Panda, Ananya

PurposeTo evaluate multicenter repeatability and reproducibility of T1 and T2 maps generated using MR fingerprinting (MRF) in the International Society for Magnetic Resonance in Medicine/National Institute of Standards and Technology MRI system phantom and in prostatic tissues.MethodsMRF experiments were performed on 5 different 3 Tesla MRI scanners at 3 different institutions: University Hospitals Cleveland Medical Center (Cleveland, OH), Brigham and Women's Hospital (Boston, MA) in the United States, and Diagnosticos da America (Rio de Janeiro, RJ) in Brazil. Raw MRF data were reconstructed using a Gadgetron‐based MRF online reconstruction pipeline to yield quantitative T1 and T2 maps. The repeatability of T1 and T2 values over 6 measurements in the International Society for Magnetic Resonance in Medicine/National Institute of Standards and Technology MRI system phantom was assessed to demonstrate intrascanner variation. The reproducibility between the 4 clinical scanners was assessed to demonstrate interscanner variation. The same‐day test–retest normal prostate mean T1 and T2 values from peripheral zone and transitional zone were also compared using the intraclass correlation coefficient and Bland–Altman analysis.ResultsThe intrascanner variation of values measured using MRF was less than 2% for T1 and 4.7% for T2 for relaxation values, within the range of 307.7 to 2360 ms for T1 and 19.1 to 248.5 ms for T2. Interscanner measurements showed that the T1 variation was less than 4.9%, and T2 variation was less than 8.1% between multicenter scanners. Both T1 and T2 values in in vivo prostatic tissue demonstrated high test–retest reliability (intraclass correlation coefficient > 0.92) and strong linear correlation (R2 > 0.840).ConclusionProstate MRF measurements of T1 and T2 are repeatable and reproducible between MRI scanners at different centers on different continents for the above measurement ranges.

Abstract Purpose: Around 30% of PDAC less than 2-cm tend to go undetected on CT due to their subtle imaging signatures. Automated detection of PDAC using AI represents an opportunity to augment physician expertise and to improve outcomes through early detection of PDAC. Our purpose was to develop a 3D-CNN for fully automated detection of PDAC and to further evaluate the impact of inclusion of pancreas segmentation on the accuracy of this 3D-CNN. Methods: A Medical Imaging Data Readiness Scale (MIDaR) level A dataset (portal venous phase CTs, slice thickness ≤ 3.75 mm) of 466 treatment-naïve biopsy-proven PDAC and 1994 subjects with normal pancreas was created after exclusion of CTs with suboptimal image quality or biliary stents. Volumetric pancreas and tumor segmentations on CTs were done by two radiologists using 3D Slicer. A total of 370 CTs with PDAC and 370 CTs with normal pancreas were randomly selected for separate training and validation sets, and 396 CTs (96 CTs with PDAC and 300 CTs with normal pancreas) were utilized for testing. Two separate 3D-CNNs were trained. A three-stage bounding-box-only model (A): stage 1 was based on a UNET-like architecture and localized the pancreas on CT with a bounding box; stage 2 utilized an Inception ResNet architecture and classified each slice through the pancreas into PDAC vs. normal; and stage 3 utilized the output of stage 2 to generate final classification for a given CT. Conversely, a four-stage pancreas segmentation-based model (B) included stage 1 of model A followed by an additional stage of automated pancreas and tumor segmentation (stage 2), classification of each slice through the pancreas into PDAC vs. normal (stage 3) and, finally, generation of final classification score (stage 4) for a given CT. Area under the receiver operating characteristic curve (AUROC) of the two models were compared on the test set. Results: Mean (SD) PDAC diameter in the test set was 1.1 (0.43) cm. Model A (three-stage bounding-box-only) correctly classified 305 (77%) out of 396 CTs from the test set into PDAC vs. normal. It incorrectly classified 12/96 (12.5%) CTs with PDAC as normal and 79/300 (26%) normal CTs as PDAC. AUROC for model A was 0.85. Model B (four-stage pancreas segmentation-based) correctly classified 351 (88%) out of 396 CTs. It incorrectly classified 13/96 (13.5%) CTs with PDAC as normal and 32/300 (10.7%) normal CTs as PDAC. AUROC for model B was 0.94. AUROC for model B was significantly higher than model A (p<0.005). Conclusion: A 3D-CNN can detect small PDAC with high accuracy using automated localization of pancreas with a bounding box without relying on separate pancreas segmentation. Inclusion of an additional automated pancreas segmentation step reduced false positives with consequent incremental gain in the model’s accuracy. Prospective validation and subsequent integration of such models into clinical workflows has the potential to reduce inadvertent errors in detection of subtle or small PDAC on standard-of-care CT scans. Citation Format: Anurima Patra, Korfiatis Panagiotis, Garima Suman, Ananya Panda, Sushil Kumar Garg, Ajit Goenka. Automated detection of pancreatic ductal adenocarcinoma (PDAC) on CT scans using artificial intelligence (AI): Impact of inclusion of automated pancreas segmentation on the accuracy of 3D-convolutional neural network (CNN) [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-084.

