Deep Learning Prostate MRI Segmentation Accuracy and Robustness: A Systematic Review.

Purpose To investigate the accuracy and robustness of prostate segmentation using deep learning across various training data sizes, MRI vendors, prostate zones, and testing methods relative to fellowship-trained diagnostic radiologists. Materials and Methods In this systematic review, Embase, PubMed, Scopus, and Web of Science databases were queried for English-language articles using keywords and related terms for prostate MRI segmentation and deep learning algorithms dated to July 31, 2022. A total of 691 articles from the search query were collected and subsequently filtered to 48 on the basis of predefined inclusion and exclusion criteria. Multiple characteristics were extracted from selected studies, such as deep learning algorithm performance, MRI vendor, and training dataset features. The primary outcome was comparison of mean Dice similarity coefficient (DSC) for prostate segmentation for deep learning algorithms versus diagnostic radiologists. Results Forty-eight studies were included. Most published deep learning algorithms for whole prostate gland segmentation (39 of 42 [93%]) had a DSC at or above expert level (DSC ≥ 0.86). The mean DSC was 0.79 ± 0.06 (SD) for peripheral zone, 0.87 ± 0.05 for transition zone, and 0.90 ± 0.04 for whole prostate gland segmentation. For selected studies that used one major MRI vendor, the mean DSCs of each were as follows: General Electric (three of 48 studies), 0.92 ± 0.03; Philips (four of 48 studies), 0.92 ± 0.02; and Siemens (six of 48 studies), 0.91 ± 0.03. Conclusion Deep learning algorithms for prostate MRI segmentation demonstrated accuracy similar to that of expert radiologists despite varying parameters; therefore, future research should shift toward evaluating segmentation robustness and patient outcomes across diverse clinical settings. Keywords: MRI, Genital/Reproductive, Prostate Segmentation, Deep Learning Systematic review registration link: osf.io/nxaev © RSNA, 2024.

Body oncologic imaging subspecialty training a curriculum based on the experience in a tertiary cancer center.

PURPOSE: To describe the structure of a dedicated body oncologic imaging fellowship program. To summarize the numbers and types of cross-sectional imaging examinations reported by fellows. METHODS: The curriculum, training methods, and assessment measures utilized in the program were reviewed and described. An educational retrospective analysis was conducted. Data on the number of examinations interpreted by fellows, breakdown of modalities, and examinations by disease management team (DMT) were collected. RESULTS: A total of 38 fellows completed the fellowship program during the study period. The median number of examinations reported per fellow was 2296 [interquartile range: 2148 - 2534], encompassing all oncology-relevant imaging modalities: CT 721 [646-786], MRI 1158 [1016-1309], ultrasound 256 [209-320] and PET/CT 176 [130-202]. The breakdown of examinations by DMT revealed variations in imaging patterns, with MRIs most frequently interpreted for genitourinary, musculoskeletal, and hepatobiliary cancers, and CTs most commonly for general staging or assessment of nonspecific symptoms. CONCLUSION: This descriptive analysis may serve as a foundation for the development of similar fellowship programs and the advancement of body oncologic imaging. The volume and diversity of examinations reported by fellows highlights the comprehensive nature of body oncologic imaging.

Integrated Automatic Examination Assignment Reduces Radiologist Interruptions: A 2-Year Cohort Study of 232,022 Examinations.

Radiology departments face challenges in delivering timely and accurate imaging reports, especially in high-volume, subspecialized settings. In this retrospective cohort study at a tertiary cancer center, we assessed the efficacy of an Automatic Assignment System (AAS) in improving radiology workflow efficiency by analyzing 232,022 CT examinations over a 12-month period post-implementation and compared it to a historical control period. The AAS was integrated with the hospital-wide scheduling system and set up to automatically prioritize and distribute unreported CT examinations to available radiologists based on upcoming patient appointments, coupled with an email notification system. Following this AAS implementation, despite a 9% rise in CT volume, coupled with a concurrent 8% increase in the number of available radiologists, the mean daily urgent radiology report requests (URR) significantly decreased by 60% (25 ± 12 to 10 ± 5, t = -17.6, p < 0.001), and URR during peak days (95th quantile) was reduced by 52.2% from 46 to 22 requests. Additionally, the mean turnaround time (TAT) for reporting was significantly reduced by 440 min for patients without immediate appointments and by 86 min for those with same-day appointments. Lastly, patient waiting time sampled in one of the outpatient clinics was not negatively affected. These results demonstrate that AAS can substantially decrease workflow interruptions and improve reporting efficiency.

