Delineation Of Lesions Research Articles

Estimation of Optimum Flip Angle using 3D VANE XD Technique: Focused on Pre-Contrast and Hepatobiliary Phase

To investigate the optimum flip angle that can enhance image quality, SNR (signal to noise ratio), and CNR (contrast to noise ratio), comparing the images obtained, applying flip angles, 11°, 14°, 17°, 20°, and 23° in getting Liver Hepatobiliary Phase image using 3D VANE XD(3D Multivane mDixon, Philips Healthcare) technique. Experiments were conducted on a total of 30 outpatients and inpatients to our hospital (HCC:10, Metastasis:10, Abscess:10). As for the equipment used in the experiments, Philips Ingenia 3.0T CX was used, and all parameters other than the flip angle were set the same to conduct the tests. As for the image analysis method, using the Image-J program (National Institutes of Health and LOCI), the SNR of the liver, kidney, and pancreas obtained from the images by flip angle before the contrast medium injection and the CNR between the lesion and the normal tissue after the contrast medium injection were measured to conduct comparative analysis. As a result of a comparison of images before and after the contrast medium injection by disease, when the flip angle of 17° was applied, SNR and CNR were measured higher than in the images of other flip angles (p0.05). In the comparisons of the images taken before and after the injection of contrast medium by disease, when a flip angle of 17° was applied, the SNR before contrast medium injection was 28-29 % higher, and the SNR after the injection of contrast medium was 11 % up to 49 % higher than that at other flip angles. There was a difference in CNR before contrast medium injection of 30-43 % and CNR after contrast medium injection of 58-68 %. The measured value increased up to 17° and then decreased after that. Additionally, in the qualitative evaluation, Lesion Conspicuity (p=0.003), Image Artifact (p=0.0001), Lesion Delineation (p=0.0002), and Vascular Anatomy (p=0.0002) received the most excellent evaluations at 17°. In conclusion, in this study, the flip angle of 17° provided the highest SNR and CNR values when the tests were conducted using the free breath hold technique, 3D VANE XD Sequence. Thus, in liver MRI protocol tests, the overall diagnostic information was provided, including hypervascular tumor.

The Evaluation of Zoomed Echo-Planar Diffusion-Weighted Magnetic Resonance Imaging With Two-dimensional Spatial-Selective Radiofrequency Excitation Pulses in Patients With Hilar Cholangiocarcinoma.

BackgroundWe aimed to investigate the feasibility and application of using the zoomed diffusion-weighted echo-planar imaging (z-EPI DWI) sequences for hilar cholangiocarcinoma assessment compared with conventional single-shot EPI diffusion-weighted imaging (c-EPI DWI).MethodsBoth c-EPI DWI and z-EPI DWI were preoperatively performed in 16 patients with histopathologically-confirmed hilar cholangiocarcinoma. A two-dimensional spatial-selective radiofrequency (RF) pulse was applied to the z-EPI DWI using an echo-planar transmit trajectory. Anatomic structural visualization, lesion conspicuity, artifact presence and overall image quality were evaluated and compared between the two sequence images. The ratio of differences regarding hilar cholangiocarcinoma lesion sizes measured on T2-weighted imaging (T2WI) and diffusion-weighted imaging (DWI) were compared from both EPI techniques. The DW images for tumor involvement of the bile duct were reviewed based on histopathological examination of the surgical intraoperative evaluation. ADC measurements of DWIs in the hilar cholangiocarcinoma lesions were conducted.ResultsThe hepatic hilar region was better delineated by visualization of anatomical structures, lesion conspicuity and overall image quality using the z-EPI DWI and these analyses were compared with the c-EPI DWI method (all p<0.05). Better lesion delineation of bile duct walls and lumens was noted in four patients with z-EPI DWI compared with those of c-EPI DWI. No significant differences were noted between the two image datasets for artifacts (p=0.876). The ratio of differences regarding hilar cholangiocarcinoma lesion sizes was significantly lower (p= 0.018) on T2WI and DWI, as determined by the z-EPI DWI than that determined by the c-EPI method. The use of z-EPI DWI resulted in the accurate diagnosis of the Bismuth-Corlette classification of 15 tumors (15/16, 93.75%), whereas the use of c-EPI DWI resulted in correct diagnosis of 12 tumors (12/16, 75.00%). There were no significant differences between c-EPI DWI and z-EPI DWI in the ADC values of hilar cholangiocarcinoma lesions (p= 0.48).Conclusionz-EPI DWI resulted in remarkable image quality improvements for hilar cholangiocarcinoma. The ability to detect and delineate lesions using z-EPI DWI was superior to that of c-EPI DWI.

