Evaluation of acute pulmonary embolism and clot burden on CTPA with deep learning

Unveiling the potential of artificial intelligence in revolutionizing disease diagnosis and prediction: a comprehensive review of machine learning and deep learning approaches

The rapid advancement of Machine Learning (ML) and Deep Learning (DL) technologies has revolutionized healthcare, particularly in the domains of disease prediction and diagnosis. This study provides a comprehensive review of ML and DL applications across sixteen diverse diseases, synthesizing findings from research conducted between 2015 and 2024. We explore these technologies' methodologies, effectiveness, and clinical outcomes, highlighting their transformative potential in healthcare settings. Although ML and DL demonstrate remarkable accuracy and efficiency in disease prediction and diagnosis, challenges including quality of data, interpretability of models, and their integration into clinical workflows remain significant barriers. By evaluating advanced approaches and their outcomes, this review not only underscores the current capabilities of ML and DL but also identifies key areas for future research. Ultimately, this work aims to serve as a roadmap for advancing healthcare practices, enhancing clinical decision making, and strengthening patient outcomes through the effective and responsible implementation of AI-driven technologies.

Meta-analysis of AI-based pulmonary embolism detection: How reliable are deep learning models?

RATIONALE AND OBJECTIVES

Deep learning (DL)-based methods show promise in detecting pulmonary embolism (PE) on CT pulmonary angiography (CTPA), potentially improving diagnostic accuracy and workflow efficiency. This meta-analysis aimed to (1) determine pooled performance estimates of DL algorithms for PE detection; and (2) compare the diagnostic efficacy of convolutional neural network (CNN)- versus U-Net-based architectures.

MATERIALS AND METHODS

Following PRISMA guidelines, we searched PubMed and EMBASE through April 15, 2025 for English-language studies (2010-2025) reporting DL models for PE detection with extractable 2 × 2 data or performance metrics. True/false positives and negatives were reconstructed when necessary under an assumed 50 % PE prevalence (with 0.5 continuity correction). We approximated AUROC as the mean of sensitivity and specificity if not directly reported. Sensitivity, specificity, accuracy, PPV and NPV were pooled using a DerSimonian-Laird random-effects model with Freeman-Tukey transformation; AUROC values were combined via a fixed-effect inverse-variance approach. Heterogeneity was assessed by Cochran's Q and I2. Subgroup analyses contrasted CNN versus U-Net models.

RESULTS

Twenty-four studies (n = 22,984 patients) met inclusion criteria. Pooled estimates were: AUROC 0.895 (95 % CI: 0.874-0.917), sensitivity 0.894 (0.856-0.923), specificity 0.871 (0.831-0.903), accuracy 0.857 (0.833-0.882), PPV 0.832 (0.794-0.869) and NPV 0.902 (0.874-0.929). Between-study heterogeneity was high (I2 ≈ 97 % for sensitivity/specificity). U-Net models exhibited higher sensitivity (0.899 vs 0.893) and CNN models higher specificity (0.926 vs 0.900); subgroup Q-tests confirmed significant differences for both sensitivity (p = 0.0002) and specificity (p < 0.001).

CONCLUSIONS

DL algorithms demonstrate high diagnostic accuracy for PE detection on CTPA, with complementary strengths: U-Net architectures excel in true-positive identification, whereas CNNs yield fewer false positives. However, marked heterogeneity underscores the need for standardized, prospective validation before routine clinical implementation.

Pulmonary Embolism Survival Prediction Using Multimodal Learning Based on Computed Tomography Angiography and Clinical Data

PURPOSE

Pulmonary embolism (PE) is a significant cause of mortality in the United States. The objective of this study is to implement deep learning (DL) models using computed tomography pulmonary angiography (CTPA), clinical data, and PE Severity Index (PESI) scores to predict PE survival.

MATERIALS AND METHODS

In total, 918 patients (median age 64 y, range 13 to 99 y, 48% male) with 3978 CTPAs were identified via retrospective review across 3 institutions. To predict survival, an AI model was used to extract disease-related imaging features from CTPAs. Imaging features and clinical variables were then incorporated into independent DL models to predict survival outcomes. Cross-modal fusion CoxPH models were used to develop multimodal models from combinations of DL models and calculated PESI scores. Five multimodal models were developed as follows: (1) using CTPA imaging features only, (2) using clinical variables only, (3) using both CTPA and clinical variables, (4) using CTPA and PESI score, and (5) using CTPA, clinical variables, and PESI score. Performance was evaluated using the concordance index (c-index). Kaplan-Meier analysis was performed to stratify patients into high-risk and low-risk groups. Additional factor-risk analysis was conducted to account for right ventricular (RV) dysfunction.

RESULTS

For both data sets, the multimodal models incorporating CTPA features, clinical variables, and PESI score achieved higher c-indices than PESI alone. Following the stratification of patients into high-risk and low-risk groups by models, survival outcomes differed significantly (both P<0.001). A strong correlation was found between high-risk grouping and RV dysfunction.

