Prospective evaluation of MR-TRG (Tumor Regression Grade) in esophageal cancer after neo-adjuvant therapy: Preliminary results.

PURPOSE: To develop MRI-based criteria to assess tumor response to neoadjuvant therapies (NAT) of esophageal cancers (EC) and to evaluate its diagnostic performance in predicting the pathological Tumor Regression Grade (pTRG). METHOD: From 2018 to 2022, patients with newly diagnosed locally advanced EC underwent MRI examinations for initial staging and restaging after NAT. Magnetic Resonance TRG (MR-TRG), equivalent to the Mandard and Becker classifications, were developed and independently assessed by two radiologists, blinded to pTRG, using T2W and DW-MR Images. All patients underwent surgery and benefited from a blinded pTRG evaluation by two pathologists. The agreement between readers and between MR-TRG and pTRG were assessed with Cohen's Kappa. The correlation of MR-TRG and pTRG was determined using Spearman's correlation. RESULTS: 28 patients were included. Interrater agreement was substantial between radiologists, improved when grouping grade 1 and 2 (κ = 0.78 rose to 0,84 for Mandard and 0.68 to 0,78 for Becker score). Agreement between pTRG and MR-TRG was moderate with a percentaged agreement (p) = 87.5 %, kappa (κ) = 0.54 and p = 83.3 %, κ = 0.49 for Mandard and Becker, respectively. Agreement was improved to substantial when grouping grades 1-2 for Mandard and 1a-1b for Becker with p = 89.3 %, κ = 0.65 and p = 85.2 %, κ = 0.65 respectively. Sensitivity and specificity of MR-TRG in predicting pTRG were 88.2 % and 72.7 % for Mandard system (scores 1-2 versus 3-5), and 83.3 % and 80 % for Becker system (scores 1a-1b versus 2-3). CONCLUSION: A substantial agreement between MR-TRG and pTRG was achieved when grouping grade 1-2. Hence, MR-TRG could be used as a surrogate of complete and near-complete pTRG.

Automatic Head and Neck Tumor segmentation and outcome prediction relying on FDG-PET/CT images: Findings from the second edition of the HECKTOR challenge.

By focusing on metabolic and morphological tissue properties respectively, FluoroDeoxyGlucose (FDG)-Positron Emission Tomography (PET) and Computed Tomography (CT) modalities include complementary and synergistic information for cancerous lesion delineation and characterization (e.g. for outcome prediction), in addition to usual clinical variables. This is especially true in Head and Neck Cancer (HNC). The goal of the HEad and neCK TumOR segmentation and outcome prediction (HECKTOR) challenge was to develop and compare modern image analysis methods to best extract and leverage this information automatically. We present here the post-analysis of HECKTOR 2nd edition, at the 24th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI) 2021. The scope of the challenge was substantially expanded compared to the first edition, by providing a larger population (adding patients from a new clinical center) and proposing an additional task to the challengers, namely the prediction of Progression-Free Survival (PFS). To this end, the participants were given access to a training set of 224 cases from 5 different centers, each with a pre-treatment FDG-PET/CT scan and clinical variables. Their methods were subsequently evaluated on a held-out test set of 101 cases from two centers. For the segmentation task (Task 1), the ranking was based on a Borda counting of their ranks according to two metrics: mean Dice Similarity Coefficient (DSC) and median Hausdorff Distance at 95th percentile (HD95). For the PFS prediction task, challengers could use the tumor contours provided by experts (Task 3) or rely on their own (Task 2). The ranking was obtained according to the Concordance index (C-index) calculated on the predicted risk scores. A total of 103 teams registered for the challenge, for a total of 448 submissions and 29 papers. The best method in the segmentation task obtained an average DSC of 0.759, and the best predictions of PFS obtained a C-index of 0.717 (without relying on the provided contours) and 0.698 (using the expert contours). An interesting finding was that best PFS predictions were reached by relying on DL approaches (with or without explicit tumor segmentation, 4 out of the 5 best ranked) compared to standard radiomics methods using handcrafted features extracted from delineated tumors, and by exploiting alternative tumor contours (automated and/or larger volumes encompassing surrounding tissues) rather than relying on the expert contours. This second edition of the challenge confirmed the promising performance of fully automated primary tumor delineation in PET/CT images of HNC patients, although there is still a margin for improvement in some difficult cases. For the first time, the prediction of outcome was also addressed and the best methods reached relatively good performance (C-index above 0.7). Both results constitute another step forward toward large-scale outcome prediction studies in HNC.

