Significant Interobserver Variability Research Articles

Improved concordance of challenging human epidermal growth factor receptor 2 dual in-situ hybridisation cases with the use of a digital image analysis algorithm in breast cancer.

Accurate assessment of human epidermal growth factor receptor 2 (HER2) expression by HER2 immunohistochemistry and in-situ hybridisation (ISH) is critical for the management of patients with breast cancer. The revised 2018 ASCO/CAP guidelines define 5 groups based on HER2 expression and copy number. Manual pathologist quantification by light microscopy of equivocal and less common HER2 ISH groups (groups 2-4) can be challenging, and there are no data on interobserver variability in reporting of these cases. We sought to determine whether a digital algorithm could improve interobserver variability in the assessment of difficult HER2 ISH cases. HER2 ISH was evaluated in a cohort enriched for less common HER2 patterns using standard light microscopy versus analysis of whole slide images using the Roche uPath HER2 dual ISH image analysis algorithm. Standard microscopy demonstrated significant interobserver variability with a Fleiss's kappa value of 0.471 (fair-moderate agreement) improving to 0.666 (moderate-good) with the use of the algorithm. For HER2 group designation (groups 1-5), there waspoor-moderate reliability between pathologists bymicroscopy [intraclass correlation coefficient (ICC) = 0.526], improving to moderate-good agreement (ICC = 0.763) with the use of the algorithm. In subgroup analysis, the algorithm improved concordance particularly in groups 2, 4 and 5. Time to enumerate cases was also significantly reduced. This work demonstrates the potential of a digital image analysis algorithm to improve the concordance of pathologist HER2 amplification status reporting in less common HER2 groups. This has the potential to improve therapy selection and outcomes for patients with HER2-low and borderline HER2-amplified breast cancers.

Ejection Fraction Estimation from Echocardiograms Using Optimal Left Ventricle Feature Extraction Based on Clinical Methods.

Echocardiography is one of the imaging systems most often utilized for assessing heart anatomy and function. Left ventricle ejection fraction (LVEF) is an important clinical variable assessed from echocardiography via the measurement of left ventricle (LV) parameters. Significant inter-observer and intra-observer variability is seen when LVEF is quantified by cardiologists using huge echocardiography data. Machine learning algorithms have the capability to analyze such extensive datasets and identify intricate patterns of structure and function of the heart that highly skilled observers might overlook, hence paving the way for computer-assisted diagnostics in this field. In this study, LV segmentation is performed on echocardiogram data followed by feature extraction from the left ventricle based on clinical methods. The extracted features are then subjected to analysis using both neural networks and traditional machine learning algorithms to estimate the LVEF. The results indicate that employing machine learning techniques on the extracted features from the left ventricle leads to higher accuracy than the utilization of Simpson's method for estimating the LVEF. The evaluations are performed on a publicly available echocardiogram dataset, EchoNet-Dynamic. The best results are obtained when DeepLab, a convolutional neural network architecture, is used for LV segmentation along with Long Short-Term Memory Networks (LSTM) for the regression of LVEF, obtaining a dice similarity coefficient of 0.92 and a mean absolute error of 5.736%.

HER2 in situ hybridization testing in breast cancer: Applying algorithm-assisted assessment to reduce interobserver variability in difficult cases.