Streamlining Radiology Workflows Through the Development and Deployment of Automated Microservices.

Microservices are a software development approach where an application is structured as a collection of loosely coupled, independently deployable services, each focusing on executing a specific purpose. The development of microservices could have a significant impact on radiology workflows, allowing routine tasks to be automated and improving the efficiency and accuracy of radiologic tasks. This technical report describes the development of several microservices that have been successfully deployed in a tertiary cancer center, resulting in substantial time savings for radiologists and other staff involved in radiology workflows. These microservices include the automatic generation of shift emails, notifying administrative staff and faculty about fellows on rotation, notifying referring physicians about outside examinations, and populating report templates with information from PACS and RIS. The report outlines the common thought process behind developing these microservices, including identifying a problem, connecting various APIs, collecting data in a database, writing a prototype and deploying it, gathering feedback and refining the service, putting it in production, and identifying staff who are in charge of maintaining the service. The report concludes by discussing the benefits and challenges of microservices in radiology workflows, highlighting the importance of multidisciplinary collaboration, interoperability, security, and privacy.

Deep learning for [18F]fluorodeoxyglucose-PET-CT classification in patients with lymphoma: a dual-centre retrospective analysis.

BACKGROUND: The rising global cancer burden has led to an increasing demand for imaging tests such as [18F]fluorodeoxyglucose ([18F]FDG)-PET-CT. To aid imaging specialists in dealing with high scan volumes, we aimed to train a deep learning artificial intelligence algorithm to classify [18F]FDG-PET-CT scans of patients with lymphoma with or without hypermetabolic tumour sites. METHODS: In this retrospective analysis we collected 16 583 [18F]FDG-PET-CTs of 5072 patients with lymphoma who had undergone PET-CT before or after treatment at the Memorial Sloa Kettering Cancer Center, New York, NY, USA. Using maximum intensity projection (MIP), three dimensional (3D) PET, and 3D CT data, our ResNet34-based deep learning model (Lymphoma Artificial Reader System [LARS]) for [18F]FDG-PET-CT binary classification (Deauville 1-3 vs 4-5), was trained on 80% of the dataset, and tested on 20% of this dataset. For external testing, 1000 [18F]FDG-PET-CTs were obtained from a second centre (Medical University of Vienna, Vienna, Austria). Seven model variants were evaluated, including MIP-based LARS-avg (optimised for accuracy) and LARS-max (optimised for sensitivity), and 3D PET-CT-based LARS-ptct. Following expert curation, areas under the curve (AUCs), accuracies, sensitivities, and specificities were calculated. FINDINGS: In the internal test cohort (3325 PET-CTs, 1012 patients), LARS-avg achieved an AUC of 0·949 (95% CI 0·942-0·956), accuracy of 0·890 (0·879-0·901), sensitivity of 0·868 (0·851-0·885), and specificity of 0·913 (0·899-0·925); LARS-max achieved an AUC of 0·949 (0·942-0·956), accuracy of 0·868 (0·858-0·879), sensitivity of 0·909 (0·896-0·924), and specificity of 0·826 (0·808-0·843); and LARS-ptct achieved an AUC of 0·939 (0·930-0·948), accuracy of 0·875 (0·864-0·887), sensitivity of 0·836 (0·817-0·855), and specificity of 0·915 (0·901-0·927). In the external test cohort (1000 PET-CTs, 503 patients), LARS-avg achieved an AUC of 0·953 (0·938-0·966), accuracy of 0·907 (0·888-0·925), sensitivity of 0·874 (0·843-0·904), and specificity of 0·949 (0·921-0·960); LARS-max achieved an AUC of 0·952 (0·937-0·965), accuracy of 0·898 (0·878-0·916), sensitivity of 0·899 (0·871-0·926), and specificity of 0·897 (0·871-0·922); and LARS-ptct achieved an AUC of 0·932 (0·915-0·948), accuracy of 0·870 (0·850-0·891), sensitivity of 0·827 (0·793-0·863), and specificity of 0·913 (0·889-0·937). INTERPRETATION: Deep learning accurately distinguishes between [18F]FDG-PET-CT scans of lymphoma patients with and without hypermetabolic tumour sites. Deep learning might therefore be potentially useful to rule out the presence of metabolically active disease in such patients, or serve as a second reader or decision support tool. FUNDING: National Institutes of Health-National Cancer Institute Cancer Center Support Grant.