CONCLUSIONS

Multiomic DL models incorporating CTPA features, clinical data, and PESI achieved higher c-indices than PESI alone for PE survival prediction.

Application of Artificial Intelligence in Thoracic Radiology: A Narrative Review

Thoracic radiology has emerged as a primary field in which artificial intelligence (AI) is extensively researched. Recent advancements highlight the potential to enhance radiologists' performance through AI. AI aids in detecting and classifying abnormalities, and in quantifying both normal and abnormal anatomical structures. Additionally, it facilitates prognostication by leveraging these quantitative values. This review article will discuss the recent achievements of AI in thoracic radiology, focusing primarily on deep learning, and explore the current limitations and future directions of this cutting-edge technique.

The promise and limitations of artificial intelligence in CTPA-based pulmonary embolism detection

Computed tomography pulmonary angiography (CTPA) is an essential diagnostic tool for identifying pulmonary embolism (PE). The integration of AI has significantly advanced CTPA-based PE detection, enhancing diagnostic accuracy and efficiency. This review investigates the growing role of AI in the diagnosis of pulmonary embolism using CTPA imaging. The review examines the capabilities of AI algorithms, particularly deep learning models, in analyzing CTPA images for PE detection. It assesses their sensitivity and specificity compared to human radiologists. AI systems, using large datasets and complex neural networks, demonstrate remarkable proficiency in identifying subtle signs of PE, aiding clinicians in timely and accurate diagnosis. In addition, AI-powered CTPA analysis shows promise in risk stratification, prognosis prediction, and treatment optimization for PE patients. Automated image interpretation and quantitative analysis facilitate rapid triage of suspected cases, enabling prompt intervention and reducing diagnostic delays. Despite these advancements, several limitations remain, including algorithm bias, interpretability issues, and the necessity for rigorous validation, which hinder widespread adoption in clinical practice. Furthermore, integrating AI into existing healthcare systems requires careful consideration of regulatory, ethical, and legal implications. In conclusion, AI-driven CTPA-based PE detection presents unprecedented opportunities to enhance diagnostic precision and efficiency. However, addressing the associated limitations is critical for safe and effective implementation in routine clinical practice. Successful utilization of AI in revolutionizing PE care necessitates close collaboration among researchers, medical professionals, and regulatory organizations.

Imaging of Acute Pulmonary Embolism: An Update

Imaging is essential in the evaluation and management of acute pulmonary embolism. Advances in multi-energy CT including dual-energy CT and photon-counting CT have allowed faster scans with lower radiation dose and optimal quality. Artificial intelligence has a potential role in triaging potentially positive examinations and could serve as a second reader.

Applications of Artificial Intelligence in Acute Thoracic Imaging

The applications of artificial intelligence (AI) in radiology are rapidly advancing with AI algorithms being used in a wide range of disease pathologies and clinical settings. Acute thoracic pathologies including rib fractures, pneumothoraces, and acute PE are associated with significant morbidity and mortality and their identification is crucial for prompt treatment. AI models which increase diagnostic accuracy, improve radiologist efficiency and reduce time to diagnosis of acute abnormalities in the thorax have the potential to significantly improve patient outcomes. The purpose of this review is to summarize the current applications of AI in acute thoracic imaging, highlighting their strengths, limitations, and future research opportunities.

Prediction of short-term adverse clinical outcomes of acute pulmonary embolism using conventional machine learning and deep Learning based on CTPA images

To explore the predictive value of traditional machine learning (ML) and deep learning (DL) algorithms based on computed tomography pulmonary angiography (CTPA) images for short-term adverse outcomes in patients with acute pulmonary embolism (APE). This retrospective study enrolled 132 patients with APE confirmed by CTPA. Thrombus segmentation and texture feature extraction was performed using 3D-Slicer software. The least absolute shrinkage and selection operator (LASSO) algorithm was used for feature dimensionality reduction and selection, with optimal λ values determined using leave-one-fold cross-validation to identify texture features with non-zero coefficients. ML models (logistic regression, random forest, decision tree, support vector machine) and DL models (ResNet 50 and Vgg 19) were used to construct the prediction models. Model performance was evaluated using receiver operating characteristic (ROC) curves and the area under the curve (AUC). The cohort included 84 patients in the good prognosis group and 48 patients in the poor prognosis group. Univariate and multivariate logistic regression analyses showed that diabetes, RV/LV ≥ 1.0, and Qanadli index form independent risk factors predicting poor prognosis in patients with APE(P < 0.05). A total of 750 texture features were extracted, with 4 key features identified through screening. There was a weak positive correlation between texture features and clinical parameters. ROC curves analysis demonstrated AUC values of 0.85 (0.78-0.92), 0.76 (0.67-0.84), and 0.89 (0.83-0.95) for the clinical, texture feature, and combined models, respectively. In the ML models, the random forest model achieved the highest AUC (0.85), and the support vector machine model achieved the lowest AUC (0.62). And the AUCs for the DL models (ResNet 50 and Vgg 19) were 0.91 (95%CI: 0.90-0.92) and 0.94(95%CI: 0.93-0.95), respectively. Vgg 19 model demonstrated exceptional precision (0.93), recall (0.76), specificity (0.95) and F1 score (0.84). Both ML and DL models based on thrombus texture features from CTPA images demonstrated higher predictive efficacy for short-term adverse outcomes in patients with APE, especially the random forest and Vgg 19 models, potentially assisting clinical management in timely interventions to improve patient prognosis.