e13561 Background: Accurate assessment of HER2 expression by HER2 immunohistochemistry and in-situ hybridization (ISH) is critical for the management of patients with breast cancer. The revised 2018 ASCO/CAP guidelines defined 5 subgroups based on HER2 expression and copy number. Few reports have investigated interobserver variability when applying the revised guidelines. Digital algorithms may have a role in assisting pathologists in assessing cases with equivocal and uncommon patterns of HER2 amplification. This study was designed to evaluate interobserver variability in the gold standard manual assessment of HER2 ISH, particularly in less common HER2 ISH groups. Subsequently, we evaluated whether a new digital image analysis algorithm (Roche uPath HER2 Dual ISH image analysis) could reduce interobserver variability and potentially improve patient selection for targeted anti-HER2 therapy. Methods: Of 2982 breast cancer cases, we derived an enriched cohort of 99 patient cases with HER2 IHC 2+ or 3+ and selective ISH results (groups 2-4, evidence of HER2 heterogeneity, borderline negative group 5 and borderline positive group 1). As per guidelines, a minimum of 20 tumour cells per slide were manually and independently counted by 4 anatomical pathologists. After a washout period and second round of de-identification, the slides were scanned (DP200; Roche) and re-scored using the digital algorithm. Results: Pathologist reporting of patient HER2 status (HER2 positive vs. HER2 negative) using manual microscopy demonstrated significant interobserver variability, with a Fleiss’s kappa value of 0.471 (95% CI: 0.378-0.564; P= 1.24E-21) indicating fair-moderate agreement. Interobserver variability in manual scoring of HER2 ISH at the level of HER2 group designation (groups 1-5) showed poor-moderate reliability between pathologists, with an intraclass correlation coefficient (ICC) of 0.526 (95% CI: 0.407-0.641; P= 9.28E-21). Repeat analysis of the cohort using algorithm-assisted quantification resulted in moderate-substantial concordance in HER2 patient status reporting, with a Fleiss’s kappa value of 0.666 (95% CI: 0.572-0.759; P= 1.70E-40). Algorithm-assisted quantification resulted in moderate-good agreement of HER2 group designation, with an ICC of 0.763 (95% CI: 0.683-0.832; P= 8.76E-48). In sub-group analysis, use of the algorithm improved concordance particularly in groups 2,4 and 5. Time to report cases was also significantly reduced with the use of the algorithm. Conclusions: Our study demonstrates the potential to improve the concordance of HER2 amplification status reporting in less common HER2 subgroups using a digital algorithm and potentially improve therapy selection and outcomes for patients with HER2-low and borderline HER2-amplified breast cancers.

Quick Assessment of the Lower Limit of Cerebral Autoregulation Using Transcranial Doppler during Cardiopulmonary Bypass in Cardiac Surgery: A Feasibility Study.

During cardiac surgery, maintaining a mean arterial pressure (MAP) within the range of cerebral autoregulation (CA) may prevent postoperative morbidity. The lower limit of cerebral autoregulation (LLA) can be determined using the mean velocity index (Mx). The standard Mx is averaged over a ten second period ( ) while using a two second averaging period ( ) is faster and may record more rapid variations in LLA. The objective of this study is to compare a quick determination of LLA (qLLA) using with the reference LLA (rLLA) using . Single center, retrospective, observational study. Patients undergoing cardiac surgery with cardiopulmonary bypass. From January 2020 to April 2021, perioperative transcranial doppler measuring cerebral artery velocity was placed on cardiac surgery patients in order to correlate with continuous MAP values. Calculation of each patient's Mx was manually determined after the surgery and qLLA and rLLA were then calculated using a threshold value of Mx 0.4. 55 patients were included. qLLA was found in 78% of the cases versus 47% for rLLA. Despite a -3 mmHg mean bias, limits of agreement were large [-19 mmHg (95% CI: -13; -25), and +13 mmHg (95% CI: +6; +19)]. There was an important interobserver variability (kappa rLLA = 0.46; 95% CI: 0.24-0.66; and Kappa qLLA = 0.36; 95% CI: 0.20-0.52). Calculation of qLLA is feasible. However, the large limits of agreement and significant interobserver variability prevent any clinical utility or interchangeability with rLLA.

Automated Deep Learning-Based Classification of Wilms Tumor Histopathology.

(1) Background: Histopathological assessment of Wilms tumors (WT) is crucial for risk group classification to guide postoperative stratification in chemotherapy pre-treated WT cases. However, due to the heterogeneous nature of the tumor, significant interobserver variation between pathologists in WT diagnosis has been observed, potentially leading to misclassification and suboptimal treatment. We investigated whether artificial intelligence (AI) can contribute to accurate and reproducible histopathological assessment of WT through recognition of individual histopathological tumor components. (2) Methods: We assessed the performance of a deep learning-based AI system in quantifying WT components in hematoxylin and eosin-stained slides by calculating the Sørensen-Dice coefficient for fifteen predefined renal tissue components, including six tumor-related components. We trained the AI system using multiclass annotations from 72 whole-slide images of patients diagnosed with WT. (3) Results: The overall Dice coefficient for all fifteen tissue components was 0.85 and for the six tumor-related components was 0.79. Tumor segmentation worked best to reliably identify necrosis (Dice coefficient 0.98) and blastema (Dice coefficient 0.82). (4) Conclusions: Accurate histopathological classification of WT may be feasible using a digital pathology-based AI system in a national cohort of WT patients.