Evaluating residual tumor after neoadjuvant chemotherapy for muscle-invasive urothelial bladder cancer: diagnostic performance and outcomes using biparametric vs. multiparametric MRI.

BACKGROUND: Neoadjuvant chemotherapy (NAC) before radical cystectomy is standard of care in patients with muscle-invasive bladder cancer (MIBC). Response assessment after NAC is important but suboptimal using CT. We assessed MRI without vs. with intravenous contrast (biparametric [BP] vs. multiparametric [MP]) for identifying residual disease on cystectomy and explored its prognostic role. METHODS: Consecutive MIBC patients that underwent NAC, MRI, and cystectomy between January 2000-November 2022 were identified. Two radiologists reviewed BP-MRI (T2 + DWI) and MP-MRI (T2 + DWI + DCE) for residual tumor. Diagnostic performances were compared using receiver operating characteristic curve analysis. Kaplan-Meier curves and Cox proportional-hazards models were used to evaluate association with disease-free survival (DFS). RESULTS: 61 patients (36 men and 25 women; median age 65 years, interquartile range 59-72) were included. After NAC, no residual disease was detected on pathology in 19 (31.1%) patients. BP-MRI was more accurate than MP-MRI for detecting residual disease after NAC: area under the curve = 0.75 (95% confidence interval (CI), 0.62-0.85) vs. 0.58 (95% CI, 0.45-0.70; p = 0.043). Sensitivity were identical (65.1%; 95% CI, 49.1-79.0) but specificity was higher in BP-MRI compared with MP-MRI for determining residual disease: 77.8% (95% CI, 52.4-93.6) vs. 38.9% (95% CI, 17.3-64.3), respectively. Positive BP-MRI and residual disease on pathology were both associated with worse DFS: hazard ratio (HR) = 4.01 (95% CI, 1.70-9.46; p = 0.002) and HR = 5.13 (95% CI, 2.66-17.13; p = 0.008), respectively. Concordance between MRI and pathology results was significantly associated with DFS. Concordant positive (MRI+/pathology+) patients showed worse DFS than concordant negative (MRI-/pathology-) patients (HR = 8.75, 95% CI, 2.02-37.82; p = 0.004) and compared to the discordant group (MRI+/pathology- or MRI-/pathology+) with HR = 3.48 (95% CI, 1.39-8.71; p = 0.014). CONCLUSION: BP-MRI was more accurate than MP-MRI for identifying residual disease after NAC. A negative BP-MRI was associated with better outcomes, providing complementary information to pathological assessment of cystectomy specimens.

Improving Radiology Oncologic Imaging Trainee Case Diversity through Automatic Examination Assignment: Retrospective Study from a Tertiary Cancer Center.

In a retrospective single-center study, the authors assessed the efficacy of an automated imaging examination assignment system for enhancing the diversity of subspecialty examinations reported by oncologic imaging fellows. The study aimed to mitigate traditional biases of manual case selection and ensure equitable exposure to various case types. Methods included evaluating the proportion of "uncommon" to "common" cases reported by fellows before and after system implementation and measuring the weekly Shannon Diversity Index to determine case distribution equity. The proportion of reported uncommon cases more than doubled from 8.6% to 17.7% in total, at the cost of a concurrent 9.0% decrease in common cases from 91.3% to 82.3%. The weekly Shannon Diversity Index per fellow increased significantly from 0.66 (95% CI: 0.65, 0.67) to 0.74 (95% CI: 0.72, 0.75; P < .001), confirming a more balanced case distribution among fellows after introduction of the automatic assignment. © RSNA, 2023 Keywords: Computer Applications, Education, Fellows, Informatics, MRI, Oncologic Imaging.

