of AI studies may help facilitate this.

Data-driven gated PET/CT: implications for lesion segmentation and quantitation

Background

Data-driven gating (DDG) can improve PET quantitation and alleviate many issues with patient motion. However, misregistration between DDG-PET and CT may occur due to the distinct temporal resolutions of PET and CT and can be mitigated by DDG-CT. Here, the effects of misregistration and respiratory motion on PET quantitation and lesion segmentation were assessed with a new DDG-PET/CT method.

Methods

A low-dose cine-CT was acquired in misregistered regions to enable both average CT (ACT) and DDG-CT. The following were compared: (1) baseline PET/CT, (2) PET/ACT (attenuation correction, AC = ACT), (3) DDG-PET (AC = helical CT), and (4) DDG-PET/CT (AC = DDG-CT). For DDG-PET, end-expiration (EE) data were derived from 50% of the total PET data at 30% from end-inspiration. For DDG-CT, EE phase CT data were extracted from cine-CT data by lung Hounsfield unit (HU) value and body contour. A total of 91 lesions from 16 consecutive patients were assessed for changes in standard uptake value (SUV), lesion glycolysis (LG), lesion volume, centroid-to-centroid distance (CCD), and DICE coefficients.

Results

Relative to baseline PET/CT, median changes in SUV_max ± σ for all 91 lesions were 20 ± 43%, 26 ± 23%, and 66 ± 66%, respectively, for PET/ACT, DDG-PET, and DDG-PET/CT. Median changes in lesion volume were 0 ± 58%, − 36 ± 26%, and − 26 ± 40%. LG for individual lesions increased for PET/ACT and decreased for DDG-PET, but was not different for DDG-PET/CT. Changes in mean HU from baseline PET/CT were dramatic for most lesions in both PET/ACT and DDG-PET/CT, especially for lesions with mean HU < 0 at baseline. CCD and DICE were both affected more by motion correction with DDG-PET than improved registration with ACT or DDG-CT.

Conclusion

As misregistration becomes more prominent, the impact of motion correction with DDG-PET is diminished. The potential benefits of DDG-PET toward accurate lesion segmentation and quantitation could only be fully realized when combined with DDG-CT. These results impress upon the necessity of ensuring both misregistration and motion correction are accounted for together to optimize the clinical utility of PET/CT.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40658-021-00411-5.

Incorporating radiomics into clinical trials: expert consensus endorsed by the European Society of Radiology on considerations for data-driven compared to biologically driven quantitative biomarkers

Existing quantitative imaging biomarkers (QIBs) are associated with known biological tissue characteristics and follow a well-understood path of technical, biological and clinical validation before incorporation into clinical trials. In radiomics, novel data-driven processes extract numerous visually imperceptible statistical features from the imaging data with no a priori assumptions on their correlation with biological processes. The selection of relevant features (radiomic signature) and incorporation into clinical trials therefore requires additional considerations to ensure meaningful imaging endpoints. Also, the number of radiomic features tested means that power calculations would result in sample sizes impossible to achieve within clinical trials. This article examines how the process of standardising and validating data-driven imaging biomarkers differs from those based on biological associations. Radiomic signatures are best developed initially on datasets that represent diversity of acquisition protocols as well as diversity of disease and of normal findings, rather than within clinical trials with standardised and optimised protocols as this would risk the selection of radiomic features being linked to the imaging process rather than the pathology. Normalisation through discretisation and feature harmonisation are essential pre-processing steps. Biological correlation may be performed after the technical and clinical validity of a radiomic signature is established, but is not mandatory. Feature selection may be part of discovery within a radiomics-specific trial or represent exploratory endpoints within an established trial; a previously validated radiomic signature may even be used as a primary/secondary endpoint, particularly if associations are demonstrated with specific biological processes and pathways being targeted within clinical trials. KEY POINTS: • Data-driven processes like radiomics risk false discoveries due to high-dimensionality of the dataset compared to sample size, making adequate diversity of the data, cross-validation and external validation essential to mitigate the risks of spurious associations and overfitting. • Use of radiomic signatures within clinical trials requires multistep standardisation of image acquisition, image analysis and data mining processes. • Biological correlation may be established after clinical validation but is not mandatory.

Learning fuzzy clustering for SPECT/CT segmentation via convolutional neural networks

PURPOSE

Quantitative bone single-photon emission computed tomography (QBSPECT) has the potential to provide a better quantitative assessment of bone metastasis than planar bone scintigraphy due to its ability to better quantify activity in overlapping structures. An important element of assessing the response of bone metastasis is accurate image segmentation. However, limited by the properties of QBSPECT images, the segmentation of anatomical regions-of-interests (ROIs) still relies heavily on the manual delineation by experts. This work proposes a fast and robust automated segmentation method for partitioning a QBSPECT image into lesion, bone, and background.

METHODS

We present a new unsupervised segmentation loss function and its semi- and supervised variants for training a convolutional neural network (ConvNet). The loss functions were developed based on the objective function of the classical Fuzzy C-means (FCM) algorithm. The first proposed loss function can be computed within the input image itself without any ground truth labels, and is thus unsupervised; the proposed supervised loss function follows the traditional paradigm of the deep learning-based segmentation methods and leverages ground truth labels during training. The last loss function is a combination of the first and the second and includes a weighting parameter, which enables semi-supervised segmentation using deep learning neural network.

EXPERIMENTS AND RESULTS

We conducted a comprehensive study to compare our proposed methods with ConvNets trained using supervised, cross-entropy and Dice loss functions, and conventional clustering methods. The Dice similarity coefficient (DSC) and several other metrics were used as figures of merit as applied to the task of delineating lesion and bone in both simulated and clinical SPECT/CT images. We experimentally demonstrated that the proposed methods yielded good segmentation results on a clinical dataset even though the training was done using realistic simulated images. On simulated SPECT/CT, the proposed unsupervised model's accuracy was greater than the conventional clustering methods while reducing computation time by 200-fold. For the clinical QBSPECT/CT, the proposed semi-supervised ConvNet model, trained using simulated images, produced DSCs of 0.75 and 0.74 for lesion and bone segmentation in SPECT, and a DSC of 0.79 bone segmentation of CT images. These DSCs were larger than that for standard segmentation loss functions by > 0.4 for SPECT segmentation, and > 0.07 for CT segmentation with P-values < 0.001 from a paired t-test.

CONCLUSIONS

A ConvNet-based image segmentation method that uses novel loss functions was developed and evaluated. The method can operate in unsupervised, semi-supervised, or fully-supervised modes depending on the availability of annotated training data. The results demonstrated that the proposed method provides fast and robust lesion and bone segmentation for QBSPECT/CT. The method can potentially be applied to other medical image segmentation applications.

Metrics to evaluate the performance of auto-segmentation for radiation treatment planning: A critical review

Advances in artificial intelligence-based methods have led to the development and publication of numerous systems for auto-segmentation in radiotherapy. These systems have the potential to decrease contour variability, which has been associated with poor clinical outcomes and increased efficiency in the treatment planning workflow. However, there are no uniform standards for evaluating auto-segmentation platforms to assess their efficacy at meeting these goals. Here, we review the most frequently used evaluation techniques which include geometric overlap, dosimetric parameters, time spent contouring, and clinical rating scales. These data suggest that many of the most commonly used geometric indices, such as the Dice Similarity Coefficient, are not well correlated with clinically meaningful endpoints. As such, a multi-domain evaluation, including composite geometric and/or dosimetric metrics with physician-reported assessment, is necessary to gauge the clinical readiness of auto-segmentation for radiation treatment planning.