Lung CT stabilization with high-frequency non-invasive ventilation (HF-NIV) and breath-hold (BH) in lung nodule assessment by PET/CT.

PURPOSE: To evaluate the effect of lung stabilization using high-frequency non-invasive ventilation (HF-NIV) and breath-hold (BH) techniques on lung nodule detection and texture assessment in PET/CT compared to a free-breathing (FB) standard lung CT acquisition in PET/CT. MATERIALS AND METHODS: Six patients aged 65 ± 7 years, addressed for initial assessment of at least one suspicious lung nodule with 18F-FDG PET/CT, underwent three consecutive lung PET/CT acquisitions with FB, HF-NIV and BH. Lung nodules were assessed on all three CT acquisitions of the PET/CT and characterized for any size, volume and solid/sub-solid nature. RESULTS: BH detected a significantly higher number of nodules (n = 422) compared to HF-NIV (n = 368) and FB (n = 191) (p < 0.001). The mean nodule size (mm) was 2.4 ± 2.1, 2.6 ± 1.9 and 3.2 ± 2.4 in BH, HF-NIV and FB, respectively, for long axis and 1.5 ± 1.3, 1.6 ± 1.2 and 2.1 ± 1.7 in BH, HF-NIV and FB, respectively, for short axis. Long- and short-axis diameters were significantly different between BH and FB (p < 0.001) and between HF-NIV and FB (p < 0.001 and p = 0.008), but not between BH and HF-NIV. A trend for higher volume was shown in FB compared to BH (p = 0.055) and HF-NIV (p = 0.068) without significant difference between BH and HF-NIV (p = 1). We found a significant difference in detectability of sub-solid nodules between the three acquisitions, with BH showing a higher number of sub-solid nodules (n = 128) compared to HF-NIV (n = 72) and FB (n = 44) (p = 0.002). CONCLUSION: We observed a higher detection rate of pulmonary nodules on CT under BH or HF-NIV conditions applied to PET/CT than with FB. BH and HF-NIV demonstrated comparable texture assessment and performed better than FB in assessing size and volume. BH showed a better performance for detecting sub-solid nodules compared to HF-NIV and FB. The addition of BH or HF-NIV to PET/CT can help improve the detection and texture characterization of lung nodules by CT, therefore improving the accuracy of oncological lung disease assessment. The ease of use of BH and its added value should prompt its use in routine practice.

The combination of non-contrast abbreviated MRI and alpha foetoprotein has high performance for hepatocellular carcinoma screening.

OBJECTIVES: This study aimed to compare two abbreviated MRI (AMRI) protocols to complete MRI for HCC detection: non-contrast (NC)-AMRI without/with alpha foetoprotein (AFP) and dynamic contrast-enhanced (Dyn)-AMRI. METHODS: This retrospective single-center study included 351 patients (M/F: 264/87, mean age: 57y) with chronic liver disease, who underwent MRI for HCC surveillance between 2014 and 2020. Two reconstructed AMRI sets were obtained based on complete MRI: NC-AMRI (T2-weighted imaging (WI) + diffusion-WI) and Dyn-AMRI (T2-WI + dynamic T1-WI) and were assessed by 2 radiologists who reported all suspicious lesions, using LI-RADS/adapted LI-RADS classification. The reference standard was based on all available patient data. Inter-reader agreement was assessed and MRI diagnostic performance was compared to the reference standard. RESULTS: The reference standard demonstrated 83/351 HCC-positive patients (prevalence: 23.6%, median size: 22 mm, and positive MRIs: 83/631). Inter-reader agreement was substantial for all sets. Sensitivities of Dyn-AMRI and complete MRI (both 92.8%) were similar, higher than NC-AMRI (72.3%, p < 0.001). Specificities were not different between sets. NC-AMRI + AFP (92.8%) had similar sensitivity to Dyn-AMRI and complete MRI. In patients with small size HCCs (≤ 2 cm), sensitivities of Dyn-AMRI (85.3%) and complete MRI (88.2%) remained similar (p = 0.564), also outperforming NC-AMRI (52.9%, p < 0.05). NC-AMRI + AFP had similar sensitivity (88.2%) to Dyn-AMRI and complete MRI (p = 0.706 and p = 1, respectively). CONCLUSIONS: Dyn-AMRI has similar diagnostic performance to complete MRI for HCC detection, while both outperform NC-AMRI, especially for small size HCCs. NC-AMRI + AFP demonstrates similar sensitivity to Dyn-AMRI and complete MRI. CLINICAL RELEVANCE STATEMENT: Due to the low sensitivity of ultrasound for hepatocellular screening, new screening methods are needed. Abbreviated MRI (AMRI) is a candidate, especially non-contrast AMRI with serum alpha foetoprotein as the acquisition time is low, without the need for contrast medium injection. KEY POINTS: • Dynamic contrast-enhanced abbreviated MRI using extracellular gadolinium-based contrast agent and complete MRI have similar diagnostic performance for hepatocellular carcinoma detection in an at-risk population. • Non-contrast abbreviated MRI with alpha foetoprotein has similar diagnostic performance to dynamic contrast-enhanced abbreviated MRI and complete MRI, including when considering small size hepatocellular carcinoma ≤ 2 cm. • Non-contrast abbreviated MRI and dynamic contrast-enhanced abbreviated MRI can be performed in 7 and 10 min, excluding patient setup time.