Taking PI-QUAL beyond the prostate: Towards a standardized radiological image quality score (RI-QUAL).

PURPOSE: To compare the interreader agreement of a novel quality score, called the Radiological Image Quality Score (RI-QUAL), to a slighly modified version of the existing Prostate Imaging Quality (mPI-QUAL) score for magnetic resonance imaging (MRI) of the prostate. METHODS: A total of 43 consecutive scans were evaluated by two subspecialized radiologists who assigned scores using both the RI-QUAL and mPI-QUAL methods. The interreader agreement was analyzed using three statistical methods: concordance correlation coefficient (CCC), intraclass correlation coefficient (ICC), and Cohen's kappa. Time needed to arrive at a quality judgment was measured and compared using the Wilcoxon signed rank test. RESULTS: The interreader agreement for RI-QUAL and mPI-QUAL scores was comparable, as evidenced by the high CCC (0.76 vs. 0.77, p = 0.93), ICC (0.86 vs. 0.87, p = 0.93), and moderate Cohen's kappa (0.61 vs. 0.64, p = 0.85) values. Moreover, RI-QUAL assessment was faster than mPI-QUAL (19 vs. 40 s, p = 0.001). CONCLUSION: RI-QUAL is a new quality score that has comparable interreader agreement to the mPI-QUAL score, but with the potential to be applied to different MRI protocols and even different modalities. Like PI-QUAL, RI-QUAL may also facilitate communication about quality to referring physicians, as it provides a standardized and easily interpretable score. Further studies are warranted to validate the usefulness of RI-QUAL in larger patient cohorts and for other imaging modalities.

Sarcomatoid renal cell carcinoma: MRI features and their association with survival.

OBJECTIVE: To evaluate MRI features of sarcomatoid renal cell carcinoma (RCC) and their association with survival. METHODS: This retrospective single-center study included 59 patients with sarcomatoid RCC who underwent MRI before nephrectomy during July 2003-December 2019. Three radiologists reviewed MRI findings of tumor size, non-enhancing areas, lymphadenopathy, and volume (and percentage) of T2 low signal intensity areas (T2LIA). Clinicopathological factors of age, gender, ethnicity, baseline metastatic status, pathological details (subtype and extent of sarcomatoid differentiation), treatment type, and follow-up were extracted. Survival was estimated using Kaplan-Meier method and Cox proportional-hazards regression model was used to identify factors associated with survival. RESULTS: Forty-one males and eighteen females (median age 62 years; interquartile range 51-68) were included. T2LIAs were present in 43 (72.9%) patients. At univariate analysis, clinicopathological factors associated with shorter survival were: greater tumor size (> 10 cm; HR [hazard ratio] = 2.44, 95% CI 1.15-5.21; p = 0.02), metastatic lymph nodes (present; HR = 2.10, 95% CI 1.01-4.37; p = 0.04), extent of sarcomatoid differentiation (non-focal; HR = 3.30, 95% CI 1.55-7.01; p < 0.01), subtypes other than clear cell, papillary, or chromophobe (HR = 3.25, 95% CI 1.28-8.20; p = 0.01), and metastasis at baseline (HR = 5.04, 95% CI 2.40-10.59; p < 0.01). MRI features associated with shorter survival were: lymphadenopathy (HR = 2.24, 95% CI 1.16-4.71; p = 0.01) and volume of T2LIA (> 3.2 mL, HR = 4.22, 95% CI 1.92-9.29); p < 0.01). At multivariate analysis, metastatic disease (HR = 6.89, 95% CI 2.79-16.97; p < 0.01), other subtypes (HR = 9.50, 95% CI 2.81-32.13; p < 0.01), and greater volume of T2LIA (HR = 2.51, 95% CI 1.04-6.05; p = 0.04) remained independently associated with worse survival. CONCLUSION: T2LIAs were present in approximately two thirds of sarcomatoid RCCs. Volume of T2LIA along with clinicopathological factors were associated with survival.

Artificial intelligence in oncologic imaging.