A Study on the Diagnostic Value of p57kip2 among Gestational Trophoblastic Diseases at a Philippine Hospital in 2020

Hydatidiform mole (H. mole), a gestational trophoblastic disease, is common in the Philippines. In general, the differentiation between complete hydatidiform mole and partial hydatidiform mole is based on morphologic evaluation via routine H&E. The problem lies on the lack of a standard criteria and overlapping features. The p57kip2 immunohistochemical stain is inexpensive, locally available, and could provide reliable results in differentiating between complete H. mole and partial H. mole. This study aims to determine the accuracy of the diagnosis of H. moles by routine H&E stain as compared to the diagnosis by p57kip2 IHC stain. There were a total of 31 cases of H. moles in 2020. It includes 20 cases of complete H. moles and 11 cases of partial H. moles. With the advent of the p57kip2 IHC stain, a total of 18 cases were diagnosed as complete H. mole and 13 were diagnosed as partial H. mole. Diagnostic statistics such as sensitivity (Sn), specificity (Sp), positive predictive value (PPV), negative predictive value (NPV), accuracy, and kappa statistics were computed to validate the diagnostic value of H&E staining with p57kip2 IHC staining on H. moles. This study concludes that a significant increase in accuracy and reliability are attained when supplemented with p57kip2 IHC stain. In addition, criteria in analyzing p57kip2 are simple and reproducible without significant interobserver variability. It is therefore recommended to use p57kip2 IHC stain on vesicular tissues to deduce the true nature of their pathology.

18F-FDG-Based Radiomics and Machine Learning: Useful Help for Aortic Prosthetic Valve Infective Endocarditis Diagnosis?

BackgroundFluorine-18 fluorodeoxyglucose (18F-FDG)-positron emission tomography (PET)/computed tomography (CT) results in better sensitivity for prosthetic valve endocarditis (PVE) diagnosis, but visual image analysis results in relatively weak specificity and significant interobserver variability. ObjectivesThe primary objective of this study was to evaluate the performance of a radiomics and machine learning–based analysis of 18F-FDG PET/CT (PET-ML) as a major criterion for the European Society of Cardiology score using machine learning as a major imaging criterion (ESC-ML) in PVE diagnosis. The secondary objective was to assess performance of PET-ML as a standalone examination. MethodsAll 18F-FDG-PET/CT scans performed for suspected aortic PVE at a single center from 2015 to 2021 were retrospectively included. The gold standard was expert consensus after at least 3 months’ follow-up. The machine learning (ML) method consisted of manually segmenting each prosthetic valve, extracting 31 radiomics features from the segmented region, and training a ridge logistic regressor to predict PVE. Training and hyperparameter tuning were done with a cross-validation approach, followed by an evaluation on an independent test database. ResultsA total of 108 patients were included, regardless of myocardial uptake, and were divided into training (n = 68) and test (n = 40) cohorts. Considering the latter, PET-ML findings were positive for 13 of 22 definite PVE cases and 3 of 18 rejected PVE cases (59% sensitivity, 83% specificity), thus leading to an ESC-ML sensitivity of 72% and a specificity of 83%. ConclusionsThe use of ML for analyzing 18F-FDG-PET/CT images in PVE diagnosis was feasible and beneficial, particularly when ML was included in the ESC 2015 criteria. Despite some limitations and the need for future developments, this approach seems promising to optimize the role of 18F-FDG PET/CT in PVE diagnosis.

Hepatocytic ballooning in non-alcoholic steatohepatitis: Dilemmas and future directions.

Hepatocytic ballooning is a key histological feature in the diagnosis of non-alcoholic steatohepatitis (NASH) and is an essential component of the two most widely used histological scoring systems for diagnosing and staging non-alcoholic fatty liver disease (NAFLD) [namely, the NAFLD activity score (NAS), and the steatosis, activity and fibrosis (SAF) scoring system]. As a result of the increasing incidence of NASH globally, the diagnostic challenges of hepatocytic ballooning are unprecedented. Despite the clear pathological concept of hepatocytic ballooning, there are still challenges in assessing hepatocytic ballooning in 'real life' situations. Hepatocytic ballooning can be confused with cellular oedema and microvesicular steatosis. Significant inter-observer variability does exist in assessing the presence and severity of hepatocytic ballooning. In this review article, we describe the underlying mechanisms associated with hepatocytic ballooning. Specifically, we discuss the increased endoplasmic reticulum stress and the unfolded protein response, as well as the rearrangement of the intermediate filament cytoskeleton, the appearance of Mallory-Denk bodies and activation of the sonic Hedgehog pathway. We also discuss the use of artificial intelligence in the detection and interpretation of hepatocytic ballooning, which may provide new possibilities for future diagnosis and treatment.