Abstract CT109: Radiomic features in tumor assessment, preliminary results from a phase 2 study of intratumoral administration of BO-112 with pembrolizumab in patients with anti PD1 refractory melanoma

Abstract Background: There is a lack of predictive biomarkers for immunotherapy in melanoma. Immunohistochemistry and gene sequencing are frequently assessed as part of clinical research. Radiomic signatures may also add valuable information, based on parameters which can be related to immune infiltration, therefore defining an imaging biomarkers panel for this clinical scenario. BO-112 is a double stranded synthetic RNA formulated with polyethyleneimine (PEI) that, by mimicking the effect of a viral infection, mobilizes the immune system. The role of imaging biomarkers is being explored in the present phase II clinical trial. Study design: Single arm study with BO-112 plus pembrolizumab (NCT04570332) in patients with advanced melanoma in progression to anti-PD1 therapy. As part of exploratory objectives, a radiomics analysis was performed to detect changes in lesion texture features. Quantitative features were obtained by using Quibim Precision 2.9 platform (Quibim, Valencia, Spain) after the manual delineation of lesions on the CT study of each subject at each timepoint to obtain information about injected/non-injected lesions. Images were normalized by taking into account Hounsfield Units (HU) bias across scanners in a cross-calibration phantom and the Z-score. The difference (Delta) in the features between baseline and week 8 was calculated. Results: Studies were assessed based on event (progression) and based on whether lesions had been injected. Patients with only cutaneous disease were not included in this analysis. Out of 23 patients who had at least two imaging assessments, 6 developed progressive disease by week 8, and 17 subjects had no event by that time. Due to the small sample size, the radiomic analysis was based on hypothesis testing. With regards to volume, 50% of the non-progressing lesions reduced their value from that of the baseline. Regarding injected versus non-injected changes, up to 50% of injected lesions decreased their volume after 8 weeks of injected treatment whereas in non-injected lesions volume decreased in less than 25% of lesions. From the independent sample t-test of delta radiomics features against the injected/non-injected lesions variable, there were 4 features with a statistically significant difference between groups; all of them related to the Low Grey-Level Zone Emphasis. Specifically, Delta original GLRLM Low Grey-Level Run Emphasis showed the highest significant difference between injected and non-injected lesions (p=0.004), with higher and positive delta values in the injected group (75% injected lesions were above 0). Conclusions: Imaging biomarkers provide a large number of quantitative image features with a wide span of information, from size to heterogeneity of the tissue which may be indicator of tumor progression and immune infiltrate. In the analysis of radiomics features, delta GLRLM low grey-level zone emphasis was sensitive to the tumoral changes happening in injected lesions at week 8. This might add insights into the imaging-based evaluation of immune infiltration in intratumoral immunotherapy and the creation of associated imaging biomarkers panels. Citation Format: Paula Moreno, Philippe Saiag, Luis de la Cruz Merino, Caroline Dutriaux, Eduardo Castanon Álvarez, Caroline Robert, Juan F. Rodríguez-Moreno, Pablo Cerezuela-Fuentes, Ana M. Arance, Henry Montaudié, Miguel F. Sanmamed, María González Cao, María Pilar López Criado, Julie Charles, Alfonso Berrocal, Enrique de Miguel, Elisa Funk-Brentano, Pablo Sau Llanas, Sorilla Prey, Eva Muñoz-Couselo, Delvys Rodríguez Abreu, Juan Martin-Liberal, Ángel Alberich-Bayarri, Javier Sánchez López, Sonia Macia, Marisol Quintero, Marya F. Chaney, Stéphane Dalle, Iván Márquez-Rodas. Radiomic features in tumor assessment, preliminary results from a phase 2 study of intratumoral administration of BO-112 with pembrolizumab in patients with anti PD1 refractory melanoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr CT109.

Two-Year Event-Free Survival Prediction in DLBCL Patients Based on In Vivo Radiomics and Clinical Parameters.

PurposeFor the identification of high-risk patients in diffuse large B-cell lymphoma (DLBCL), we investigated the prognostic significance of in vivo radiomics derived from baseline [18F]FDG PET/CT and clinical parameters.MethodsPre-treatment [18F]FDG PET/CT scans of 85 patients diagnosed with DLBCL were assessed. The scans were carried out in two clinical centers. Two-year event-free survival (EFS) was defined. After delineation of lymphoma lesions, conventional PET parameters and in vivo radiomics were extracted. For 2-year EFS prognosis assessment, the Center 1 dataset was utilized as the training set and underwent automated machine learning analysis. The dataset of Center 2 was utilized as an independent test set to validate the established predictive model built by the dataset of Center 1.ResultsThe automated machine learning analysis of the Center 1 dataset revealed that the most important features for building 2-year EFS are as follows: max diameter, neighbor gray tone difference matrix (NGTDM) busyness, total lesion glycolysis, total metabolic tumor volume, and NGTDM coarseness. The predictive model built on the Center 1 dataset yielded 79% sensitivity, 83% specificity, 69% positive predictive value, 89% negative predictive value, and 0.85 AUC by evaluating the Center 2 dataset.ConclusionBased on our dual-center retrospective analysis, predicting 2-year EFS built on imaging features is feasible by utilizing high-performance automated machine learning.