Radiology is integral to cancer care. Compared to molecular assays, imaging has its advantages. Imaging as a noninvasive tool can assess the entirety of tumor unbiased by sampling error and is routinely acquired at multiple time points in oncological practice. Imaging data can be digitally post-processed for quantitative assessment. The ever-increasing application of Artificial intelligence (AI) to clinical imaging is challenging radiology to become a discipline with competence in data science, which plays an important role in modern oncology. Beyond streamlining certain clinical tasks, the power of AI lies in its ability to reveal previously undetected or even imperceptible radiographic patterns that may be difficult to ascertain by the human sensory system. Here, we provide a narrative review of the emerging AI applications relevant to the oncological imaging spectrum and elaborate on emerging paradigms and opportunities. We envision that these technical advances will change radiology in the coming years, leading to the optimization of imaging acquisition and discovery of clinically relevant biomarkers for cancer diagnosis, staging, and treatment monitoring. Together, they pave the road for future clinical translation in precision oncology.

Programmatic Implementation of a Custom Subspecialized Oncologic Imaging Workflow Manager at a Tertiary Cancer Center.

PURPOSE: To evaluate whether a custom programmatic workflow manager reduces reporting turnaround times (TATs) from a body oncologic imaging workflow at a tertiary cancer center. METHODS: A custom software program was developed and implemented in the programming language R. Other aspects of the workflow were left unchanged. TATs were measured over a 12-month period (June-May). The same prior 12-month period served as a historical control. Median TATs of magnetic resonance imaging (MRI) and computed tomography (CT) examinations were compared with a Wilcoxon test. A chi-square test was used to compare the numbers of examinations reported within 24 hours and after 72 hours as well as the proportions of examinations assigned according to individual radiologist preferences. RESULTS: For all MRI and CT examinations (124,507 in 2019/2020 and 138,601 in 2020/2021), the median TAT decreased from 4 (interquartile range: 1-22 hours) to 3 hours (1-17 hours). Reports completed within 24 hours increased from 78% (124,127) to 89% (138,601). For MRI, TAT decreased from 22 (5-49 hours) to 8 hours (2-21 hours), and reports completed within 24 hours increased from 55% (14,211) to 80% (23,744). For CT, TAT decreased from 3 (1-19 hours) to 2 hours (1-13 hours), and reports completed within 24 hours increased from 84% (82,342) to 92% (99,922). Delayed reports (with a TAT > 72 hours) decreased from 17.0% (4,176) to 2.2% (649) for MRI and from 2.5% (2,500) to 0.7% (745) for CT. All differences were statistically significant (P < .001). CONCLUSION: The custom workflow management software program significantly decreased MRI and CT report TATs.

68Ga-PSMA-11 PET/MRI versus multiparametric MRI in men referred for prostate biopsy: primary tumour localization and interreader agreement.

BACKGROUND: Magnetic resonance imaging (MRI) is recommended by the European Urology Association guidelines as the standard modality for imaging-guided biopsy. Recently positron emission tomography with prostate-specific membrane antigen (PSMA PET) has shown promising results as a tool for this purpose. The aim of this study was to compare the accuracy of positron emission tomography with prostate-specific membrane antigen/magnetic resonance imaging (PET/MRI) using the gallium-labeled prostate-specific membrane antigen (68Ga-PSMA-11) and multiparametric MRI (mpMRI) for pre-biopsy tumour localization and interreader agreement for visual and semiquantitative analysis. Semiquantitative parameters included apparent diffusion coefficient (ADC) and maximum lesion diameter for mpMRI and standardized uptake value (SUVmax) and PSMA-positive volume (PSMAvol) for PSMA PET/MRI. RESULTS: Sensitivity and specificity were 61.4% and 92.9% for mpMRI and 66.7% and 92.9% for PSMA PET/MRI for reader one, respectively. RPE was available in 23 patients and 41 of 47 quadrants with discrepant findings. Based on RPE results, the specificity for both imaging modalities increased to 98% and 99%, and the sensitivity improved to 63.9% and 72.1% for mpMRI and PSMA PET/MRI, respectively. Both modalities yielded a substantial interreader agreement for primary tumour localization (mpMRI kappa = 0.65 (0.52-0.79), PSMA PET/MRI kappa = 0.73 (0.61-0.84)). ICC for SUVmax, PSMAvol and lesion diameter were almost perfect (≥ 0.90) while for ADC it was only moderate (ICC = 0.54 (0.04-0.78)). ADC and lesion diameter did not correlate significantly with Gleason score (ρ = 0.26 and ρ = 0.16) while SUVmax and PSMAvol did (ρ = - 0.474 and ρ = - 0.468). CONCLUSIONS: PSMA PET/MRI has similar accuracy and reliability to mpMRI regarding primary prostate cancer (PCa) localization. In our cohort, semiquantitative parameters from PSMA PET/MRI correlated with tumour grade and were more reliable than the ones from mpMRI.