Evaluation of Strabocheck® as an objective measurement method of comitant horizontal strabismus in children undergoing surgery

ABSTRACT The gold standard of angle measurement is the prism alternating cover test (PCT). This method requires cooperation from the child, experience, and presents a significant inter-observer variability. Strabocheck®(SK) is a new simple tool for objective and semi-automated angle measurement. Our purpose is to evaluate Strabocheck® in children undergoing surgery for comitant horizontal strabismus. The study population was divided into 3 groups: infantile esotropia, partially accommodative esotropia and intermittent exotropia. The primary endpoint was the agreement between Strabocheck®and the PCT. A total of 44 children were included prospectively. The correlation between the angle measured by the PCT and the angle measured by SK was strong (R = 0.87). The mean absolute difference in the angle measured by the two methods was Δ = 11.9+/- 9.8 diopters. The Bland-Altman plot shows a 95% interval limit between −30.0 [−34.4; −25.6] and 31.0 [26.7; 35.4] diopters. SK is an interesting tool to evaluate the angle of strabismus in children. However, the residual discordance between PCT and SK leads us to question the real value of the angle, which can only be approximated. A better clinical experience of this new tool in relation to the clinical condition and the PCT will allow to have a more accurate idea on the true angle and will probably help the surgeon to adapt this procedure.

A pair of deep learning auto-contouring models for prostate cancer patients injected with a radio-transparent versus radiopaque hydrogel spacer.

Absorbable hydrogel spacer injected between prostate and rectum is gaining popularity for rectal sparing. The spacer alters patient anatomy and thus requires new auto-contouring models. To report the development and comprehensive evaluation of two deep-learning models for patients injected with a radio-transparent (model I) versus radiopaque (model II) spacer. Model I was trained and cross-validated by 135 cases with transparent spacer and tested on 24 cases. Using refined training methods, model II was trained and cross-validated by the same dataset, but with the Hounsfield Unit distribution in the spacer overridden by that obtained from ten cases with opaque spacer. Model II was tested on 64 cases. The models auto-contour eight regions of interest (ROIs): spacer, prostate, proximal seminal vesicles (SVs), left and right femurs, bladder, rectum, and penile bulb. Qualitatively, each auto contour (AC), as well as the composite set, was assessed against manual contour (MC), by a radiation oncologist using a 1 (accepted directly or after minor editing), 2 (accepted after moderate editing), 3 (accepted after major editing), and 4 (rejected) scoring scale. The efficiency gain was characterized by the mean score as nearly complete [1-1.75], substantial (1.75-2.5], meaningful (2.5-3.25], and no (3.25-4.00]. Quantitatively, the geometric similarity between AC and MC was evaluated by dice similarity coefficient (DSC) and mean distance to agreement (MDA), using tolerance recommended by AAPM TG-132 Report. The results by the two models were compared to examine the outcome of the refined training methods. The large number of testing cases for model II allowed further investigation of inter-observer variability in clinical dataset. The correlation between score and DSC/MDA was studied on the ROIs with 10 or more counts of each acceptable score (1, 2, 3). For model I/model II: the mean score was 3.63/1.30 for transparent/opaque spacer, 2.71/2.16 for prostate, 3.25/2.44 for proximal SVs, 1.13/1.02 for both femurs, 2.25/1.25 for bladder, 3.00/2.06 for rectum, 3.38/2.42 for penile bulb, and 2.79/2.20 for the composite set; the mean DSC was 0.52/0.84 for spacer, 0.84/0.85 for prostate, 0.60/0.62 for proximal SVs, 0.94/0.96 for left femur, 0.95/0.96 for right femur, 0.91/0.95 for bladder, 0.81/0.84 for rectum, and 0.65/0.65 for penile bulb; and the mean MDA was 2.9/0.9mm for spacer, 1.9/1.7mm for prostate, 2.4/2.3mm for proximal SVs, 0.8/0.5mm for left femur, 0.7/0.5mm for right femur, 1.5/0.9mm for bladder, 2.3/1.9mm for rectum, and 2.2/2.2mm for penile bulb. Model II showed significantly improved scores for all ROIs, and metrics for spacer, femurs, bladder, and rectum. Significant inter-observer variability was only found for prostate. Highly linear correlation between the score and DSC was found for the two qualified ROIs (prostate and rectum). The overall efficiency gain was meaningful for model I and substantial for model II. The ROIs meeting the clinical deployment criteria (mean score below 3.25, DSC above 0.8, and MDA below 2.5mm) included prostate, both femurs, bladder and rectum for both models, and spacer for model II.

