Assessing and improving the validity of COVID-19 autopsy studies - a multi center approach to establish essential standards for immunohistochemical and ultrastructural analyses

Background Autopsy studies have provided valuable insights into the pathophysiology of COVID-19. Controversies remain whether the clinical presentation is due to direct organ damage by SARS-CoV-2 or secondary effects, e.g. by an overshooting immune response. SARS-CoV-2 detection in tissues by RT-qPCR and immunohistochemistry (IHC) or electron microscopy (EM) can help answer these questions, but a comprehensive evaluation of these applications is missing. Methods We assessed publications using IHC and EM for SARS-CoV-2 detection in autopsy tissues. We systematically evaluated commercially available antibodies against the SARS-CoV-2 spike protein and nucleocapsid, dsRNA, and non-structural protein Nsp3 in cultured cell lines and COVID-19 autopsy tissues. In a multicenter study, we evaluated specificity, reproducibility, and inter-observer variability of SARS-CoV-2 nucleocapsid staining. We correlated RT-qPCR viral tissue loads with semiquantitative IHC scoring. We used qualitative and quantitative EM analyses to refine criteria for ultrastructural identification of SARS-CoV-2. Findings Publications show high variability in the detection and interpretation of SARS-CoV-2 abundance in autopsy tissues by IHC or EM. In our study, we show that IHC using antibodies against SARS-CoV-2 nucleocapsid yields the highest sensitivity and specificity. We found a positive correlation between presence of viral proteins by IHC and RT-qPCR-determined SARS-CoV-2 viral RNA load (r=-0.83, p-value <0.0001). For EM, we refined criteria for virus identification and also provide recommendations for optimized sampling and analysis. 116 of 122 publications misinterpret cellular structures as virus using EM or show only insufficient data. We provide publicly accessible digitized EM and IHC sections as a reference and for training purposes. Interpretation Since detection of SARS-CoV-2 in human autopsy tissues by IHC and EM is difficult and frequently incorrect, we propose criteria for a re-evaluation of available data and guidance for further investigations of direct organ effects by SARS-CoV-2.

Exceptionally rapid recurrence of malignant meningioma: case report

Anaplastic meningioma represents less than 5% of all meningiomas. It is a neoplasm with a poor prognosis due to aggressiveness and a high rate of recurrence. Patients could remain asymptomatic but clinical characteristics of mass effect are the most common presentation. Although diagnosis is made with histological study, this method is difficult to define, with inter-observer variability. When possible, surgical resection is the primary management. We discuss a case of an adult female patient with tonic-clonic seizures and weakness attributed to an anaplastic meningioma in the occipital lobe. The patient was treated with a parietal craniotomy with complete resection. One month later the patient suffered a recurrence of the tumor with the need for further intervention with incomplete resection. Due to extent of the damage the patient deceased two weeks later.

VARIATIONS IN ACCESSIBILITY OF ROUND WINDOW VIA POSTERIOR TYMPANOTOMY APPROACH IN COCHLEAR IMPLANT SURGERY

Objective: To apply the St Thomas’ Hospital (STH) classification of round window type, in a Pakistani pediatric population undergoing cochlear implantation, and rate the inter observer variability of applying this classification. Study Design: Cross sectional study. Place and Duration of Study: Combined Military Hospital Rawalpindi, from Apr 2019 to Dec 2020. Methodology: Patients were examined per-operatively by a panel of four surgeons after "optimal" posterior tympanotomy for round window variations, as per STH classification of approachability of RWM. The observations of the four surgeons were recorded and interobserver variation was assessed and analyzed. Results: A total of 100 patients were operated, 45 females and 55 males. Mean age was 3.8 years. There was minimal inter observer variability with regards to round window type and extent of "optimal" posterior tympanotomy. Three patients had type I, 76 had type IIA, 15 had type IIB and 6 patients had type III. Round window insertion/membranous cochleostomy was possible in 70 patients, whereas the rest require extended round window approach or bony cochleostomy. Conclusion: The STH classification is a useful predictor of route of CI electrode insertion and most patients can undergo RW insertion with confidence based on minimal variation between surgeons when applying the STH classification as well as when deciding the extent of surgical exposure.