Risk factors for concomitant positive midstream urine culture in patients presenting with symptomatic ureterolithiasis.

In patients with symptomatic ureterolithiasis, immediate treatment of concomitant urinary tract infection (UTI) may prevent sepsis. However, urine cultures require at least 24 h to confirm or exclude UTI, and therefore, clinical variables may help to identify patients who require immediate empirical broad-spectrum antibiotics and surgical intervention. Therefore, we divided a consecutive cohort of 705 patients diagnosed with symptomatic ureterolithiasis at a single institution between 2011 and 2017 into a training (80%) and a testing cohort (20%). A machine-learning-based variable selection approach was used for the fitting of a multivariable prognostic logistic regression model. The discriminatory ability of the model was quantified by the area under the curve (AUC) of receiver-operating curves (ROC). After validation and calibration of the model, a nomogram was created, and decision curve analysis (DCA) was used to evaluate the clinical net-benefit. UTI was observed in 40 patients (6%). LASSO regression selected the variables elevated serum CRP, positive nitrite, and positive leukocyte esterase for fitting of the model with the highest discriminatory ability. In the testing cohort, model performance evaluation for prediction of UTI showed an AUC of 82 (95% CI 71.5-95.7%). Model calibration plots showed excellent calibration. DCA showed a clinically meaningful net-benefit between a threshold probability of 0 and 80% for the novel model, which was superior to the net-benefit provided by either one of its singular components. In conclusion, we developed and internally validated a logistic regression model and a corresponding highly accurate nomogram for prediction of concomitant positive midstream urine culture in patients presenting with symptomatic ureterolithiasis.

Frequency and outcomes of MRI-detected axillary adenopathy following COVID-19 vaccination.

OBJECTIVES: To assess the frequency of ipsilateral axillary adenopathy on breast MRI after COVID-19 vaccination. To investigate the duration, outcomes, and associated variables of vaccine-related adenopathy. METHODS: In this retrospective cohort study, our database was queried for patients who underwent breast MRI following COVID-19 vaccination from January 22, 2021, to March 21, 2021. The frequency of ipsilateral axillary adenopathy and possible associated variables were evaluated, including age, personal history of ipsilateral breast cancer, clinical indication for breast MRI, type of vaccine, side of vaccination, number of doses, and number of days between the vaccine and the MRI exam. The outcomes of the adenopathy were investigated, including the duration of adenopathy and biopsy results. RESULTS: A total of 357 patients were included. The frequency of adenopathy on breast MRI was 29% (104/357 patients). Younger patients and shorter time intervals from the second dose of the vaccine were significantly associated with the development of adenopathy (p = 0.002 for both). Most adenopathy resolved or decreased on follow-up, with 11% of patients presenting persistence of adenopathy up to 64 days after the second dose of the vaccine. Metastatic axillary carcinoma was diagnosed in three patients; all three had a current ipsilateral breast cancer diagnosis. CONCLUSIONS: Vaccine-related adenopathy is a frequent event after COVID-19 vaccination; short-term follow-up is an appropriate clinical approach, except in patients with current ipsilateral breast cancer. Adenopathy may often persist 4-8 weeks after the second dose of the vaccine, thus favoring longer follow-up periods. KEY POINTS: • MRI-detected ipsilateral axillary adenopathy is a frequent benign finding after mRNA COVID-19 vaccination. • Axillary adenopathy following COVID-19 vaccination often persists > 4 weeks after vaccination, favoring longer follow-up periods. • In patients with concurrent ipsilateral breast cancer, axillary adenopathy can represent metastatic carcinoma and follow-up is not appropriate.