Development of the Pulmonary Embolism Progression (PEP) score for predicting short-term clinical deterioration in intermediate-risk pulmonary embolism: a single-center retrospective study

Accurate risk stratification in acute intermediate-risk pulmonary embolism (PE) is essential. Current prediction scores lack the ability to forecast impending clinical decline. The Pulmonary Embolism Progression (PEP) score aims to predict short-term clinical deterioration (respiratory failure or hemodynamic instability within 72 h) in patients with intermediate-risk PE. This single-center retrospective cohort study analyzed patients with intermediate PE. The outcome of interest was respiratory failure or hemodynamic instability within 72 h. A multivariate logistic regression identified five predictive variables for the final PEP score: use of > 4 L/min of supplemental oxygen above baseline, lactate > 2.0 mmol/L, high-sensitivity cardiac troponin T (hs-cTnT) > 40 ng/L, tricuspid annular plane systolic excursion (TAPSE) < 13 mm, and the combination of central and subsegmental clot. The derivation cohort included 117 patients, and the validation cohort included 70 patients. The area under the receiver operating characteristic (AUROC) curve for the derivation cohort was 0.8671 (95% CI: 0.7946, 0.9292), and for the validation cohort, it was 0.9264 (95% CI: 0.8680, 0.9847). A PEP score of 4 points yielded the highest combination of sensitivity (93%) and specificity (65%). Each incremental point increase in the PEP score raised the probability of clinical deterioration by a factor of 1.933. The PEP score is a reliable tool for predicting the likelihood of clinical deterioration in intermediate-risk PE patients within 72 h, potentially aiding in timely clinical decision-making and improving patient outcomes.

Prediction of pulmonary embolism by an explainable machine learning approach in the real world

In recent years, large amounts of researches showed that pulmonary embolism (PE) has become a common disease, and PE remains a clinical challenge because of its high mortality, high disability, high missed and high misdiagnosed rates. To address this, we employed an artificial intelligence-based machine learning algorithm (MLA) to construct a robust predictive model for PE. We retrospectively analyzed 1480 suspected PE patients hospitalized in West China Hospital of Sichuan University between May 2015 and April 2020. 126 features were screened and diverse MLAs were utilized to craft predictive models for PE. Area under the receiver operating characteristic curves (AUC) were used to evaluate their performance and SHapley Additive exPlanation (SHAP) values were utilized to elucidate the prediction model. Regarding the efficacy of the single model that most accurately predicted the outcome, RF demonstrated the highest efficacy in predicting outcomes, with an AUC of 0.776 (95% CI 0.774-0.778). The SHAP summary plot delineated the positive and negative effects of features attributed to the RF prediction model, including D-dimer, activated partial thromboplastin time (APTT), fibrin and fibrinogen degradation products (FFDP), platelet count, albumin, cholesterol, and sodium. Furthermore, the SHAP dependence plot illustrated the impact of individual features on the RF prediction model. Finally, the MLA based PE predicting model was designed as a web page that can be applied to the platform of clinical management. In this study, PE prediction model was successfully established and designed as a web page, facilitating the optimization of early diagnosis and timely treatment strategies to enhance PE patient outcomes.

SPE-YOLO: A deep learning model focusing on small pulmonary embolism detection

OBJECTIVES

By developing the deep learning model SPE-YOLO, the detection of small pulmonary embolism has been improved, leading to more accurate identification of this condition. This advancement aims to better serve medical diagnosis and treatment.

METHODS

This retrospective study analyzed images of 142 patients from Tianjin Medical University General Hospital using YOLOv8 as the foundational framework. Firstly, a small detection head P2 was added to better capture and identify small targets. Secondly, the SEAttention mechanism was integrated into the model to enhance focus on crucial features and optimize detection accuracy. Lastly, the feature extraction process was refined by introducing ODConv convolution to capture more comprehensive contextual information, thereby enhancing the detection performance of small pulmonary embolisms. The model's practical application ability was evaluated using 2000 cases from the RSNA dataset as an external test set.

RESULTS

In the Tianjin test set, our model achieved a precision of 84.20 % and an accuracy of 81.50 %. This represents an improvement of approximately 5 % and 4 % respectively compared to the original YOLOv8. F1 scores, recall rates and average accuracy have also increased by 4 %, 5 %, 6 %, respectively. In data from the external validation set of RSNA, SPE-YOLO exhibited its effectiveness with a sensitivity of 90.70 % and an accuracy of 86.45 %.