PurposeTo develop a two‐stage three‐dimensional (3D) convolutional neural networks (CNNs) for fully automated volumetric segmentation of pancreas on computed tomography (CT) and to further evaluate its performance in the context of intra‐reader and inter‐reader reliability at full dose and reduced radiation dose CTs on a public dataset.MethodsA dataset of 1994 abdomen CT scans (portal venous phase, slice thickness ≤ 3.75‐mm, multiple CT vendors) was curated by two radiologists (R1 and R2) to exclude cases with pancreatic pathology, suboptimal image quality, and image artifacts (n = 77). Remaining 1917 CTs were equally allocated between R1 and R2 for volumetric pancreas segmentation [ground truth (GT)]. This internal dataset was randomly divided into training (n = 1380), validation (n = 248), and test (n = 289) sets for the development of a two‐stage 3D CNN model based on a modified U‐net architecture for automated volumetric pancreas segmentation. Model’s performance for pancreas segmentation and the differences in model‐predicted pancreatic volumes vs GT volumes were compared on the test set. Subsequently, an external dataset from The Cancer Imaging Archive (TCIA) that had CT scans acquired at standard radiation dose and same scans reconstructed at a simulated 25% radiation dose was curated (n = 41). Volumetric pancreas segmentation was done on this TCIA dataset by R1 and R2 independently on the full dose and then at the reduced radiation dose CT images. Intra‐reader and inter‐reader reliability, model’s segmentation performance, and reliability between model‐predicted pancreatic volumes at full vs reduced dose were measured. Finally, model’s performance was tested on the benchmarking National Institute of Health (NIH)‐Pancreas CT (PCT) dataset.ResultsThree‐dimensional CNN had mean (SD) Dice similarity coefficient (DSC): 0.91 (0.03) and average Hausdorff distance of 0.15 (0.09) mm on the test set. Model’s performance was equivalent between males and females (P = 0.08) and across different CT slice thicknesses (P > 0.05) based on noninferiority statistical testing. There was no difference in model‐predicted and GT pancreatic volumes [mean predicted volume 99 cc (31cc); GT volume 101 cc (33 cc), P = 0.33]. Mean pancreatic volume difference was −2.7 cc (percent difference: −2.4% of GT volume) with excellent correlation between model‐predicted and GT volumes [concordance correlation coefficient (CCC)=0.97]. In the external TCIA dataset, the model had higher reliability than R1 and R2 on full vs reduced dose CT scans [model mean (SD) DSC: 0.96 (0.02), CCC = 0.995 vs R1 DSC: 0.83 (0.07), CCC = 0.89, and R2 DSC:0.87 (0.04), CCC = 0.97]. The DSC and volume concordance correlations for R1 vs R2 (inter‐reader reliability) were 0.85 (0.07), CCC = 0.90 at full dose and 0.83 (0.07), CCC = 0.96 at reduced dose datasets. There was good reliability between model and R1 at both full and reduced dose CT [full dose: DSC: 0.81 (0.07), CCC = 0.83 and reduced dose DSC:0.81 (0.08), CCC = 0.87]. Likewise, there was good reliability between model and R2 at both full and reduced dose CT [full dose: DSC: 0.84 (0.05), CCC = 0.89 and reduced dose DSC:0.83(0.06), CCC = 0.89]. There was no difference in model‐predicted and GT pancreatic volume in TCIA dataset (mean predicted volume 96 cc (33); GT pancreatic volume 89 cc (30), p = 0.31). Model had mean (SD) DSC: 0.89 (0.04) (minimum–maximum DSC: 0.79 −0.96) on the NIH‐PCT dataset.ConclusionA 3D CNN developed on the largest dataset of CTs is accurate for fully automated volumetric pancreas segmentation and is generalizable across a wide range of CT slice thicknesses, radiation dose, and patient gender. This 3D CNN offers a scalable tool to leverage biomarkers from pancreas morphometrics and radiomics for pancreatic diseases including for early pancreatic cancer detection.