Assessment of Artificial Intelligence in Echocardiography Diagnostics in Differentiating Takotsubo Syndrome From Myocardial Infarction.

Importance: Machine learning algorithms enable the automatic classification of cardiovascular diseases based on raw cardiac ultrasound imaging data. However, the utility of machine learning in distinguishing between takotsubo syndrome (TTS) and acute myocardial infarction (AMI) has not been studied. Objectives: To assess the utility of machine learning systems for automatic discrimination of TTS and AMI. Design, Settings, and Participants: This cohort study included clinical data and transthoracic echocardiogram results of patients with AMI from the Zurich Acute Coronary Syndrome Registry and patients with TTS obtained from 7 cardiovascular centers in the International Takotsubo Registry. Data from the validation cohort were obtained from April 2011 to February 2017. Data from the training cohort were obtained from March 2017 to May 2019. Data were analyzed from September 2019 to June 2021. Exposure: Transthoracic echocardiograms of 224 patients with TTS and 224 patients with AMI were analyzed. Main Outcomes and Measures: Area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity of the machine learning system evaluated on an independent data set and 4 practicing cardiologists for comparison. Echocardiography videos of 228 patients were used in the development and training of a deep learning model. The performance of the automated echocardiogram video analysis method was evaluated on an independent data set consisting of 220 patients. Data were matched according to age, sex, and ST-segment elevation/non-ST-segment elevation (1 patient with AMI for each patient with TTS). Predictions were compared with echocardiographic-based interpretations from 4 practicing cardiologists in terms of sensitivity, specificity, and AUC calculated from confidence scores concerning their binary diagnosis. Results: In this cohort study, apical 2-chamber and 4-chamber echocardiographic views of 110 patients with TTS (mean [SD] age, 68.4 [12.1] years; 103 [90.4%] were female) and 110 patients with AMI (mean [SD] age, 69.1 [12.2] years; 103 [90.4%] were female) from an independent data set were evaluated. This approach achieved a mean (SD) AUC of 0.79 (0.01) with an overall accuracy of 74.8 (0.7%). In comparison, cardiologists achieved a mean (SD) AUC of 0.71 (0.03) and accuracy of 64.4 (3.5%) on the same data set. In a subanalysis based on 61 patients with apical TTS and 56 patients with AMI due to occlusion of the left anterior descending coronary artery, the model achieved a mean (SD) AUC score of 0.84 (0.01) and an accuracy of 78.6 (1.6%), outperforming the 4 practicing cardiologists (mean [SD] AUC, 0.72 [0.02]) and accuracy of 66.9 (2.8%). Conclusions and Relevance: In this cohort study, a real-time system for fully automated interpretation of echocardiogram videos was established and trained to differentiate TTS from AMI. While this system was more accurate than cardiologists in echocardiography-based disease classification, further studies are warranted for clinical application.

Impact of ComBat Harmonization on PET Radiomics-Based Tissue Classification: A Dual-Center PET/MRI and PET/CT Study.

Our purpose was to determine whether ComBat harmonization improves 18F-FDG PET radiomics-based tissue classification in pooled PET/MRI and PET/CT datasets. Methods: Two hundred patients who had undergone 18F-FDG PET/MRI (2 scanners and vendors; 50 patients each) or PET/CT (2 scanners and vendors; 50 patients each) were retrospectively included. Gray-level histogram, gray-level cooccurrence matrix, gray-level run-length matrix, gray-level size-zone matrix, and neighborhood gray-tone difference matrix radiomic features were calculated for volumes of interest in the disease-free liver, spleen, and bone marrow. For individual feature classes and a multiclass radiomic signature, tissue was classified on ComBat-harmonized and unharmonized pooled data, using a multilayer perceptron neural network. Results: Median accuracies in training and validation datasets were 69.5% and 68.3% (harmonized), respectively, versus 59.5% and 58.9% (unharmonized), respectively, for gray-level histogram; 92.1% and 86.1% (harmonized), respectively, versus 53.6% and 50.0% (unharmonized), respectively, for gray-level cooccurrence matrix; 84.8% and 82.8% (harmonized), respectively, versus 62.4% and 58.3% (unharmonized), respectively, for gray-level run-length matrix; 87.6% and 85.6% (harmonized), respectively, versus 56.2% and 52.8% (unharmonized), respectively, for gray-level size-zone matrix; 79.5% and 77.2% (harmonized), respectively, versus 54.8% and 53.9% (unharmonized), respectively, for neighborhood gray-tone difference matrix; and 86.9% and 84.4% (harmonized), respectively, versus 62.9% and 58.3% (unharmonized), respectively, for radiomic signature. Conclusion: ComBat harmonization may be useful for multicenter 18F-FDG PET radiomics studies using pooled PET/MRI and PET/CT data.