Updates on Lymphovascular Invasion in Breast Cancer.

Traditionally, lymphovascular invasion (LVI) has represented one of the foremost pathological features of malignancy and has been associated with a worse prognosis in different cancers, including breast carcinoma. According to the most updated reporting protocols, the assessment of LVI is required in the pathology report of breast cancer surgical specimens. Importantly, strict histological criteria should be followed for LVI assessment, which nevertheless is encumbered by inconsistency in interpretation among pathologists, leading to significant interobserver variability and scarce reproducibility. Current guidelines for breast cancer indicate biological factors as the main determinants of oncological and radiation therapy, together with TNM staging and age. In clinical practice, the widespread use of genomic assays as a decision-making tool for hormone receptor-positive, HER2-negative breast cancer and the subsequent availability of a reliable prognostic predictor have likely scaled back interest in LVI's predictive value. However, in selected cases, the presence of LVI impacts adjuvant therapy. This review summarizes current knowledge on LVI in breast cancer with regard to definition, histopathological assessment, its biological understanding, clinicopathological association, and therapeutic implications.

Adverse neonatal outcome and veno-arterial differences in umbilical cord blood pH (ΔpH) at birth: a population-based study of 108,629 newborns

BackgroundUmbilical cord blood gases are routinely used by midwives and obstetricians for quality assurance of birth management and in clinical research. They can form the basis for solving medicolegal issues in the identification of severe intrapartum hypoxia at birth. However, the scientific value of veno-arterial differences in cord blood pH, also known as ΔpH, is largely unknown. By tradition, the Apgar score is frequently used to predict perinatal morbidity and mortality, however significant inter-observer and regional variations decrease its reliability and there is a need to identify more accurate markers of perinatal asphyxia. The aim of our study was to investigate the association of small and large veno-arterial differences in umbilical cord pH, ΔpH, with adverse neonatal outcome.MethodsThis retrospective, population-based study collected obstetric and neonatal data from women giving birth in nine maternity units from Southern Sweden from 1995 to 2015. Data was extracted from the Perinatal South Revision Register, a quality regional health database. Newborns at ≥37 gestational weeks with a complete and validated set of umbilical cord blood samples from both cord artery and vein were included. Outcome measures included: ΔpH percentiles, ‘Small ΔpH’ (10th percentile), ‘Large ΔpH’ (90th percentile), Apgar score (0–6), need for continuous positive airway pressure (CPAP) and admission to neonatal intensive care unit (NICU). Relative risks (RR) were calculated with modified Poisson regression model.ResultsThe study population comprised of 108,629 newborns with complete and validated data. Mean and median ΔpH was 0.08 ± 0.05. Analyses of RR showed that ‘Large ΔpH’ was associated with a decreased RR of adverse perinatal outcome with increasing UApH (at UApH ≥7.20: RR for low Apgar 0.29, P = 0.01; CPAP 0.55, P = 0.02; NICU admission 0.81, P = 0.01). ‘Small ΔpH’ was associated with an increased RR for low Apgar score and NICU admission only at higher UApH values (at UApH 7.15–7.199: RR for low Apgar 1.96, P = 0.01; at UApH ≥7.20: RR for low Apgar 1.65, P = 0.00, RR for NICU admission 1.13, P = 0.01).ConclusionLarge differences between cord venous and arterial pH (ΔpH) at birth were associated with a lower risk for perinatal morbidity including low 5-minute Apgar Score, the need for continuous positive airway pressure and NICU admission when UApH was above 7.15. Clinically, ΔpH may be a useful tool in the assessment of the newborn’s metabolic condition at birth. Our findings may stem from the ability of the placenta to adequately replenish acid-base balance in fetal blood. ‘Large ΔpH’ may therefore be a marker of effective gas exchange in the placenta during birth.