CONCLUSION

The SPE-YOLO algorithm demonstrates strong capability in identifying small pulmonary embolisms, offering clinicians a more accurate and efficient diagnostic tool. This advancement is expected to enhance the quality of pulmonary embolism diagnosis and support the progress of medical services.

Current Role of CT Pulmonary Angiography in Pulmonary Embolism: A State-of-the-Art Review

The purpose of this study is to conduct a literature review on the current role of computed tomography pulmonary angiography (CTPA) in the diagnosis and prognosis of pulmonary embolism (PE). It addresses key topics such as the quantification of the thrombotic burden, its role as a predictor of mortality, new diagnostic techniques that are available, the possibility of analyzing the thrombus composition to differentiate its evolutionary stage, and the applicability of artificial intelligence (AI) in PE through CTPA. The only finding from CTPA that has been validated as a prognostic factor so far is the right ventricle/left ventricle (RV/LV) diameter ratio being >1, which is associated with a 2.5-fold higher risk of all-cause mortality or adverse events, and a 5-fold higher risk of PE-related mortality. The increasing use of techniques such as dual-energy computed tomography allows for the more accurate diagnosis of perfusion defects, which may go undetected in conventional computed tomography, identifying up to 92% of these defects compared to 78% being detected by CTPA. Additionally, it is essential to explore the latest advances in the application of AI to CTPA, which are currently expanding and have demonstrated a 23% improvement in the detection of subsegmental emboli compared to manual interpretation. With deep image analysis, up to a 95% accuracy has been achieved in predicting PE severity based on the thrombus volume and perfusion deficits. These advancements over the past 10 years significantly contribute to early intervention strategies and, therefore, to the improvement of morbidity and mortality outcomes for these patients.

Heart lung axis in acute pulmonary embolism: Role of CT in risk stratification

Pulmonary embolism (PE) remains a significant cause of mortality requiring prompt diagnosis and risk stratification. This review focuses on the role of computed tomography (CT) in the risk stratification of acute PE, highlighting its impact on patient management. We will explore basic pathophysiology of pulmonary embolism (PE) and review current guidelines, which will help radiologists interpret images within a broader clinical context. This review covers key CT findings which can be used for risk stratification including indicators of right ventricular (RV) dysfunction, clot burden, clot location and left atrial volume. We will discuss the measurement of RV/LV diameter ratio as a key indicator of RV dysfunction and its limitations and challenges within various patient populations. While these parameters should be included in a radiologist's report, their predictive value for mortality depends on the patient's existing cardiopulmonary reserve and should not be interpreted in isolation.

CT-Derived Features as Predictors of Clot Burden and Resolution

Objectives: To evaluate the prognostic utility of CT-imaging-derived biomarkers in distinguishing acute pulmonary embolism (PE) resolution and its progression to chronic PE, as well as their association with clot burden. Materials and Methods: We utilized a cohort of 45 patients (19 male (42.2%)) and 96 corresponding CT scans with exertional dyspnea following an acute PE. These patients were referred for invasive cardiopulmonary exercise testing (CPET) at the University of Pittsburgh Medical Center from 2018 to 2022, for whom we have ground truth classification of chronic PE, as well as CT-derived features related to body composition, cardiopulmonary vasculature, and PE clot burden using artificial intelligence (AI) algorithms. We applied Lasso regularization to select parameters, followed by (1) Ordinary Least Squares (OLS) regressions to analyze the relationship between clot burden and the selected parameters and (2) logistic regressions to differentiate between chronic and resolved patients. Results: Several body composition and cardiopulmonary factors showed statistically significant association with clot burden. A multivariate model based on cardiopulmonary features demonstrated superior performance in predicting PE resolution (AUC: 0.83, 95% CI: 0.71-0.95), indicating significant associations between airway ratio (negative correlation), aorta diameter, and heart volume (positive correlation) with PE resolution. Other multivariate models integrating demographic features showed comparable performance, while models solely based on body composition and baseline clot burden demonstrated inferior performance. Conclusions: Our analysis suggests that cardiopulmonary and demographic features hold prognostic value for predicting PE resolution, whereas body composition and baseline clot burden do not. Clinical Relevance: Our identified prognostic factors may facilitate the follow-up procedures for patients diagnosed with acute PE.

Performance and clinical utility of an artificial intelligence-enabled tool for pulmonary embolism detection

PURPOSE

Diagnosing pulmonary embolism (PE) is still challenging due to other conditions that can mimic its appearance, leading to incomplete or delayed management and several inter-observer variabilities. This study evaluated the performance and clinical utility of an artificial intelligence (AI)-based application designed to assist clinicians in the detection of PE on CT pulmonary angiography (CTPA).

PATIENTS AND METHODS

CTPAs from 230 US cities acquired on 57 scanner models from 6 different vendors were retrospectively collected. Three US board certified expert radiologists defined the ground truth by majority agreement. The same cases were analyzed by CINA-PE, an AI-driven algorithm capable of detecting and highlighting suspected PE locations. The algorithm's performance at a per-case and per-finding level was evaluated. Furthermore, cases with PE not mentioned in the clinical report but correctly detected by the algorithm were analyzed.