Label-free coronavirus disease 2019 lesion segmentation based on synthetic healthy lung image subtraction.

PurposeCoronavirus disease 2019 (COVID‐19) has become a global pandemic and is still posing a severe health risk to the public. Accurate and efficient segmentation of pneumonia lesions in computed tomography (CT) scans is vital for treatment decision‐making. We proposed a novel unsupervised approach using a cycle consistent generative adversarial network (cycle‐GAN) which automates and accelerates the process of lesion delineation.MethodThe workflow includes lung volume segmentation, healthy lung image synthesis, infected and healthy image subtraction, and binary lesion mask generation. The lung volume was first delineated using a pre‐trained U‐net and worked as the input for the following network. A cycle‐GAN was trained to generate synthetic healthy lung CT images from infected lung images. After that, the pneumonia lesions were extracted by subtracting the synthetic healthy lung CT images from the infected lung CT images. A median filter and k‐means clustering were then applied to contour the lesions. The auto segmentation approach was validated on three different datasets.ResultsThe average Dice coefficient reached 0.666 ± 0.178 on the three datasets. Especially, the dice reached 0.748 ± 0.121 and 0.730 ± 0.095, respectively, on two public datasets Coronacases and Radiopedia. Meanwhile, the average precision and sensitivity for lesion segmentation on the three datasets were 0.679 ± 0.244 and 0.756 ± 0.162. The performance is comparable to existing supervised segmentation networks and outperforms unsupervised ones.ConclusionThe proposed label‐free segmentation method achieved high accuracy and efficiency in automatic COVID‐19 lesion delineation. The segmentation result can serve as a baseline for further manual modification and a quality assurance tool for lesion diagnosis. Furthermore, due to its unsupervised nature, the result is not influenced by physicians’ experience which otherwise is crucial for supervised methods.

Image reconstruction using small-voxel size improves small lesion detection for positron emission tomography.

BackgroundPET/CT imaging is widely used in oncology and provides both metabolic and anatomic information. Because of the relatively poor spatial resolution of PET, the detection of small lesions is limited. The low spatial resolution introduces the partial-volume effect (PVE) which negatively affects images both qualitatively and quantitatively. The aim of the study was to investigate the effect of small-voxel (2 mm in-line pixel size) vs. standard-voxel (4 mm in-line pixel size) reconstruction on lesion detection and image quality in a range of activity ratios.Materials and methodsThe National Electrical Manufacturers Association (NEMA) body phantom and the Micro Hollow-Sphere phantom spheres were filled with a solution of [18F]fluorodeoxyglucose ([18F]FDG) in sphere-to-background ratios of 2:1, 3:1, 4:1 and 8:1. In all images reconstructed with 2 mm and 4 mm in-line pixel size the visual lesion delineation, contrast recovery coefficient (CRC) and contrast-to-noise ratio (CNR) were evaluated.ResultsFor smaller (≤ 13 mm) phantom spheres, significantly higher CRC and CNR using small-voxel reconstructions were found, also improving visual lesion delineation. CRC did not differ significantly for larger (≥ 17 mm) spheres using 2 mm and 4 mm in-line pixel size, but CNR was significantly lower; however, lower CNR did not affect visual lesion delineation.ConclusionsSmall-voxel reconstruction consistently improves precise small lesion delineation, lesion contrast and image quality.

Neuroanatomic Markers of Posttraumatic Epilepsy Based on MR Imaging and Machine Learning.

Although posttraumatic epilepsy is a common complication of traumatic brain injury, the relationship between these conditions is unclear and early posttraumatic epilepsy detection and prevention remain major unmet clinical challenges. This study aimed to identify imaging biomarkers that predict posttraumatic epilepsy among survivors of traumatic brain injury on the basis of an MR imaging data set. We performed tensor-based morphometry to analyze brain-shape changes associated with traumatic brain injury and to derive imaging features for statistical group comparison. Additionally, machine learning was used to identify structural anomalies associated with brain lesions. Automatically generated brain lesion maps were used to identify brain regions where lesion load may indicate an increased incidence of posttraumatic epilepsy. We used 138 non-posttraumatic epilepsy subjects for training the machine learning method. Validation of lesion delineation was performed on 15 subjects. Group analysis of the relationship between traumatic brain injury and posttraumatic epilepsy was performed on an independent set of 74 subjects (37 subjects with and 37 randomly selected subjects without epilepsy). We observed significant F-statistics related to tensor-based morphometry analysis at voxels close to the pial surface, which may indicate group differences in the locations of edema, hematoma, or hemorrhage. The results of the F-test on lesion data showed significant differences between groups in both the left and right temporal lobes. We also saw significant differences in the right occipital lobe and cerebellum. Statistical analysis suggests that lesions in the temporal lobes, cerebellum, and the right occipital lobe are associated with an increased posttraumatic epilepsy incidence.