Esophageal cancer T-staging on MRI: A preliminary study using cine and static MR sequences.

OBJECTIVES: To evaluate the added value of cine MR in addition to static MRI for T-Staging assessment of esophageal cancer (EC). MATERIALS AND METHODS: This prospective monocentric study included 54 patients (mean age 66.3 ± 9.4 years, 46 men) with histologically proven EC. They underwent MRI on a 3 T-scanner in addition to the standard workup. Acquisitions included static and cine sequences (steady-state-free-precession and real-time True-FISP during water ingestion). Three radiologists independently assessed T-staging and diagnosis confidence by reviewing (1) static sequences (S-MRI) and (2) adding cine sequences (SC-MRI). Inter-reader agreement was performed. MRI T-staging was correlated to reference standard T-staging (histopathology or consensus on endoscopic ultrasonography and imaging findings) and to clinical outcome by log-rank test. RESULTS: Both S-MRI and SC-MRI T-staging showed a significant correlation with reference T-staging (rs = 0.667, P < 0.001). SC-MRI showed a slightly better performance in distinguishing T1-T3 from T4 with a sensitivity, specificity and AUC of 76.5% (95% CI: 50.1-93.2), 83.8% (68-93.8) and 0.801 (0.681-0.921) vs 70.6% (44-89.7), 83% (68-93.8) and 0.772 (0.645-0.899) for S-MRI. Compared to S-MRI, SC-MRI increased inter-reader agreement for T4a and T4b (κ = 0.403 and 0.498) and T-staging confidence. CONCLUSION: MRI is accurate for T-staging of EC. The addition of cine sequences allows better differentiation between T1-T3 and T4 tumors with increased diagnostic confidence and inter-reader agreement.

Feasibility and Performance of Free-Hand Single-Photon Computed Tomography/Ultrasonography for Preoperative Parathyroid Adenoma Localization: A Pilot Study.

The aim of this prospective pilot study was to evaluate the feasibility of a new hybrid imaging modality, free-hand single-photon computed tomography/ultrasonography (fhSPECT/US), for preoperative localization of parathyroid adenomas and to compare its performance with conventional ultrasonography and SPECT/CT. Twelve patients diagnosed with primary hyperparathyroidism underwent sequentially US and parathyroid scintigraphy, including SPECT/CT, followed by fhSPECT/US, allowing for real-time fusion between US and freehand-generated gamma-camera images. The fhSPECT/US detection rates were correlated with histopathology, when available, or with the imaging modality showing the most lesions. Based on a per patient analysis, the detection rate was significantly different when comparing SPECT/CT to fhSPECT/US (p = 0.03), and not significantly different when comparing SPECT/CT to US (p = 0.16) and US to fhSPECT/US (p = 0.08). Based on a per-lesion analysis, the detection rate of SPECT/CT was significantly higher than that of US (p = 0.01) and fhSEPCT/US (p = 0.003), and there was no significant difference in detection rate when comparing US to fhSPECT/US (p = 0.08). The main perceived limitations of fhSPECT/US in lesion detection were: (i) lesions localized at a depth ≥4.5 cm; (ii) imperfect image fusion due to tissue compression; (iii) limited spatial manipulation ability of the SPECT mobile camera handheld probe; and (iv) a wide spread of detected activity. In conclusion, clinical use of fhSPECT/US for localization of parathyroid adenomas is feasible, but shows lower sensitivity than conventional modalities and requires technical improvements.

Modeling tumor size dynamics based on real-world electronic health records and image data in advanced melanoma patients receiving immunotherapy.