RESULTS

A total of 1204 CTPAs (mean age 62.1 years ± 16.6[SD], 44.4 % female, 14.9 % positive) were included in the study. Per-case sensitivity and specificity were 93.9 % (95%CI: 89.3 %-96.9 %) and 94.8 % (95%CI: 93.3 %-96.1 %), respectively. Per-finding positive predictive value was 89.5 % (95%CI: 86.7 %-91.9 %). Among the 196 positive cases, 29 (15.6 %) were not mentioned in the clinical report. The algorithm detected 22/29 (76 %) of these cases, leading to a reduction in the miss rate from 15.6 % to 3.8 % (7/186).

CONCLUSIONS

The AI-based application may improve diagnostic accuracy in detecting PE and enhance patient outcomes through timely intervention. Integrating AI tools in clinical workflows can reduce missed or delayed diagnoses, and positively impact healthcare delivery and patient care.

PE-Ynet: a novel attention-based multi-task model for pulmonary embolism detection using CT pulmonary angiography (CTPA) scan images

Pulmonary Embolism (PE) has diverse manifestations with different etiologies such as venous thromboembolism, septic embolism, and paradoxical embolism. In this study, a novel attention-based multi-task model is proposed for PE segmentation and detection from Computed Tomography Pulmonary Angiography (CTPA) images. A Y-Net architecture is used to implement this model, which facilitates segmentation and classification jointly, improving performance and efficiency. It is leveraged with Multi Head Attention (MHA), which allows the model to focus on important regions of the image while suppressing irrelevant information, improving the accuracy of the segmentation and detection tasks. The proposed PE-YNet model is tested with two public datasets, achieving a maximum mean detection and segmentation accuracy of 99.89% and 99.83%, respectively, on the CAD-PE challenge dataset. Similarly, it also achieves a detection accuracy of 99.75% and a segmentation accuracy of 99.81% on the FUMPE dataset. Additionally, sensitivity analysis also shows a high sensitivity of 0.9885 for the localization error ɛ = 0 for the CAD-PE dataset, demonstrating the model's robustness against false predictions compared to state-of-the-art models. Further, this model also exhibits lower inference time, size, and memory usage compared to representative models. An automated PE-YNet tool can assist physicians with PE diagnosis, treatment, and prognosis monitoring in the clinical management of CoVID-19.

nnU-Net-based deep-learning for pulmonary embolism: detection, clot volume quantification, and severity correlation in the RSPECT dataset

OBJECTIVES

CT pulmonary angiography is the gold standard for diagnosing pulmonary embolism, and DL algorithms are being developed to manage the increase in demand. The nnU-Net is a new auto-adaptive DL framework that minimizes manual tuning, making it easier to develop effective algorithms for medical imaging even without specific expertise. This study assesses the performance of a locally developed nnU-Net algorithm on the RSPECT dataset for PE detection, clot volume measurement, and correlation with right ventricle overload.

MATERIALS & METHODS

User input was limited to segmentation using 3DSlicer. We worked with the RSPECT dataset and trained an algorithm from 205 PE and 340 negatives. The test dataset comprised 6573 exams. Performance was tested against PE characteristics, such as central, non-central, and RV overload. Blood clot volume (BCV) was extracted from each exam. We employed ROC curves and logistic regression for statistical validation.

RESULTS

Negative studies had a median BCV of 1 μL, which increased to 345 μL in PE-positive cases and 7,378 μL in central PEs. Statistical analysis confirmed a significant BCV correlation with PE presence, central PE, and increased RV/LV ratio (p < 0.0001). The model's AUC for PE detection was 0.865, with an 83 % accuracy at a 55 μL threshold. Central PE detection AUC was 0.937 with 91 % accuracy at 850 μL. The RV overload AUC stood at 0.848 with 79 % accuracy.

CONCLUSION

The nnU-Net algorithm demonstrated accurate PE detection, particularly for central PE. BCV is an accurate metric for automated severity stratification and case prioritization.

CLINICAL RELEVANCE STATEMENT

The nnU-Net framework can be utilized to create a dependable DL for detecting PE. It offers a user-friendly approach to those lacking expertise in AI and rapidly extracts the Blood Clot Volume, a metric that can evaluate the PE's severity.

Automatic detection of pulmonary embolism on computed tomography pulmonary angiogram scan using a three-dimensional convolutional neural network

OBJECTIVE

To propose a convolutional neural network (EmbNet) for automatic pulmonary embolism detection on computed tomography pulmonary angiogram (CTPA) scans and to assess its diagnostic performance.

METHODS

305 consecutive CTPA scans between January 2019 and December 2021 were enrolled in this study (142 for training, 163 for internal validation), and 250 CTPA scans from a public dataset were used for external validation. The framework comprised a preprocessing step to segment the pulmonary vessels and the EmbNet to detect emboli. Emboli were divided into three location-based subgroups for detailed evaluation: central arteries, lobar branches, and peripheral regions. Ground truth was established by three radiologists.