A168 UTILITY OF FOLLOW-UP PANCOLONIC DYE SPRAY CHROMOENDOSCOPY FOR THE DETECTION AND TREATMENT OF DYSPLASTIC LESIONS IDENTIFIED DURING HIGH-DEFINITION WHITE-LIGHT ENDOSCOPY IN PATIENTS WITH COLONIC INFLAMMATORY BOWEL DISEASE

Abstract Background Patients with inflammatory bowel disease (IBD) involving the colorectum are at higher risk for the development of flat and poorly delineated dysplastic lesions. Pancolonic dye spray chromoendoscopy (DCE) can represent an adjunct to white light endoscopy (WLE) that enhances detection and delineation of such lesions. Aims We evaluated the utility of follow-up DCE for lesion detection and treatment in patients with IBD who had visible (polypoid or flat lesions) or invisible (identified in non-targeted biopsies) dysplasia detected during high-definition (HD)-WLE. Methods We retrospectively studied patients with colonic IBD from The Ottawa Hospital who underwent DCE by a trained endoscopist, as follow-up for dysplasia detected during HD-WLE, over a 7-year time period (2013–2020). We collected demographic and disease-specific variables. Results Twenty-four patients were included (mean age 56.7±13.8 years, 50.0% male, 70.8% ulcerative colitis, mean disease duration 18.0±11.0 years, 70.8% moderate or severe disease activity historically, and 41.7% remote history of dysplasia). Seventeen (70.8%) patients were referred following detection of invisible dysplasia; the remainder (29.2%) were referred for surveillance of poorly defined visible dysplastic lesions. For those referred following detection of invisible dysplasia, DCE identified visible dysplasia at the same site in 8/17 (47.1%) patients, at a different site in 6/17 (35.3%) patients, and no dysplasia in 3/17 (17.6%) patients. DCE identified 1.39±1.50 new visible dysplastic lesions per patient that were not identified on index HD-WLE. DCE upgraded the highest grade of dysplasia and diagnosed colorectal cancer in 2/24 (8.3%) patients, which was missed on index HD-WLE. Endoscopic resection was successful in 23/34 (76.7%) dysplastic lesions identified on DCE. However, DCE was unable to facilitate lesion resection in three patients due to advanced lesion characteristics. Follow-up data was available for 17/24 (70.8%) patients (mean 0.88±0.58 years). Of these patients, 7/17 (41.2%) developed subsequent dysplasia, including 4/17 (23.5%) at a site of prior invisible dysplasia, 2/17 (11.8%) at sites of prior visible dysplasia, and 1/17 (5.88%) at a different site. One of these patients developed high-grade dysplasia requiring resection. Conclusions In this study, DCE is a valuable tool for dysplasia detection and treatment in persons with colonic IBD who have ill-defined dysplasia identified during HD-WLE. Given some patients developed dysplasia following DCE in our cohort, it is important to maintain close follow-up and obtain biopsies around sites of prior invisible and visible dysplasia. Further studies are required to validate our findings. Funding Agencies None

Accurate Delineation of Mucosal Lesions in Treatment-Planning Computed Tomography Using Iodine Paste Markers for Oral Mucosal Melanoma

This technical report introduces the utility of iodine paste markers using endodontic materials for the accurate contouring of mucosal lesions of oral mucosal melanoma, which are difficult to delineate on imaging during the planning of carbon-ion radiation therapy (CIRT). The patient had a primary oral mucosal melanoma located in the palatal mucosa without palatal or maxillary bone invasion. A dental root canal filling material, which is a calcium hydroxide/iodoform nonhardenable paste, was used as a marker. We first performed treatment-planning computed tomography (CT) without an iodine paste marker for mucosal lesions. Subsequently, we placed an iodine paste marker on the palatal mucosal lesion to accurately delineate the mucosal lesions of the palate. Finally, we obtained a reference CT image with an iodine paste marker. Computed tomography without the marker was fused to the reference CT with markers during treatment planning, and the gross tumor volume was contoured. Thereafter, CIRT was delivered without markers. During CIRT, expected acute mucositis was observed in the area of the planning target volume, including melanosis, in accordance with the dose distribution. The use of iodine paste markers for localized mucosal lesions, which are difficult to delineate on CT and magnetic resonance imaging, may be useful for accurately contouring gross tumor volumes on treatment-planning CT.

Noninvasive Visualization of Sub-5 mm Orthotopic Hepatic Tumors by a Nanoprobe-Mediated Positive and Reverse Contrast-Balanced Imaging Strategy.