The development of immune checkpoint inhibitors (ICIs) has revolutionized cancer therapy but only a fraction of patients benefits from this therapy. Model-informed drug development can be used to assess prognostic and predictive clinical factors or biomarkers associated with treatment response. Most pharmacometric models have thus far been developed using data from randomized clinical trials, and further studies are needed to translate their findings into the real-world setting. We developed a tumor growth inhibition model based on real-world clinical and imaging data in a population of 91 advanced melanoma patients receiving ICIs (i.e., ipilimumab, nivolumab, and pembrolizumab). Drug effect was modeled as an ON/OFF treatment effect, with a tumor killing rate constant identical for the three drugs. Significant and clinically relevant covariate effects of albumin, neutrophil to lymphocyte ratio, and Eastern Cooperative Oncology Group (ECOG) performance status were identified on the baseline tumor volume parameter, as well as NRAS mutation on tumor growth rate constant using standard pharmacometric approaches. In a population subgroup (n = 38), we had the opportunity to conduct an exploratory analysis of image-based covariates (i.e., radiomics features), by combining machine learning and conventional pharmacometric covariate selection approaches. Overall, we demonstrated an innovative pipeline for longitudinal analyses of clinical and imaging RWD with a high-dimensional covariate selection method that enabled the identification of factors associated with tumor dynamics. This study also provides a proof of concept for using radiomics features as model covariates.

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.

Semiautomated Pipeline to Quantify Tumor Evolution From Real-World Positron Emission Tomography/Computed Tomography Imaging.

PURPOSE: A semiautomated pipeline for the collection and curation of free-text and imaging real-world data (RWD) was developed to quantify cancer treatment outcomes in large-scale retrospective real-world studies. The objectives of this article are to illustrate the challenges of RWD extraction, to demonstrate approaches for quality assurance, and to showcase the potential of RWD for precision oncology. METHODS: We collected data from patients with advanced melanoma receiving immune checkpoint inhibitors at the Lausanne University Hospital. Cohort selection relied on semantically annotated electronic health records and was validated using process mining. The selected imaging examinations were segmented using an automatic commercial software prototype. A postprocessing algorithm enabled longitudinal lesion identification across imaging time points and consensus malignancy status prediction. Resulting data quality was evaluated against expert-annotated ground-truth and clinical outcomes obtained from radiology reports. RESULTS: The cohort included 108 patients with melanoma and 465 imaging examinations (median, 3; range, 1-15 per patient). Process mining was used to assess clinical data quality and revealed the diversity of care pathways encountered in a real-world setting. Longitudinal postprocessing greatly improved the consistency of image-derived data compared with single time point segmentation results (classification precision increased from 53% to 86%). Image-derived progression-free survival resulting from postprocessing was comparable with the manually curated clinical reference (median survival of 286 v 336 days, P = .89). CONCLUSION: We presented a general pipeline for the collection and curation of text- and image-based RWD, together with specific strategies to improve reliability. We showed that the resulting disease progression measures match reference clinical assessments at the cohort level, indicating that this strategy has the potential to unlock large amounts of actionable retrospective real-world evidence from clinical records.

QuantImage v2: a comprehensive and integrated physician-centered cloud platform for radiomics and machine learning research.

BACKGROUND: Radiomics, the field of image-based computational medical biomarker research, has experienced rapid growth over the past decade due to its potential to revolutionize the development of personalized decision support models. However, despite its research momentum and important advances toward methodological standardization, the translation of radiomics prediction models into clinical practice only progresses slowly. The lack of physicians leading the development of radiomics models and insufficient integration of radiomics tools in the clinical workflow contributes to this slow uptake. METHODS: We propose a physician-centered vision of radiomics research and derive minimal functional requirements for radiomics research software to support this vision. Free-to-access radiomics tools and frameworks were reviewed to identify best practices and reveal the shortcomings of existing software solutions to optimally support physician-driven radiomics research in a clinical environment. RESULTS: Support for user-friendly development and evaluation of radiomics prediction models via machine learning was found to be missing in most tools. QuantImage v2 (QI2) was designed and implemented to address these shortcomings. QI2 relies on well-established existing tools and open-source libraries to realize and concretely demonstrate the potential of a one-stop tool for physician-driven radiomics research. It provides web-based access to cohort management, feature extraction, and visualization and supports "no-code" development and evaluation of machine learning models against patient-specific outcome data. CONCLUSIONS: QI2 fills a gap in the radiomics software landscape by enabling "no-code" radiomics research, including model validation, in a clinical environment. Further information about QI2, a public instance of the system, and its source code is available at https://medgift.github.io/quantimage-v2-info/ . Key points As domain experts, physicians play a key role in the development of radiomics models. Existing software solutions do not support physician-driven research optimally. QuantImage v2 implements a physician-centered vision for radiomics research. QuantImage v2 is a web-based, "no-code" radiomics research platform.