Prospectively-validated deep learning model for segmenting swallowing and chewing structures in CT

ObjectiveDelineating swallowing and chewing structures aids in radiotherapy (RT) treatment planning to limit dysphagia, trismus, and speech dysfunction. We aim to develop an accurate and efficient method to automate this process.ApproachCT scans of 242 head and neck (H&N) cancer patients acquired from 2004-2009 at our institution were used to develop auto-segmentation models for the masseters, medial pterygoids, larynx, and pharyngeal constrictor muscle using DeepLabV3+. A cascaded architecture was used, wherein models were trained sequentially to spatially constrain each structure group based on prior segmentations. Additionally, an ensemble of models, combining contextual information from axial, coronal, and sagittal views was used to improve segmentation accuracy. Prospective evaluation was conducted by measuring the amount of manual editing required in 91 H&N CT scans acquired February-May 2021.Main resultsMedians and inter-quartile ranges of Dice Similarity Coefficients (DSC) computed on the retrospective testing set (N=24) were 0.87 (0.85-0.89) for the masseters, 0.80 (0.79- 0.81) for the medial pterygoids, 0.81 (0.79-0.84) for the larynx, and 0.69 (0.67-0.71) for the constrictor. Auto-segmentations, when compared to inter-observer variability in 10 randomly selected scans, showed better agreement (DSC) with each observer as compared to inter-observer DSC. Prospective analysis showed most manual modifications needed for clinical use were minor, suggesting auto-contouring could increase clinical efficiency. Trained segmentation models are available for research use upon request via https://github.com/cerr/CERR/wiki/Auto-Segmentation-models.SignificanceWe developed deep learning-based auto-segmentation models for swallowing and chewing structures in CT and demonstrated its potential for use in treatment planning to limit complications post-RT. To the best of our knowledge, this is the only prospectively-validated deep learning-based model for segmenting chewing and swallowing structures in CT. Additionally, the segmentation models have been made open-source to facilitate reproducibility and multi-institutional research.

MVnet: automated time-resolved tracking of the mitral valve plane in CMR long-axis cine images with residual neural networks: a multi-center, multi-vendor study

Background Mitral annular plane systolic excursion (MAPSE) and left ventricular (LV) early diastolic velocity (e’) are key metrics of systolic and diastolic function, but not often measured by cardiovascular magnetic resonance (CMR). Its derivation is possible with manual, precise annotation of the mitral valve (MV) insertion points along the cardiac cycle in both two and four-chamber long-axis cines, but this process is highly time-consuming, laborious, and prone to errors. A fully automated, consistent, fast, and accurate method for MV plane tracking is lacking. In this study, we propose MVnet, a deep learning approach for MV point localization and tracking capable of deriving such clinical metrics comparable to human expert-level performance, and validated it in a multi-vendor, multi-center clinical population. Methods The proposed pipeline first performs a coarse MV point annotation in a given cine accurately enough to apply an automated linear transformation task, which standardizes the size, cropping, resolution, and heart orientation, and second, tracks the MV points with high accuracy. The model was trained and evaluated on 38,854 cine images from 703 patients with diverse cardiovascular conditions, scanned on equipment from 3 main vendors, 16 centers, and 7 countries, and manually annotated by 10 observers. Agreement was assessed by the intra-class correlation coefficient (ICC) for both clinical metrics and by the distance error in the MV plane displacement. For inter-observer variability analysis, an additional pair of observers performed manual annotations in a randomly chosen set of 50 patients. Results MVnet achieved a fast segmentation (<1 s/cine) with excellent ICCs of 0.94 (MAPSE) and 0.93 (LV e’) and a MV plane tracking error of −0.10 ± 0.97 mm. In a similar manner, the inter-observer variability analysis yielded ICCs of 0.95 and 0.89 and a tracking error of −0.15 ± 1.18 mm, respectively. Conclusion A dual-stage deep learning approach for automated annotation of MV points for systolic and diastolic evaluation in CMR long-axis cine images was developed. The method is able to carefully track these points with high accuracy and in a timely manner. This will improve the feasibility of CMR methods which rely on valve tracking and increase their utility in a clinical setting.