Delineation of small malignant lesions and their vasculature enables early and accurate diagnosis of hepatocellular carcinoma (HCC). However, it remains challenging to identify these features simultaneously by noninvasive imaging technology. Reverse contrast imaging emerges as a powerful means to detect early-stage HCC by taking inspiration from the intrinsic liver phagocytosis toward exogenous agents to generate negative tumor-to-normal tissue signals. However, this mechanism conflicts with the signal-enhancing requirements for vasculature visualization. Here, we solve this conundrum by designing a positive and reverse contrast-balanced imaging strategy based on a multifunctional PEG-Ta2O5@CuS nanoprobe that combines advanced gemstone spectral computer tomography (GSCT) with photoacoustic (PA) imaging. The nanoprobe exhibits preferential accumulation in Kupffer cells and hepatocytes over tumor cells, and its spectral properties are well matched with GSCT, leading to the enhancement of reverse contrast signals that enable clear delineation of 2-4 mm orthotopic HCC lesions. Meanwhile, its strong PA imaging capability at the second near-infrared (NIR-II) window makes vascular evaluation accessible by monitoring the positive signal enhancement derived from the limited tumor accumulation of the nanoprobe. In addition, the nanoprobe enables NIR-II photohyperthermia for timely tumor ablation. Overall, this proposed strategy shows potential in early detection and theranostics of HCC for improved clinical outcomes.

Application of Highly Flexible Adaptive Image Receive Coil for Lung MR Imaging Using Zero TE Sequence: Comparison with Conventional Anterior Array Coil.

(1) Background: Highly flexible adaptive image receive (AIR) coil has become available for clinical use. The present study aimed to evaluate the performance of AIR anterior array coil in lung MR imaging using a zero echo time (ZTE) sequence compared with conventional anterior array (CAA) coil. (2) Methods: Sixty-six patients who underwent lung MR imaging using both AIR coil (ZTE-AIR) and CAA coil (ZTE-CAA) were enrolled. Image quality of ZTE-AIR and ZTE-CAA was quantified by calculating blur metric value, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) of lung parenchyma. Image quality was qualitatively assessed by two independent radiologists. Lesion detection capabilities for lung nodules and emphysema and/or lung cysts were evaluated. Patients’ comfort levels during examinations were assessed. (3) Results: SNR and CNR of lung parenchyma were higher (both p < 0.001) in ZTE-AIR than in ZTE-CAA. Image sharpness was superior in ZTE-AIR (p < 0.001). Subjective image quality assessed by two independent readers was superior (all p < 0.05) in ZTE-AIR. AIR coil was preferred by 64 of 66 patients. ZTE-AIR showed higher (all p < 0.05) sensitivity for sub-centimeter nodules than ZTE-CAA by both readers. ZTE-AIR showed higher (all p < 0.05) sensitivity and accuracy for detecting emphysema and/or cysts than ZTE-CAA by both readers. (4) Conclusions: The use of highly flexible AIR coil in ZTE lung MR imaging can improve image quality and patient comfort. Application of AIR coil in parenchymal imaging has potential for improving delineation of low-density parenchymal lesions and tiny nodules.

Value of ultra-high field MRI in patients with suspected focal epilepsy and negative 3\xa0T MRI (EpiUltraStudy): protocol for a prospective, longitudinal therapeutic study

PurposeResective epilepsy surgery is a well-established, evidence-based treatment option in patients with drug-resistant focal epilepsy. A major predictive factor of good surgical outcome is visualization and delineation of a potential epileptogenic lesion by MRI. However, frequently, these lesions are subtle and may escape detection by conventional MRI (≤ 3 T). MethodsWe present the EpiUltraStudy protocol to address the hypothesis that application of ultra-high field (UHF) MRI increases the rate of detection of structural lesions and functional brain aberrances in patients with drug-resistant focal epilepsy who are candidates for resective epilepsy surgery. Additionally, therapeutic gain will be addressed, testing whether increased lesion detection and tailored resections result in higher rates of seizure freedom 1 year after epilepsy surgery. Sixty patients enroll the study according to the following inclusion criteria: aged ≥ 12 years, diagnosed with drug-resistant focal epilepsy with a suspected epileptogenic focus, negative conventional 3 T MRI during pre-surgical work-up.ResultsAll patients will be evaluated by 7 T MRI; ten patients will undergo an additional 9.4 T MRI exam. Images will be evaluated independently by two neuroradiologists and a neurologist or neurosurgeon. Clinical and UHF MRI will be discussed in the multidisciplinary epilepsy surgery conference. Demographic and epilepsy characteristics, along with postoperative seizure outcome and histopathological evaluation, will be recorded.ConclusionThis protocol was reviewed and approved by the local Institutional Review Board and complies with the Declaration of Helsinki and principles of Good Clinical Practice. Results will be submitted to international peer-reviewed journals and presented at international conferences.Trial registration numberwww.trialregister.nl: NTR7536.

Dynamic susceptibility contrast MRI may contribute in prediction of stereotactic radiosurgery outcome in brain metastases.