Echo Doppler Parameters of Diastolic Function.

The purpose of this review is to highlight the echo Doppler parameters that form the cornerstone for the evaluation of diastolic function as per the guideline documents of the American Society of Echocardiography (ASE) and the European Association of Cardiovascular Imaging (EACVI). In addition, the individual Doppler-based parameters will be explored, with commentary on the rationale behind their use and the multi-parametric approach to the assessment of diastolic dysfunction (DD) using echocardiography. Previous guidelines for assessment of diastolic function are complex with modest diagnostic performance and significant inter-observer variability. The most recent guidelines have made the evaluation of DD more streamlined with excellent correlation with invasive measures of LV filling pressures. This is a review of the echo-derived Doppler parameters that are integral in the diagnosis and gradation of DD. A brief description of the physiological principles that govern changes in echocardiographic parameters during normal and abnormal diastolic function is also discussed for the appropriate diagnosis of DD using non-invasive Doppler echocardiography techniques.

Video-Based Deep Learning for Automated Assessment of Left Ventricular Ejection Fraction in Pediatric Patients

Significant interobserver and interstudy variability occurs for left ventricular functional indices despite standardization of measurement techniques. Artificial intelligence models trained on adult echocardiograms are not likely to be applicable to a pediatric population. We present EchoNet-Peds, a video-based deep learning algorithm, which matches human expert performance of left ventricular (LV) segmentation and ejection fraction (EF). A large pediatric dataset of 4,467 echocardiograms were used to develop EchoNet-Peds. EchoNet-Peds was trained on 80% of the data for segmentation of the left ventricle and estimation of left ventricular EF. The remaining 20% was used to fine tune and validate the algorithm. In both apical 4-chamber (A4C) and parasternal short-axis views (PSAX), EchoNet-Peds segments the left ventricle with a Dice similarity coefficient of 0.89. EchoNet-Peds estimates EF with a mean absolute error of 3.66% and can routinely identify pediatric patients with systolic dysfunction (area under the curve of 0.95). EchoNet-Peds was trained on pediatric echocardiograms and performed significantly better to estimate EF (p < 0.001) than an adult model applied to the same data. Accurate, rapid automation of EF assessment and recognition of systolic dysfunction in a pediatric population are feasible using EchoNet-Peds with the potential for far-reaching clinical impact. In addition, the first large pediatric dataset of annotated echocardiograms is now publicly available for efforts to develop pediatric-specific artificial intelligence algorithms.

Inter-observer variation of target volume delineation for CT-guided cervical cancer brachytherapy.

Delineation is a critical and challenging step in radiotherapy planning. Differences in delineation among observers are common, despite the presence of contouring guidelines. This study aimed to identify the inter-observer variability in the target volume delineation of computed tomography (CT)-guided brachytherapy for cervical cancer. Four radiation oncologists (ROs) with different expertise levels delineated high-risk (HR) and intermediate-risk (IR) clinical target volume (CTV) according to GYN GEC-ESTRO recommendations, in a blinded manner on every CT set of ten locally advanced cervical cancer cases. The most experienced RO's contours were determined as the index and used for comparison. Dice similarity coefficient (DSC) and pairwise Hausdorff distance (HD) metrics were applied to compare the overlap and gross deviations of all contours. Median DSC for HR-CTV and IR-CTV were 0.73 and 0.76, respectively, and a good concordance was achieved for both in majority of contours. While there was no difference in DSC measurements for HR-CTV among the three ROs, RO-3 provided improved DSC values for IR-CTV (p = 0.01). Median HD95 was 5.02 mm and 6.83 mm, and median HDave was 1.69 mm and 2.21 mm for HR-CTV and IR-CTV, respectively. There was no significant difference among ROs in HR-CTV for HD95 or HDave; however, IR-CTV value was significantly improved according to RO-3 (p = 0.01). Case-by-case HD analysis showed no significant inter-observer variations, except for two cases. The inter-observer agreement is generally high for target volumes in CT-guided brachytherapy for cervical cancer. The agreement is lower for IR-CTV than HR-CTV. The individual characteristics of each case and different expertise levels of the ROs may have caused the differences. Despite the good concordance for delineation, dosimetric consequences can still be clinically significant.