RESULTS

The EmbNet's per-scan level sensitivity, specificity, positive predictive value (PPV), and negative predictive value were 90.9%, 75.4%, 48.4%, and 97.0% (internal validation) and 88.0%, 70.5%, 42.7%, and 95.9% (external validation). At the per-embolus level, the overall sensitivity and PPV of the EmbNet were 86.0% and 61.3% (internal validation), and 83.5% and 57.5% (external validation). The sensitivity and PPV of central emboli were 89.7% and 52.0% (internal validation), and 94.4% and 43.0% (external validation); of lobar emboli were 95.2% and 76.9% (internal validation), and 93.5% and 72.5% (external validation); and of peripheral emboli were 82.6% and 61.7% (internal validation), and 80.2% and 59.4% (external validation). The average false positive rate was 0.45 false emboli per scan (internal validation) and 0.69 false emboli per scan (external validation).

CONCLUSION

The EmbNet provides high sensitivity across embolus locations, suggesting its potential utility for initial screening in clinical practice.

AI-based automated segmentation for ovarian/adnexal masses and their internal components on ultrasound imaging

Purpose

Segmentation of ovarian/adnexal masses from surrounding tissue on ultrasound images is a challenging task. The separation of masses into different components may also be important for radiomic feature extraction. Our study aimed to develop an artificial intelligence-based automatic segmentation method for transvaginal ultrasound images that (1) outlines the exterior boundary of adnexal masses and (2) separates internal components.

Approach

A retrospective ultrasound imaging database of adnexal masses was reviewed for exclusion criteria at the patient, mass, and image levels, with one image per mass. The resulting 54 adnexal masses (36 benign/18 malignant) from 53 patients were separated by patient into training (26 benign/12 malignant) and independent test (10 benign/6 malignant) sets. U-net segmentation performance on test images compared to expert detailed outlines was measured using the Dice similarity coefficient (DSC) and the ratio of the Hausdorff distance to the effective diameter of the outline (R HD - D ) for each mass. Subsequently, in discovery mode, a two-level fuzzy c-means (FCM) unsupervised clustering approach was used to separate the pixels within masses belonging to hypoechoic or hyperechoic components.

Results

The DSC (median [95% confidence interval]) was 0.91 [0.78, 0.96], andR HD - D was 0.04 [0.01, 0.12], indicating strong agreement with expert outlines. Clinical review of the internal separation of masses into echogenic components demonstrated a strong association with mass characteristics.

Conclusion

A combined U-net and FCM algorithm for automatic segmentation of adnexal masses and their internal components achieved excellent results compared with expert outlines and review, supporting future radiomic feature-based classification of the masses by components.

Modern imaging of acute pulmonary embolism

The first-choice imaging test for visualization of thromboemboli in the pulmonary vasculature in patients with suspected acute pulmonary embolism (PE) is multidetector computed tomography pulmonary angiography (CTPA) - a readily available and widely used imaging technique. Through technological advancements over the past years, alternative imaging techniques for the diagnosis of PE have become available, whilst others are still under investigation. In particular, the evolution of artificial intelligence (AI) is expected to enable further innovation in diagnostic management of PE. In this narrative review, current CTPA techniques and the emerging technology photon-counting CT (PCCT), as well as other modern imaging techniques of acute PE are discussed, including CTPA with iodine maps based on subtraction or dual-energy acquisition, single-photon emission CT (SPECT), magnetic resonance angiography (MRA), and magnetic resonance direct thrombus imaging (MRDTI). Furthermore, potential applications of AI are discussed.

Machine learning in cancer-associated thrombosis: hype or hope in untangling the clot

The goal of machine learning (ML) is to create informative signals and useful tasks by leveraging large datasets to derive computational algorithms. ML has the potential to revolutionize the healthcare industry by boosting productivity, enhancing safe and effective patient care, and lightening the load on clinicians. In addition to gaining mechanistic insights into cancer-associated thrombosis (CAT), ML can be used to improve patient outcomes, streamline healthcare delivery, and spur innovation. Our review paper delves into the present and potential applications of this cutting-edge technology, encompassing three areas: i) computer vision-assisted diagnosis of thromboembolism from radiology data; ii) case detection from electronic health records using natural language processing; iii) algorithms for CAT prediction and risk stratification. The availability of large, well-annotated, high-quality datasets, overfitting, limited generalizability, the risk of propagating inherent bias, and a lack of transparency among patients and clinicians are among the challenges that must be overcome in order to effectively develop ML in the health sector. To guarantee that this powerful instrument can be utilized to maximize innovation in CAT, clinicians can collaborate with stakeholders such as computer scientists, regulatory bodies, and patient groups.