BackgroundFollowing stereotactic radiosurgery (SRS), predicting treatment response is not possible at an early stage using structural imaging alone. Hence, the current study aims at investigating whether dynamic susceptibility contrast (DSC)-MRI estimated prior to SRS can provide predictive biomarkers in response to SRS treatment and characterize vascular characteristics of pseudo-progression.MethodsIn this retrospective study, perfusion-weighted DSC-MRI image data acquired with a temporal resolution of 1.45 seconds were collected from 41 patients suffering from brain metastases. Outcome was defined based on lesion volume changes in time (determined on structural images) or death. Motion correction and manual lesion delineation were performed prior to semi-automated, voxel-wise perfusion analysis. Statistical testing was performed using linear regression and a significance threshold at P = .05. Age, sex, primary cancers (pulmonary cancer and melanoma), lesion volume, and dichotomized survival time were added as covariates in the linear regression models (ANOVA).ResultsRelative cerebral blood volume (rCBV) and relative cerebral blood flow (rCBF) were found to be significantly lower prior to SRS treatment in patients with increasing lesion volume or early death post-SRS (P ≤ .01).ConclusionUnfavorable treatment outcome may be linked to low perfusion prior to SRS. Pseudo-progression may be preceded by a transient rCBF increase post-SRS. However, results should be verified in different or larger patient material.

On Unsupervised Methods for Medical Image Segmentation: Investigating Classic Approaches in Breast Cancer DCE-MRI

Unsupervised segmentation techniques, which do not require labeled data for training and can be more easily integrated into the clinical routine, represent a valid solution especially from a clinical feasibility perspective. Indeed, large-scale annotated datasets are not always available, undermining their immediate implementation and use in the clinic. Breast cancer is the most common cause of cancer death in women worldwide. In this study, breast lesion delineation in Dynamic Contrast Enhanced MRI (DCE-MRI) series was addressed by means of four popular unsupervised segmentation approaches: Split-and-Merge combined with Region Growing (SMRG), k-means, Fuzzy C-Means (FCM), and spatial FCM (sFCM). They represent well-established pattern recognition techniques that are still widely used in clinical research. Starting from the basic versions of these segmentation approaches, during our analysis, we identified the shortcomings of each of them, proposing improved versions, as well as developing ad hoc pre- and post-processing steps. The obtained experimental results, in terms of area-based—namely, Dice Index (DI), Jaccard Index (JI), Sensitivity, Specificity, False Positive Ratio (FPR), False Negative Ratio (FNR)—and distance-based metrics—Mean Absolute Distance (MAD), Maximum Distance (MaxD), Hausdorff Distance (HD)—encourage the use of unsupervised machine learning techniques in medical image segmentation. In particular, fuzzy clustering approaches (namely, FCM and sFCM) achieved the best performance. In fact, for area-based metrics, they obtained DI = 78.23% ± 6.50 (sFCM), JI = 65.90% ± 8.14 (sFCM), sensitivity = 77.84% ± 8.72 (FCM), specificity = 87.10% ± 8.24 (sFCM), FPR = 0.14 ± 0.12 (sFCM), and FNR = 0.22 ± 0.09 (sFCM). Concerning distance-based metrics, they obtained MAD = 1.37 ± 0.90 (sFCM), MaxD = 4.04 ± 2.87 (sFCM), and HD = 2.21 ± 0.43 (FCM). These experimental findings suggest that further research would be useful for advanced fuzzy logic techniques specifically tailored to medical image segmentation.

Effects of Tracer Uptake Time in Non-Small Cell Lung Cancer 18F-FDG PET Radiomics.

PET radiomics applied to oncology allow the measurement of intratumoral heterogeneity. This quantification can be affected by image protocols; hence, there is an increased interest in understanding how radiomic expression on PET images is affected by different imaging conditions. To address that interest, this study explored how radiomic features are affected by changes in 18F-FDG uptake time, image reconstruction, lesion delineation, and radiomic binning settings. Methods: Ten non-small cell lung cancer patients underwent 18F-FDG PET on 2 consecutive days. On each day, scans were obtained at 60 and 90 min after injection and reconstructed following EARL version 1 and with point-spread-function resolution modeling (PSF-EARL2). Lesions were delineated with an SUV threshold of 4.0, with 40% of SUVmax, and with a contrast-based isocontour. PET image intensity was discretized with both a fixed bin width (FBW) and a fixed bin number before the calculation of the radiomic features. Repeatability of features was measured with the intraclass correlation coefficient, and the change in feature value over time was calculated as a function of its repeatability. Features were then classified into use-case scenarios based on their repeatability and susceptibility to tracer uptake time. Results: With PSF-EARL2 reconstruction, 40% of SUVmax lesion delineation, and FBW intensity discretization, most features (94%) were repeatable at both uptake times (intraclass correlation coefficient > 0.9), 35% being classified for dual-time-point use cases as being sensitive to changes in uptake time, 39% were classified for cross-sectional studies with an unclear dependency on time, 20% were classified for cross-sectional use while being robust to uptake time changes, and 6% were discarded for poor repeatability. EARL version 1 images had 1 fewer repeatable feature (neighborhood gray-level different matrix coarseness) than PSF-EARL2; the contrast-based delineation had the poorest repeatability of the delineation methods, with 45% of features being discarded; and fixed bin number resulted in lower repeatability than FBW (45% and 6% of features were discarded, respectively). Conclusion: Repeatability was maximized with PSF-EARL2 reconstruction, lesion delineation at 40% of SUVmax, and FBW intensity discretization. On the basis of their susceptibility to uptake time, radiomic features were classified into specific non-small cell lung cancer PET radiomics use cases.