A novel approach for fibrous dysplasia assessment using combined planar and quantitative SPECT/CT analysis of Tc-99m-diphosphonate bone scan in correlation with biological bone turnover markers of disease activity.

Purpose: To investigate the emerging role of Tc-99m-labeled diphosphonate (Tc-99m-DPD) uptake quantification by SPECT/CT in fibrous dysplasia (FD) bone lesions and its correlation with biological bone turnover markers (BTMs) of disease activity. Materials and methods: Seven patients (49 ± 16 years) with a confirmed diagnosis of FD were included in this retrospective study. Bone scans with Tc-99m-DPD and quantitative SPECT/CT (xSPECT/CT) were performed. SUVmax (maximum standard unit value) and SUVmean (mean standard unit value) were measured in all FD bone lesions. The skeletal burden score (SBS) was assessed on planar scintigraphy and multiplied by mean SUV max and SUVmean to generate two new parameters, SBS_SUVmax and SBS_SUVmean, respectively. Planar and xSPECT/CT quantitative measures were correlated with biological BTMs of disease activity, including fibroblast growth factor 23 (FGF-23), alkaline phosphatase (ALP), procollagen 1 intact N-terminal propeptide (P1NP) and C-terminal telopeptide (CTX), as well as scoliosis angle measured on radiographs. Statistical significance was evaluated with Spearman's correlations. Results: A total of 76 FD bone lesions were analyzed, showing an average SUVmax and SUVmean (g/mL) of 13 ± 7.3 and 8 ± 4.5, respectively. SBS, SBS_SUVmax and SBS_SUVmean values were 30.8 ± 25.6, 358 ± 267 and 220.1 ± 164.5, respectively. Mean measured values of FGF-23 (pg/mL), ALP (U/L), P1NP (µg/L) and CTX (pg/mL) were 98.4 (22-175), 283.5 (46-735), 283.1 (31-1,161) and 494 (360-609), respectively. Mean scoliosis angle was 15.7 (7-22) degrees. We found a very strong positive correlation between planar-derived SBS and CTX (r = 0.96, p = 0.010), but no significant correlation between SUVmax or SUVmean and biological BTMs. SBS_SUVmax showed a strong to very strong positive correlation with CTX (ρ = 0.99, p = 0.002), FGF-23 (ρ = 0.91, p = 0.010), ALP (ρ = 0.82, p = 0.020), and P1NP (ρ = 0.78, p = 0.039), respectively. Conclusion: This study showed that biological BTMs are significantly correlated with diphosphonate uptake on bone scan, quantified by a new parameter combining information from both planar and quantitative SPECT/CT. Further analysis of bone scan quantitative SPECT/CT data in a larger patient population might help better characterize the skeletal disease burden in FD, and guide treatment and follow-up.

Reproducibility of lung cancer radiomics features extracted from data-driven respiratory gating and free-breathing flow imaging in [18F]-FDG PET/CT.

BACKGROUND: Quality and reproducibility of radiomics studies are essential requirements for the standardisation of radiomics models. As recent data-driven respiratory gating (DDG) [18F]-FDG has shown superior diagnostic performance in lung cancer, we evaluated the impact of DDG on the reproducibility of radiomics features derived from [18F]-FDG PET/CT in comparison to free-breathing flow (FB) imaging. METHODS: Twenty four lung nodules from 20 patients were delineated. Radiomics features were derived on FB flow PET/CT and on the corresponding DDG reconstruction using the QuantImage v2 platform. Lin's concordance factor (Cb) and the mean difference percentage (DIFF%) were calculated for each radiomics feature using the delineated nodules which were also classified by anatomical localisation and volume. Non-reproducible radiomics features were defined as having a bias correction factor Cb < 0.8 and/or a mean difference percentage DIFF% > 10. RESULTS: In total 141 features were computed on each concordance analysis, 10 of which were non-reproducible on all pulmonary lesions. Those were first-order features from Laplacian of Gaussian (LoG)-filtered images (sigma = 1 mm): Energy, Kurtosis, Minimum, Range, Root Mean Squared, Skewness and Variance; Texture features from Gray Level Cooccurence Matrix (GLCM): Cluster Prominence and Difference Variance; First-order Standardised Uptake Value (SUV) feature: Kurtosis. Pulmonary lesions located in the superior lobes had only stable radiomics features, the ones from the lower parts had 25 non-reproducible radiomics features. Pulmonary lesions with a greater size (defined as long axis length > median) showed a higher reproducibility (9 non-reproducible features) than smaller ones (20 non-reproducible features). CONCLUSION: Calculated on all pulmonary lesions, 131 out of 141 radiomics features can be used interchangeably between DDG and FB PET/CT acquisitions. Radiomics features derived from pulmonary lesions located inferior to the superior lobes are subject to greater variability as well as pulmonary lesions of smaller size.