Adverse perinatal outcome and veno-arterial differences in umbilical cord pH at birth: a population-based study

Analysis of cord blood gases, especially cord arterial pH (UApH), can identify newborns at risk of adverse outcomes, yet the scientific value of veno-arterial differences in cord pH (ΔpH) is largely unknown. By tradition, the Apgar score is frequently used to predict perinatal morbidity and mortality, however, significant inter-observer variations decrease its reliability and there is a need for more accurate markers of perinatal asphyxia. The objective of our study was to investigate the association of ΔpH to risk for adverse perinatal outcome.

SPECT-MPI for Coronary Artery Disease: A Deep Learning Approach.

Worldwide, coronary artery disease (CAD) is a leading cause of mortality and morbidity and remains to be a top health priority in many countries. A non-invasive imaging modality for diagnosis of CAD such as single photon emission computed tomography-myocardial perfusion imaging (SPECT-MPI) is usually requested by cardiologists as it displays radiotracer distribution in the heart reflecting myocardial perfusion. The interpretation of SPECT-MPI is done visually by a nuclear medicine physician and is largely dependent on his clinical experience and showing significant inter-observer variability. The aim of the study is to apply a deep learning approach in the classification of SPECT-MPI for perfusion abnormalities using convolutional neural networks (CNN). A publicly available anonymized SPECT-MPI from a machine learning repository (https://www.kaggle.com/selcankaplan/spect-mpi) was used in this study involving 192 patients who underwent stress-test-rest Tc99m MPI. An exploratory approach of CNN hyperparameter selection to search for optimum neural network model was utilized with particular focus on various dropouts (0.2, 0.5, 0.7), batch sizes (8, 16, 32, 64), and number of dense nodes (32, 64, 128, 256). The base CNN model was also compared with the commonly used pre-trained CNNs in medical images such as VGG16, InceptionV3, DenseNet121 and ResNet50. All simulations experiments were performed in Kaggle using TensorFlow 2.6.0., Keras 2.6.0, and Python language 3.7.10. The best performing base CNN model with parameters consisting of 0.7 dropout, batch size 8, and 32 dense nodes generated the highest normalized Matthews Correlation Coefficient at 0.909 and obtained 93.75% accuracy, 96.00% sensitivity, 96.00% precision, and 96.00% F1-score. It also obtained higher classification performance as compared to the pre-trained architectures. The results suggest that deep learning approaches through the use of CNN models can be deployed by nuclear medicine physicians in their clinical practice to further augment their decision skills in the interpretation of SPECT-MPI tests. These CNN models can also be used as a dependable and valid second opinion that can aid physicians as a decision-support tool as well as serve as teaching or learning materials for the less-experienced physicians particularly those still in their training career. These highlights the clinical utility of deep learning approaches through CNN models in the practice of nuclear cardiology.

Diagnosis of Coronary Artery Disease from Myocardial Perfusion Imaging Using Convolutional Neural Networks

Cardiovascular disease is a highly prevalent health problem in both underdeveloped and developing countries worldwide. As such, it remains to be one of the top health priorities in many countries. In coronary artery disease (CAD), the formation of an atherosclerotic plaque is evident in the lumen of blood vessels leading to the derangement in blood flow resulting to diminished delivery of oxygen to the myocardium. Single Photon Emission Computed Tomography – Myocardial Perfusion Imaging (SPECT-MPI) is a usually requested imaging modality to evaluate for CAD. Visual evaluation of the MPI images is performed by a nuclear medicine doctor and is largely dependent on his experience showing significant inter-observer variability. The study aims to assess the performance of convolutional neural networks (CNN) using transfer learning to classify SPECT-MPI for perfusion abnormalities using an anonymized publicly available SPECT-MPI dataset. The pre-processing methods that were applied to the dataset were the following: (a) normalization of images, (b) shuffling of images, (c) train-test split, and (d) geometric augmentation. The pre-processed data was then entered to the popular pre-trained CNNs typically applied to medical images: VGG16, DenseNet121, InceptionV3 and ResNet50. The best performing models were obtained by VGG16 and InceptionV3 with the highest accuracy rate of 84.38%. However, VGG16 had higher recall and F1-scores as compared to InceptionV3 while InceptionV3 had higher precision. Nonetheless, VGG16, InceptionV3 and DenseNet121 obtained similar performance metrics with each other (recall:80-100%, precision: 80.65-100%, F1-scores: 88.89-90.91%) while ResNet50 generated the lowest performance metrics. Overall findings suggest that any of these 3 CNN models (VGG16, InceptionV3, DenseNet121) can be deployed by nuclear medicine physicians in their clinical practice to further augment their decision skills in the interpretation of SPECT-MPI tests. The models can also be adopted as dependable and trusted secondary assessment which can guide junior doctors seeking consultation for a reliable diagnosis. These models can likewise serve as teaching or learning materials for the less experienced physicians particularly those still in their training career. This highlights the clinical utility of these models in the practice of nuclear cardiology. The results of the research exhibited encouraging outcomes which may possibly be incorporated clinical work. The study has the potential to enrich CAD discernment and monitoring.