Objective Our aim was to determine the radiologic and clinicopathologic characteristics of thyroid nodules with focal 68Ga-DOTATATE activity. Methods In this retrospective study of 1927 consecutive 68Ga-DOTATATE PET scans, 85 patients with incidental and nonincidental focal 68Ga-DOTATATE avid thyroid nodules were identified, of which 31 patients with 33 thyroid nodules underwent fine-needle aspiration (FNA) or surgery. These 33 nodules were reviewed for Krenning score and SUVmax of the thyroid nodule, contralateral thyroid lobe and left atrium. Results Cytology/histopathology included 58% (19/33) with benign findings, 18% (6/33) medullary thyroid carcinoma (MTC), 9% (3/33) atypia or follicular lesion of undetermined significance (AUS/FLUS), 9% (3/33) suspicious for follicular neoplasm and Hurthle cell adenoma (SFN/HCA) and 6% (2/33) nondiagnostic cytology. Median serum calcitonin was 1156 pg/mL (range, 460–1828) in MTC and was <5.0 pg/mL (<5.0–12.5) in patients with benign nodules. Nodules had Krenning score of 1, 2 and 3 in 46% (15/33), 27% (9/33) and 27% (9/33). The majority of MTC and AUS/FLUS nodules had a Krenning score of 3, and there was substantial intragroup variation in Krenning score among the benign nodules. The mean SUVmax for the entire cohort was 5.5 ± 2.9 (mean ± SD), and the range was 2.0–13.0. There was overlap in the nodule/contralateral thyroid SUVmax ratios between groups. The MTC and AUS/FLUS nodules tended to have a higher nodule/blood pool SUVmax ratio than the other pathologic groups. Conclusion There was considerable variation in radiologic characteristics among benign thyroid nodules. The ratio of thyroid nodule SUVmax/blood pool SUVmax may be useful to differentiate pathologic groups, but larger studies are needed to investigate this further. Given the potential for malignancy in thyroid nodules with focal 68Ga-DOTATATE activity, further evaluation with serum calcitonin and FNA may be considered. Video Abstract: http://links.lww.com/NMC/A186

Abstract Background: MR fingerprinting (MRF) is a novel imaging method proposed for the diagnosis of Multiple Sclerosis (MS). This study aims to determine if MR Fingerprinting (MRF) relaxometry can differentiate frontal normal appearing white matter (F-NAWM) and splenium in patients diagnosed with MS as compared to controls and to characterize the relaxometry of demyelinating plaques relative to the time of diagnosis.Methods: Three-dimensional (3D) MRF data were acquired on a 3.0T MRI system resulting in isotropic voxels (1x1x1mm3) and a total acquisition time of 4min 38s. Data were collected on 18 subjects paired with 18 controls. Regions of interested were drawn over MRF-derived T1 relaxometry maps encompassing selected MS lesions, F-NAWM and splenium. T1 and T2 relaxometry features from those segmented areas were used to classify MS lesions from F-NAWM and splenium with T-distributed stochastic neighbor embedding algorithms (T-SNE). Partial least squares discriminant analysis (PLS-DA) was performed to discriminate NAWM and Splenium in MS compared with controls. Results: Mean out-of-fold machine learning prediction accuracy for discriminant results between MS patients and controls for F-NAWM was 65% and approached 90% for the splenium. There was significant positive correlation between time since diagnosis and MS lesions mean T2 (p=0.015), minimum T1 (p=0.03) and negative correlation with splenium uniformity (p=0.04). Perfect discrimination (AUC=1) was achieved between selected features from MS lesions and F-NAWM.Conclusions: 3D-MRF has the ability to differentiate between MS and controls based on relaxometry properties from the F-NAWM and splenium. Whole brain coverage allows the assessment of quantitative properties within lesions that provide chronological assessment of the time from MS diagnosis.

Objective Studies evaluating leptomeningeal disease on whole-body 18F-FDG PET are lacking. The purpose was to evaluate PET imaging of leptomeningeal disease and investigate the incremental utility of newer PET reconstructions in leptomeningeal disease. Methods PET imaging of 56 patients with leptomeningeal disease detected initially on MRI (n = 53) or cytopathology (n = 35) were retrospectively reviewed. Regular 3-dimensional iterative reconstruction (3D IR, n = 56) and advanced reconstruction (AdvRecon, n = 41) PET images were evaluated by readers blinded to clinical and MRI findings for uptake involving cauda equina, posterior fossa and spinal cord. Spinal cord uptake pattern was classified as normal (uptake < liver), uptake = liver, conus uptake > liver, conus and cervical cord uptake > liver and multifocal/diffuse uptake > liver. SUVmax ratios of conus/liver, conus/left atrium and conus/cervical cord were compared between 3D IR and AdvRecon datasets. Results Cauda equina uptake was seen in 64% and 78% on 3D IR and AdvRecon; posterior fossa uptake was seen in 52% and 54% on 3D IR and AdvRecon, respectively. Twelve percent had cauda equina or posterior fossa uptake visible only on AdvRecon. On 3D IR, normal spinal cord uptake was most common (27%); on AdvRecon, conus and cervical cord uptake > liver was most common (32%). Seven of 11 patients with normal spinal cord uptake on 3D IR were upgraded to increased uptake on AdvRecon. AdvRecon showed significantly higher conus/liver, conus/blood pool and conus/cervical cord SUVmax ratios (P < 0.0001). Conclusion Abnormal uptake in cauda equina, posterior fossa and spinal cord uptake are visible on FDG PET in leptomeningeal disease with increased conspicuity advanced PET reconstructions.