Anatomically aware dual-hop learning for pulmonary embolism detection in CT pulmonary angiograms

Pulmonary Embolisms (PE) represent a leading cause of cardiovascular death. While medical imaging, through computed tomographic pulmonary angiography (CTPA), represents the gold standard for PE diagnosis, it is still susceptible to misdiagnosis or significant diagnosis delays, which may be fatal for critical cases. Despite the recently demonstrated power of deep learning to bring a significant boost in performance in a wide range of medical imaging tasks, there are still very few published researches on automatic pulmonary embolism detection. Herein we introduce a deep learning based approach, which efficiently combines computer vision and deep neural networks for pulmonary embolism detection in CTPA. Our method brings novel contributions along three orthogonal axes: (1) automatic detection of anatomical structures; (2) anatomical aware pretraining, and (3) a dual-hop deep neural net for PE detection. We obtain state-of-the-art results on the publicly available multicenter large-scale RSNA dataset.

A deep learning-based algorithm improves radiology residents' diagnoses of acute pulmonary embolism on CT pulmonary angiograms

PURPOSE

To compare radiology residents' diagnostic performances to detect pulmonary emboli (PEs) on CT pulmonary angiographies (CTPAs) with deep-learning (DL)-based algorithm support and without.

METHODS

Fully anonymized CTPAs (n = 207) of patients suspected of having acute PE served as input for PE detection using a previously trained and validated DL-based algorithm. Three residents in their first three years of training, blinded to the index report and clinical history, read the CTPAs first without, and 2 months later with the help of artificial intelligence (AI) output, to diagnose PE as present, absent or indeterminate. We evaluated concordances and discordances with the consensus-reading results of two experts in chest imaging.

RESULTS

Because the AI algorithm failed to analyze 11 CTPAs, 196 CTPAs were analyzed; 31 (15.8 %) were PE-positive. Good-classification performance was higher for residents with AI-algorithm support than without (AUROCs: 0.958 [95 % CI: 0.921-0.979] vs. 0.894 [95 % CI: 0.850-0.931], p < 0.001, respectively). The main finding was the increased sensitivity of residents' diagnoses using the AI algorithm (92.5 % vs. 81.7 %, respectively). Concordance between residents (kappa: 0.77 [95 % CI: 0.76-0.78]; p < 0.001) improved with AI-algorithm use (kappa: 0.88 [95 % CI: 0.87-0.89]; p < 0.001).

CONCLUSION

The AI algorithm we used improved between-resident agreements to interpret CTPAs for suspected PE and, hence, their diagnostic performances.

[The contribution of artificial intelligence (AI) subsequent to the processing of thoracic imaging]

The contribution of artificial intelligence (AI) to medical imaging is currently the object of widespread experimentation. The development of deep learning (DL) methods, particularly convolution neural networks (CNNs), has led to performance gains often superior to those achieved by conventional methods such as machine learning. Radiomics is an approach aimed at extracting quantitative data not accessible to the human eye from images expressing a disease. The data subsequently feed machine learning models and produce diagnostic or prognostic probabilities. As for the multiple applications of AI methods in thoracic imaging, they are undergoing evaluation. Chest radiography is a practically ideal field for the development of DL algorithms able to automatically interpret X-rays. Current algorithms can detect up to 14 different abnormalities present either in isolation or in combination. Chest CT is another area offering numerous AI applications. Various algorithms have been specifically formed and validated for the detection and characterization of pulmonary nodules and pulmonary embolism, as well as segmentation and quantitative analysis of the extent of diffuse lung diseases (emphysema, infectious pneumonias, interstitial lung disease). In addition, the analysis of medical images can be associated with clinical, biological, and functional data (multi-omics analysis), the objective being to construct predictive approaches regarding disease prognosis and response to treatment.

Does Adding the Pulmonary Infarction and Right Ventricle to Left Ventricle Diameter Ratio to the Qanadli Index (A Combined Qanadli Index) More Accurately, Predict Short-Term Mortality in Patients with Pulmonary Embolism?

Background  The Qanadli index can be used to assess the severity of pulmonary arterial involvement in patients with acute pulmonary embolism. However, it seems that considering pulmonary infarction and right ventricle/left ventricle (RV/LV) ratio along with this index (called the combined Qanadli index) can provide a more accurate view of changes in cardiovascular parameters in these patients and help predict mortality in a better manner. In this regard, we evaluated the ability of the combined Qanadli index versus the Qanadli index in predicting short-term mortality in patients with pulmonary embolism. Methods  This retrospective study enrolled 234 patients with acute pulmonary embolism. Patients were divided into two groups: those who expired in 30 days and who survived. Then they were evaluated by computed tomography angiography of pulmonary arteries. The RV/LV diameter ratio and also pulmonary artery obstruction index (PAOI) were calculated. The patient's computed tomography scans were reviewed for pulmonary infarction. By adding the RV/LV ratio and pulmonary infarction to PAOI, a new index called the modified Qanadli score was made. Univariable and multivariable logistic regression was done for finding predictors of mortality. Results  Nine cases (40%) of patients in the mortality group and 42 (20%) of survivors had ischemic heart disease and the difference was significantly meaningful. The mean Qanadli index in the mortality group was 16.8 ± 8.45 and in survivors was 8.3 ± 4.2. By adding the pulmonary infarction score and PAOI score to RV/LV ratio score, the odds ratio (OR) for predicting mortality increased significantly to 13 and 16, respectively, which were significantly meaningful. Based on our findings, the highest OR for predicting short-term mortality was obtained through a combined Qanadli index (PAOI score + pulmonary infarction score + RV/LV score) that was 17 in univariable and 18 in multivariable logistic regression analysis ( p -value = 0.015). Conclusion  The new combined Qanadli index has more ability than the Qanadli index and RV/LV ratio for predicting changes in cardiovascular parameters and short-term mortality in patients with pulmonary embolism.