Inter-Observer Variability in MR-Based Target Volume Delineation of Uveal Melanoma

Several efforts are being undertaken toward MRI-based treatment planning for ocular proton therapy for uveal melanoma (UM). The interobserver variability of the gross target volume (GTV) on magnetic resonance imaging (MRI) is one of the important parameters to design safety margins for a reliable treatment. Therefore, this study assessed the interobserver variation in GTV delineation of UM on MRI. Six observers delineated the GTV in 10 different patients using the Big Brother contouring software. Patients were scanned at 3T MRI with a surface coil, and tumors were delineated separately on contrast enhanced 3DT1 (T1gd) and 3DT2-weighted scans with an isotropic acquisition resolution of 0.8 mm. Volume difference and overall local variation (median standard deviation of the distance between the delineated contours and the median contour) were analyzed for each GTV. Additionally, the local variation was analyzed for 4 interfaces: sclera, vitreous, retinal detachment, and tumor-choroid interface. The average GTV was significantly larger on T1gd (0.57cm3) compared with T2 (0.51cm3, P=.01). A not significant higher interobserver variation was found on T1gd (0.41 mm) compared with T2 (0.35 mm). The largest variations were found at the tumor-choroid interface due to peritumoral enhancement (T1gd, 0.62 mm; T2, 0.52 mm). As a result, a larger part of this tumor-choroid interface appeared to be included on T1gd-based GTVs compared with T2, explaining the smaller volumes on T2. The interobserver variation of 0.4 mm on MRI are low with respect to the voxel size of 0.8 mm, enabling small treatment margins. We recommend delineation based on the T1gd-weighted scans, as choroidal tumor extensions might be missed.

Inter‐Observer variability in MRI‐based target volume delineation of uveal melanoma

AbstractPurpose: Several efforts are being undertaken towards MRI‐based treatment planning for ocular proton therapy for uveal melanoma (UM). The inter‐observer variability of the gross target volume (GTV) on MRI is one of the important parameters to design safety margins for a reliable treatment. Therefore, this study assessed the inter‐observer variation in tumour delineation of UM on MRI.Methods: Six observers delineated the tumour in 10 different UM patients using the Big Brother contouring software. Patients were scanned at 3 T MRI with a surface coil and tumours were delineated separately on contrast enhanced 3DT1 (T1gd) and 3DT2‐weighted scans with a resolution of 0.8 mm isotropic for the acquisition voxels (1). Volume difference and overall local variation were analysed for each patient. Additionally the local variation was analysed for four interfaces: sclera, vitreous, retinal detachment, and tumour‐choroid interface.Results: The average tumour contour was significantly larger on T1gd (0.57 cm3) compared to T2 (0.51cm3, p = 0.01). A not significant higher interobserver variation was found on T1gd (0.4 1 mm) compared to T2 (0.35 mm). The largest variations were found at the tumour‐choroid interface, due to peritumoral enhancement (T1gd: 0.62 mm and T2: 0.5 2 mm). As a result, a larger part of this tumour‐choroid interface appeared to be included on T1gd based GTVs compared to T2, explaining the smaller volumes on T2.Conclusions: The interobserver variation of 0.4 mm on MRI are low with respect to the voxel size of 0.8 mm. We recommend to delineate based on the T1gd‐weighted scans, as choroidal tumour extensions might be missed.Reference1. Ferreira, et al. MR imaging characteristics of uveal melanoma with histopathological validation. Neuroradiology. 2022; 64:171. doi: 10.1007/s00234‐021‐02825‐5.