Investigation of Autosegmentation Techniques on T2-Weighted MRI for Off-line Dose Reconstruction in MR-Linac Adapt to Position Workflow for Head and Neck Cancers

Purpose: In order to accurately accumulate delivered dose for head and neck cancer patients treated with the Adapt to Position workflow on the 1.5T magnetic resonance imaging (MRI)-linear accelerator (MR-linac), the low-resolution T2-weighted MRIs used for daily setup must be segmented to enable reconstruction of the delivered dose at each fraction. In this study, our goal is to evaluate various autosegmentation methods for head and neck organs at risk (OARs) on on-board setup MRIs from the MR-linac for off-line reconstruction of delivered dose. Methods: Seven OARs (parotid glands, submandibular glands, mandible, spinal cord, and brainstem) were contoured on 43 images by seven observers each. Ground truth contours were generated using a simultaneous truth and performance level estimation (STAPLE) algorithm. 20 autosegmentation methods were evaluated in ADMIRE: 1-9) atlas-based autosegmentation using a population atlas library (PAL) of 5/10/15 patients with STAPLE, patch fusion (PF), random forest (RF) for label fusion; 10-19) autosegmentation using images from a patient's 1-4 prior fractions (individualized patient prior (IPP)) using STAPLE/PF/RF; 20) deep learning (DL) (3D ResUNet trained on 43 ground truth structure sets plus 45 contoured by one observer). Execution time was measured for each method. Autosegmented structures were compared to ground truth structures using the Dice similarity coefficient, mean surface distance, Hausdorff distance, and Jaccard index. For each metric and OAR, performance was compared to the inter-observer variability using Dunn's test with control. Methods were compared pairwise using the Steel-Dwass test for each metric pooled across all OARs. Further dosimetric analysis was performed on three high-performing autosegmentation methods (DL, IPP with RF and 4 fractions (IPP_RF_4), IPP with 1 fraction (IPP_1)), and one low-performing (PAL with STAPLE and 5 atlases (PAL_ST_5)). For five patients, delivered doses from clinical plans were recalculated on setup images with ground truth and autosegmented structure sets. Differences in maximum and mean dose to each structure between the ground truth and autosegmented structures were calculated and correlated with geometric metrics. Results: DL and IPP methods performed best overall, all significantly outperforming inter-observer variability and with no significant difference between methods in pairwise comparison. PAL methods performed worst overall; most were not significantly different from the inter-observer variability or from each other. DL was the fastest method (33 seconds per case) and PAL methods the slowest (3.7 - 13.8 minutes per case). Execution time increased with number of prior fractions/atlases for IPP and PAL. For DL, IPP_1, and IPP_RF_4, the majority (95%) of dose differences were within 250 cGy from ground truth, but outlier differences up to 785 cGy occurred. Dose differences were much higher for PAL_ST_5, with outlier differences up to 1920 cGy. Dose differences showed weak but significant correlations with all geometric metrics (R2 between 0.030 and 0.314). Conclusions: The autosegmentation methods offering the best combination of performance and execution time are DL and IPP_1. Dose reconstruction on on-board T2-weighted MRIs is feasible with autosegmented structures with minimal dosimetric variation from ground truth, but contours should be visually inspected prior to dose reconstruction in an end-to-end dose accumulation workflow.

Reproducibility of non-invasive VT substrate characterization: finding the same paths

Background Delayed enhancement gadolinium MRI is a useful technique to identify myocardial scar. The objective of this study is compare the reproducibility of the scar quantification and characterization based on cardiac MRI. Methods 10 patients with ischemic ethology underwent to 1,5T DE-MRI acquisition for myocardial scar analysis. Images were processed using a commercial software (ADAS3D-Galgo Medical) and different parameters from scar tissue (mass of the scar, core of scar and border zone expressed in grams) were analysed. Conducting channels evaluation was obtained by the number of corridors and the mass of the border zone of those corridors. To perform this analysis, 2 experienced and 1 non experienced users segmented DE-MRI acquisition in order to evaluate the inter observer variability. Bland-Altman analysis was employed to evaluate the comparison between the measurements. Results Inter observer agreement between experienced users was high (table). The mean and the standard deviation of the differences between two measurements for the scar mass was −3,9±14,66 gr. Analysing the scar tissue divided in core and border zone, the mass of these volume tissues were very similar (−3,51±4,56gr and −0,4±12,87gr respectively. Regarding conducting channels characteristics, the mean of the differences was 0±2 for the number of channels and 1,71±7,76 gr for the mass on the border zone of the corridors. Comparing the measurements between one of the experienced users and the beginner user, results were similar but significant differences were found on the mass of the core and the number of channels, with a variability of ±2 channels (table). Conclusions Left ventricular scar size and characteristics derived from late gadolinium enhanced post-processed images are highly reproducible between experienced observers. FUNDunding Acknowledgement Type of funding sources: None. Table 1 Scar analysis performed by 3 users