Object Detection Improves Tumour Segmentation in MR Images of Rare Brain Tumours.

Simple SummaryThis study evaluates the impact of adding an object detection framework into brain tumour segmentation models, especially when the models are applied to different domains. In recent years, multiple models have been successfully applied to brain tumour segmentation tasks. However, the performance and stability of these models have never been evaluated when the training and target domain differ. In this study, we identify object detection as a simpler problem that can be injected into a segmentation model as an a priori, and which can increase the performance of our models. We propose an automatic segmentation model that, without model retraining or adaptation, showed good results when applied to a rare brain tumour.Tumour lesion segmentation is a key step to study and characterise cancer from MR neuroradiological images. Presently, numerous deep learning segmentation architectures have been shown to perform well on the specific tumour type they are trained on (e.g., glioblastoma in brain hemispheres). However, a high performing network heavily trained on a given tumour type may perform poorly on a rare tumour type for which no labelled cases allows training or transfer learning. Yet, because some visual similarities exist nevertheless between common and rare tumours, in the lesion and around it, one may split the problem into two steps: object detection and segmentation. For each step, trained networks on common lesions could be used on rare ones following a domain adaptation scheme without extra fine-tuning. This work proposes a resilient tumour lesion delineation strategy, based on the combination of established elementary networks that achieve detection and segmentation. Our strategy allowed us to achieve robust segmentation inference on a rare tumour located in an unseen tumour context region during training. As an example of a rare tumour, Diffuse Intrinsic Pontine Glioma (DIPG), we achieve an average dice score of 0.62 without further training or network architecture adaptation.

The role of unenhanced phase of the liver in the scanning protocol of metastatic breast cancer: implications for sensitivity, response evaluation and size measurement.

BackgroundTo analyse if performing unenhanced CT of the liver aids in the evaluation of metastatic lesions, response assessment or alter the size of the lesions, compared with portal phase alone, in patients with hepatic metastases from breast carcinoma.Patients and methodsOne-hundred and fifty-three CT scans of 36 women were included. Scans consisted of unenhanced, arterial and portal delayed phases of the liver. Two readers sorted which phase was best for visualization of metastases, evaluated the number of lesions detected in each phase, selected the best phase for assessment of response in two consecutive scans, and measured one target lesion in all the phases. Χ2 was used to compare differences among phases and paired t test for measurement differences.ResultsUnenhanced, arterial and portal phases were considered better phases by readers 1/2 in 68/67%, 27/28% and 69/70%, and some lesions were missed in 2%, 11% and 7%, respectively. Sensitivity was significantly better for unenhanced and portal phases compared to arterial phase. Comparison between consecutive scans was considered better in unenhanced (80/79%), followed by portal (70/69%) and arterial phases (31/31%). Maximum diameter of target lesions was 15% greater in unenhanced phase (p < 0.001).ConclusionsPortal and unenhanced phases of the liver allow better detection and delineation of metastatic hepatic lesions from breast carcinoma. In most cases, unenhanced CT is the best phase to assess response and provides the largest diameter. Therefore, we recommend the use of unenhanced CT in the evaluation of patients with breast carcinoma and suspected or known hepatic metastatic disease.

Abstract 17074: Primary Cardiac Sarcoma In Elderly Patient And Need For High Index Of Suspicion