Segmentation and Classification of Head and Neck Nodal Metastases and Primary Tumors in PET/CT.

The prediction of cancer characteristics, treatment planning and patient outcome from medical images generally requires tumor delineation. In Head and Neck cancer (H&N), the automatic segmentation and differentiation of primary Gross Tumor Volumes (GTVt) and malignant lymph nodes (GTVn) is a necessary step for large-scale radiomics studies to predict patient outcome such as Progression Free Survival (PFS). Detecting malignant lymph nodes is also a crucial step for Tumor-Node-Metastases (TNM) staging and to support the decision to resect the nodes. In turn, automatic TNM staging and patient outcome prediction can greatly benefit patient care by helping clinicians to find the best personalized treatment. We propose the first model to automatically individually segment GTVt and GTVn in PET/CT images. A bi-modal 3D U-Net model is trained for multi-class and multi-components segmentation on the multi-centric HECKTOR 2020 dataset containing 254 cases. The dataset has been specifically re-annotated by experts to obtain ground truth GTVn contours. The results show promising segmentation performance for the automation of radiomics pipelines and their validation on large-scale studies for which manual annotations are not available. An average test Dice Similarity Coefficients (DSC) of 0.717 is obtained for the segmentation of GTVt. The GTVn segmentation is evaluated with an aggregated DSC to account for the cases without GTVn, which is estimated at 0.729 on the test set.

Comparison Between Magnetic Resonance Imaging and Computed Tomography in the Detection and Volumetric Assessment of Lung Nodules: A Prospective Study.

Rationale and Objectives: Computed tomography (CT) lung nodule assessment is routinely performed and appears very promising for lung cancer screening. However, the radiation exposure through time remains a concern. With the overall goal of an optimal management of indeterminate lung nodules, the objective of this prospective study was therefore to evaluate the potential of optimized ultra-short echo time (UTE) MRI for lung nodule detection and volumetric assessment. Materials and Methods: Eight (54.9 ± 13.2 years) patients with at least 1 non-calcified nodule ≥4 mm were included. UTE under high-frequency non-invasive ventilation (UTE-HF-NIV) and in free-breathing at tidal volume (UTE-FB) were investigated along with volumetric interpolated breath-hold examination at full inspiration (VIBE-BH). Three experienced readers assessed the detection rate of nodules ≥4 mm and ≥6 mm, and reported their location, 2D-measurements and solid/subsolid nature. Volumes were measured by two experienced readers. Subsequently, two readers assessed the detection and volume measurements of lung nodules ≥4mm in gold-standard CT images with soft and lung kernel reconstructions. Volumetry was performed with lesion management software (Carestream, Rochester, New York, USA). Results: UTE-HF-NIV provided the highest detection rate for nodules ≥4 mm (n = 66) and ≥6 mm (n = 32) (35 and 50%, respectively). No dependencies were found between nodule detection and their location in the lung with UTE-HF-NIV (p > 0.4), such a dependency was observed for two readers with VIBE-BH (p = 0.002 and 0.03). Dependencies between the nodule's detection and their size were noticed among readers and techniques (p < 0.02). When comparing nodule volume measurements, an excellent concordance was observed between CT and UTE-HF-NIV, with an overestimation of 13.2% by UTE-HF-NIV, <25%-threshold used for nodule's growth, conversely to VIBE-BH that overestimated the nodule volume by 28.8%. Conclusion: UTE-HF-NIV is not ready to replace low-dose CT for lung nodule detection, but could be used for follow-up studies, alternating with CT, based on its volumetric accuracy.

Cleaning radiotherapy contours for radiomics studies, is it worth it? A head and neck cancer study.