Prognostic Utility of MRI Features in Intradiverticular Bladder Tumor.

BACKGROUND: Intradiverticular bladder tumors (IDBT) are rare but clinically important, as they are difficult to assess endoscopically due to limited anatomic access and risk of perforation. MRI may be helpful in assessing IDBT and providing relevant staging and prognostic information. PURPOSE: To assess MRI findings of IDBT and their relationship with overall survival. METHODS: This retrospective study included 31 consecutive patients with IDBT undergoing MRI from 2008 to 2018 identified through electronic medical records and PACS database search. Two radiologists independently assessed the following MRI features: size (>3 vs ≤3 cm), diverticular neck involvement, Vesical Imaging-Reporting and Data System (VI-RADS) score (>3 vs ≤3), perivesical fat infiltration, additional tumors and suspicious pelvic lymph nodes. Overall survival was estimated using Kaplan-Meier analysis; and the relationship with clinicopathological and MRI features was determined using the Cox proportional-hazards regression model. Inter-reader agreement was assessed using intraclass correlation coefficients (ICC) and Cohen's kappa (K). RESULTS: Median follow-up was 1044 days (interquartile range, 474-1952 days). Twenty-six (83.9%) patients underwent surgical treatment with or without neoadjuvant chemotherapy. On MRI, greater tumor size (>3 cm), diverticular neck involvement, perivesical extension, and suspicious lymph nodes were associated with lower overall survival (HR = 3.6-8.1 and 4.3-6.3 for the 2 radiologists, p ≤ 0.03). Other clinicopathological or MRI findings were not associated with survival (p = 0.27-0.65). Inter-reader agreement was excellent for tumor size (ICC = 0.991; 95% CI 0.982-0.996), fair for VI-RADS (K = 0.52, 95% CI, 0.22-0.82), and moderate for others (K = 0.61-0.79). CONCLUSION: In patients with IDBT, several MRI features were significantly associated with overall survival. Utilizing all available clinicopathological and imaging information may improve estimation of prognosis.

Automatic Forecasting of Radiology Examination Volume Trends for Optimal Resource Planning and Allocation.

The aim of the study was to evaluate the performance of the Prophet forecasting procedure, part of the Facebook open-source Artificial Intelligence portfolio, for forecasting variations in radiological examination volumes. Daily CT and MRI examination volumes from our institution were extracted from the radiology information system (RIS) database. Data from January 1, 2015, to December 31, 2019, was used for training the Prophet algorithm, and data from January 2020 was used for validation. Algorithm performance was then evaluated prospectively in February and August 2020. Total error and mean error per day were evaluated, and computational time was logged using different Markov chain Monte Carlo (MCMC) samples. Data from 610,570 examinations were used for training; the majority were CTs (82.3%). During retrospective testing, prediction error was reduced from 19 to < 1 per day in CT (total 589 to 17) and from 5 to < 1 per day (total 144 to 27) in MRI by fine-tuning the Prophet procedure. Prospective prediction error in February was 11 per day in CT (9934 predicted, 9667 actual) and 1 per day in MRI (2484 predicted, 2457 actual) and was significantly better than manual weekly predictions (p = 0.001). Inference with MCMC added no substantial improvements while vastly increasing computational time. Prophet accurately models weekly, seasonal, and overall trends paving the way for optimal resource allocation for radiology exam acquisition and interpretation.