Case Presentation: Primary cardiac tumors are rare with an incidence between 0.0017 and 0.2%. Most cardiac tumors are benign such as myxomas, lipomas, and papillary fibroelastomas. Primary malignant cardiac tumors are even more uncommon, with cardiac sarcomas being the most common. We present a case of an 87-year old female who presented initially for progressive shortness of breath and concern for potential endocarditis vs intra-cardiac thrombus noted on outside hospital transthoracic echocardiogram. Blood cultures were negative, no leukocytosis or sequelae of infection were noted. Transesophageal echocardiogram (TEE) noted a large lesion attached to the anterior mitral leaflet, with additional lesions noted on the left atrial free wall just medial to the right sided pulmonary veins. Obstruction of mitral inflow was noted with a mean gradient through the mitral valve of 5mmHg. Cardiac MRI was performed for further delineation of lesions and based on tissue characteristics were felt to be thrombus (Figure 1). Hypercoagulability workup was negative, and there was no improvement in lesion size with systemic anticoagulation. Cardiothoracic surgery was consulted for surgical removal of the mass. Pre-operative cardiac catheterization showed multi-vessel obstructive coronary artery disease. It was also noted that the left circumflex artery appeared to supply blood into a discrete area in the left atrium. The patient was ultimately taken for surgical removal. Surgical pathology demonstrated primary high grade cardiac sarcoma. Discussion: The rarity and heterogeneity of primary cardiac tumors, along with many potential mimickers pose a diagnostic and therapeutic challenge. This, combined with the location and aggressive nature of the tumors, make them difficult to treat and incur a high mortality rate. This highlights the importance of maintaining a high index of suspicion and keeping a broad differential when evaluating intra-cardiac masses.

Fuhrman nuclear grade prediction of clear cell renal cell carcinoma: influence of volume of interest delineation strategies on machine learning-based dynamic enhanced CT radiomics analysis.

To investigate the influence of different volume of interest (VOI) delineation strategies on machine learning-based predictive models for discrimination between low and high nuclear grade clear cell renal cell carcinoma (ccRCC) on dynamic contrast-enhanced CT. This study retrospectively collected 177 patients with pathologically proven ccRCC (124 low-grade; 53 high-grade). Tumor VOI was manually segmented, followed by artificially introducing uncertainties as: (i) contour-focused VOI, (ii) margin erosion of 2 or 4mm, and (iii) margin dilation (2, 4, or 6mm) inclusive of perirenal fat, peritumoral renal parenchyma, or both. Radiomics features were extracted from four-phase CT images (unenhanced phase (UP), corticomedullary phase (CMP), nephrographic phase (NP), excretory phase (EP)). Different combinations of four-phasic features for each VOI delineation strategy were used to build 176 classification models. The best VOI delineation strategy and superior CT phase were identified and the top-ranked features were analyzed. Features extracted from UP and EP outperformed features from other single/combined phase(s). Shape features and first-order statistics features exhibited superior discrimination. The best performance (ACC 81%, SEN 67%, SPE 87%, AUC 0.87) was achieved with radiomics features extracted from UP and EP based on contour-focused VOI. Shape and first-order features extracted from UP + EP images are better feature representations. Contour-focused VOI erosion by 2mm or dilation by 4mm within peritumor renal parenchyma exerts limited impact on discriminative performance. It provides a reference for segmentation tolerance in radiomics-based modeling for ccRCC nuclear grading. • Lesion delineation uncertainties are tolerated within a VOI erosion range of 2mm or dilation range of 4mm within peritumor renal parenchyma for radiomics-based ccRCC nuclear grading. • Radiomics features extracted from unenhanced phase and excretory phase are superior to other single/combined phase(s) at differentiating high vs low nuclear grade ccRCC. • Shape features and first-order statistics features showed superior discriminative capability compared to texture features.

PDMS-Parylene Hybrid, Flexible Micro-ECoG Electrode Array for Spatiotemporal Mapping of Epileptic Electrophysiological Activity from Multicortical Brain Regions.

Epilepsy detection and focus location are urgent issues that need to be solved in epilepsy research. A cortex conformable and fine spatial accuracy electrocorticogram (ECoG) sensor array, especially for real-time detection of multicortical functional regions and delineating epileptic focus remains a challenge. Here, we fabricated a polydimethylsiloxane (PDMS)-parylene hybrid, flexible micro-ECoG electrode array. The multiwalled carbon nanotubes (MWCNTs)/poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) nanocomposite-modified electrode interface significantly improved the sensing performance with low impedance (20.68 ± 6.65 kΩ), stable phase offset, and high sensitivity. The electrophysiological activities of multicortical brain regions (somatosensory cortex, parietal association cortex, and visual cortex) were simultaneously monitored during normal and epileptic statuses. The epileptic ECoG activities spread spatiotemporally from the starting point toward the adjacent cortex. Significant variations of the waveform, power, and frequency band were observed. The ECoG potential (123 ± 23 μV) at normal status was prominently up to 417 ± 87 μV at the spike wave stage. Besides, the power for epileptic activity (11.049 ± 4.513 μW) was 10 times higher than that (1.092 ± 0.369 μW) for normal activity. In addition, the theta frequency band was found to be a characteristic frequency band of epileptic signals. These joint analysis results of multicortical regions indicated that the active micron-scale region on the parietal association cortex was more likely to be the epileptogenic focus. Cortical mapping with high spatial detail provides the accurate delineation of lesions. The flexible micro-ECoG electrode array is a powerful tool for constructing a spatiotemporal map of the cortex. It provides a technical platform for epileptic focus location, biomedical diagnosis, and brain-computer interaction.