A vast majority of studies in the radiomics field are based on contours originating from radiotherapy planning. This kind of delineation (e.g. Gross Tumor Volume, GTV) is often larger than the true tumoral volume, sometimes including parts of other organs (e.g. trachea in Head and Neck, H&N studies) and the impact of such over-segmentation was little investigated so far. In this paper, we propose to evaluate and compare the performance between models using two contour types: those from radiotherapy planning, and those specifically delineated for radiomics studies. For the latter, we modified the radiotherapy contours to fit the true tumoral volume. The two contour types were compared when predicting Progression-Free Survival (PFS) using Cox models based on radiomics features extracted from FluoroDeoxyGlucose-Positron Emission Tomography (FDG-PET) and CT images of 239 patients with oropharyngeal H&N cancer collected from five centers, the data from the 2020 HECKTOR challenge. Using Dedicated contours demonstrated better performance for predicting PFS, where Harell's concordance indices of 0.61 and 0.69 were achieved for Radiotherapy and Dedicated contours, respectively. Using automatically Resegmented contours based on a fixed intensity range was associated with a C-index of 0.63. These results illustrate the importance of using clean dedicated contours that are close to the true tumoral volume in radiomics studies, even when tumor contours are already available from radiotherapy treatment planning.

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.

Pulmonary Lymphangitic Carcinomatosis: Diagnostic Performance of High-Resolution CT and 18F-FDG PET/CT in Correlation with Clinical Pathologic Outcome.

The rationale of this study was to investigate the performance of high-resolution CT (HRCT) versus 18F-FDG PET/CT for the diagnosis of pulmonary lymphangitic carcinomatosis (PLC). Methods: In this retrospective institution-approved study, 94 patients addressed for initial staging of lung cancer with suspicion of PLC were included. Using double-blind analysis, we assessed the presence of signs favoring PLC on HRCT (smooth or nodular septal lines, subpleural nodularity, peribronchovascular thickening, satellite nodules, lymph node enlargement, and pleural effusion). 18F-FDG PET/CT images were reviewed to qualitatively evaluate peritumoral uptake and to quantify tracer uptake in the tumoral and peritumoral areas. Histology performed on surgical specimens served as the gold standard for all patients. Results: Among 94 included patients, 73% (69/94) had histologically confirmed PLC. Peribronchovascular thickening, lymph node involvement, and increased peritumoral uptake were more often present in patients with PLC (P < 0.009). Metabolic variables, including tumor SUVmax, SUVmean, metabolic tumor volume, and total lesion glycolysis, as well as peritumoral SUVmax, SUVmean, and their respective ratios to background, were significantly higher in the PLC group than in the non-PLC group (P ≤ 0.0039). Sensitivity, specificity, and area under the receiver-operating-characteristic curve for peribronchovascular thickening (69%, 83%, and 0.76, respectively; 95% confidence interval [95%CI], 0.67-0.85) and increased peritumoral uptake (94%, 84%, and 0.89, respectively; 95%CI, 0.81-0.97) were similar (P = 0.054). For detecting PLC, sensitivity, specificity, and area under the receiver-operating-characteristic curve were significantly higher, at 97%, 92%, and 0.98, respectively (95%CI, 0.96-1.00), for peritumoral SUVmax and 94%, 88%, and 0.96, respectively (95%CI, 0.92-1.00), for peritumoral SUVmean (all P ≤ 0.025). Conclusion: Qualitative evaluation of 18F-FDG PET/CT and HRCT perform similarly for the diagnosis of PLC, with both being outperformed by 18F-FDG PET/CT quantitative parameters.

Quantitative bone SPECT/CT: high specificity for identification of prostate cancer bone metastases.

PURPOSE: Bone scintigraphy with 99mTc-labeled diphosphonates can identify prostate cancer bone metastases with high sensitivity, but relatively low specificity, because benign conditions such as osteoarthritis can also trigger osteoblastic reactions. We aimed to investigate the diagnostic performance of 99mTc-2,3-dicarboxy propane-1,1-diphosphonate (99mTc-DPD) uptake quantification by single-photon emission computed tomography coupled with computed tomography (SPECT/CT) for distinguishing prostate cancer bone metastases from spinal and pelvic osteoarthritic lesions. METHODS: We retrospectively assessed 26 bone scans from 26 patients with known prostate cancer bone metastases and 13 control patients with benign spinal and pelvic osteoarthritic changes without known neoplastic disease. Quantitative SPECT/CT (xSPECT, Siemens Symbia Intevo, Erlangen, Germany) was performed and standardized uptake values (SUVs) were quantified with measurements of SUVmax and SUVmean (g/mL) in all bone metastases for the prostate cancer group and in spinal and pelvic osteoarthritic changes for the control group. We used receiver operating characteristics (ROC) curves to determine the optimum SUVmax cutoff value to distinguish between bone metastases and benign spinal and pelvic lesions. RESULTS: In total, 264 prostate cancer bone metastases were analyzed, showing a mean SUVmax and SUVmean of 34.6 ± 24.6 and 20.8 ± 14.7 g/mL, respectively. In 24 spinal and pelvic osteoarthritic lesions, mean SUVmax and SUVmean were 14.2 ± 3.8 and 8.9 ± 2.2 g/mL, respectively. SUVmax and SUVmean were both significantly different between the bone metastases and osteoarthritic groups (p ≤ 0.0001). Using a SUVmax cutoff of 19.5 g/mL for prostate cancer bone metastases in the spine and pelvis, sensitivity, specificity, positive and negative predictive values were 87, 92, 99 and 49%, respectively. CONCLUSION: This study showed significant differences in quantitative 99mTc-DPD uptake on bone SPECT/CT between prostate cancer bone metastases and spinal and pelvic osteoarthritic changes, with higher SUVmax and SUVmean in metastases. Using a SUVmax cutoff of 19.5 g/mL, high specificity and positive predictive value for metastases identification in the spine and pelvis were found, thus increasing accuracy of bone scintigraphy.

Detection Rate of Culprit Tumors Causing Osteomalacia Using Somatostatin Receptor PET/CT: Systematic Review and Meta-Analysis.

BACKGROUND: Tumor-induced or oncogenic osteomalacia (TIO) is a rare paraneoplastic syndrome in which osteomalacia is a consequence of fibroblast growth factor 23 (FGF23) secretion by a mesenchymal tumor. The localization of the culprit lesion in patients with TIO is often challenging. Several studies have evaluated the detection rate (DR) of these tumors using somatostatin receptor positron emission tomography (SSTR-PET/CT). We aimed to summarize literature findings on this topic providing pooled estimates of DR. METHODS: A comprehensive literature search by screening PubMed, Embase and Cochrane library electronic databases through August 2019 was performed. The pooled DR of culprit tumors using SSTR-PET/CT in patients with TIO was calculated using a random-effects statistical model. RESULTS: Fourteen studies on the use of SSTR-PET/CT in detecting the culprit tumor in patients with TIO were included in the qualitative analysis. The pooled DR of SSTR-PET/CT on a per-patient-based analysis calculated using eleven studies (166 patients) was 87.6% (95% confidence interval (95% CI) 80.2-95.1%). Statistical heterogeneity among studies was detected (I-square = 63%), likely due to the use of different radiolabeled somatostatin analogues, as demonstrated by a subgroup analysis. CONCLUSIONS: Despite limited literature data due to the rarity of the disease, SSTR-PET/CT demonstrated a very high DR of culprit tumors in patients with TIO and it could be used as first-line imaging method for this indication.

18F-FDG PET metabolic-to-morphological volume ratio predicts PD-L1 tumour expression and response to PD-1 blockade in non-small-cell lung cancer.

PURPOSE: Anti-PD-1/PD-L1 blockade can restore tumour-specific T-cell immunity and is an emerging therapy in non-small-cell lung cancer (NSCLC). We investigated the correlation between 18F-FDG PET/CT-based markers and tumour tissue expression of PD-L1, necrosis and clinical outcome in patients receiving checkpoint inhibitor treatment. METHODS: PD-Li expression in biopsy or resection specimens from 49 patients with confirmed NSCLC was investigated by immunohistochemistry. Maximum standardized uptake value (SUVmax), mean SUV (SUVmean), metabolic tumour volume (MTV) and total lesion glycolysis (TLG) were obtained from 18F-FDG PET/CT images. The ratio of metabolic to morphological lesion volumes (MMVR) and its association with PD-L1 expression in each lesion were calculated. The associations between histologically reported necrosis and 18F-FDG PET imaging patterns and radiological outcome (evaluated by iRECIST) following anti-PD-1/PD-L1 therapy were also analysed. In 14 patients, the association between necrosis and MMVR and tumour immune contexture were analysed by multiple immunofluorescent (IF) staining for CD8, PD-1, granzyme B (GrzB) and NFATC2. RESULTS: In total, 25 adenocarcinomas and 24 squamous cell carcinomas were analysed. All tumours showed metabolic 18F-FDG PET uptake. MMVR was correlated inversely with PD-L1 expression in tumour cells. Furthermore, PD-L1 expression and low MMVR were significantly correlated with clinical benefit. Necrosis was correlated negatively with MMVR. Multiplex IF staining showed a greater frequency of activated CD8+ cells in necrotic tumours than in nonnecrotic tumours in both stromal and epithelial tumour compartments. CONCLUSION: This study introduces MMVR as a new imaging biomarker and its ability to noninvasively capture increased PD-L1 tumour expression and predict clinical benefit from checkpoint blockade in NSCLC should be further evaluated.