Pixel-level annotated dataset of computed tomography angiography images of acute pulmonary embolism

Pulmonary embolism has a high incidence and mortality, especially if undiagnosed. The examination of choice for diagnosing the disease is computed tomography pulmonary angiography. As many factors can lead to misinterpretations and diagnostic errors, different groups are utilizing deep learning methods to help improve this process. The diagnostic accuracy of these methods tends to increase by augmenting the training dataset. Deep learning methods can potentially benefit from the use of images acquired with devices from different vendors. To the best of our knowledge, we have developed the first public dataset annotated at the pixel and image levels and the first pixel-level annotated dataset to contain examinations performed with equipment from Toshiba and GE. This dataset includes 40 examinations, half performed with each piece of equipment, representing samples from two medical services. We also included measurements related to the cardiac and circulatory consequences of pulmonary embolism. We encourage the use of this dataset to develop, evaluate and compare the performance of new AI algorithms designed to diagnose PE.

Deep Learning-Based Algorithm for Automatic Detection of Pulmonary Embolism in Chest CT Angiograms

Purpose: Since the prompt recognition of acute pulmonary embolism (PE) and the immediate initiation of treatment can significantly reduce the risk of death, we developed a deep learning (DL)-based application aimed to automatically detect PEs on chest computed tomography angiograms (CTAs) and alert radiologists for an urgent interpretation. Convolutional neural networks (CNNs) were used to design the application. The associated algorithm used a hybrid 3D/2D UNet topology. The training phase was performed on datasets adequately distributed in terms of vendors, patient age, slice thickness, and kVp. The objective of this study was to validate the performance of the algorithm in detecting suspected PEs on CTAs. Methods: The validation dataset included 387 anonymized real-world chest CTAs from multiple clinical sites (228 U.S. cities). The data were acquired on 41 different scanner models from five different scanner makers. The ground truth (presence or absence of PE on CTA images) was established by three independent U.S. board-certified radiologists. Results: The algorithm correctly identified 170 of 186 exams positive for PE (sensitivity 91.4% [95% CI: 86.4–95.0%]) and 184 of 201 exams negative for PE (specificity 91.5% [95% CI: 86.8–95.0%]), leading to an accuracy of 91.5%. False negative cases were either chronic PEs or PEs at the limit of subsegmental arteries and close to partial volume effect artifacts. Most of the false positive findings were due to contrast agent-related fluid artifacts, pulmonary veins, and lymph nodes. Conclusions: The DL-based algorithm has a high degree of diagnostic accuracy with balanced sensitivity and specificity for the detection of PE on CTAs.

Artificial Intelligence Tool for Detection and Worklist Prioritization Reduces Time to Diagnosis of Incidental Pulmonary Embolism at CT

Purpose

To evaluate the diagnostic efficacy of artificial intelligence (AI) software in detecting incidental pulmonary embolism (IPE) at CT and shorten the time to diagnosis with use of radiologist reading worklist prioritization.

Materials and Methods

In this study with historical controls and prospective evaluation, regulatory-cleared AI software was evaluated to prioritize IPE on routine chest CT scans with intravenous contrast agent in adult oncology patients. Diagnostic accuracy metrics were calculated, and temporal end points, including detection and notification times (DNTs), were assessed during three time periods (April 2019 to September 2020): routine workflow without AI, human triage without AI, and worklist prioritization with AI.

Results

In total, 11 736 CT scans in 6447 oncology patients (mean age, 63 years ± 12 [SD]; 3367 men) were included. Prevalence of IPE was 1.3% (51 of 3837 scans), 1.4% (54 of 3920 scans), and 1.0% (38 of 3979 scans) for the respective time periods. The AI software detected 131 true-positive, 12 false-negative, 31 false-positive, and 11 559 true-negative results, achieving 91.6% sensitivity, 99.7% specificity, 99.9% negative predictive value, and 80.9% positive predictive value. During prospective evaluation, AI-based worklist prioritization reduced the median DNT for IPE-positive examinations to 87 minutes (vs routine workflow of 7714 minutes and human triage of 4973 minutes). Radiologists' missed rate of IPE was significantly reduced from 44.8% (47 of 105 scans) without AI to 2.6% (one of 38 scans) when assisted by the AI tool (P < .001).

Conclusion

AI-assisted workflow prioritization of IPE on routine CT scans in oncology patients showed high diagnostic accuracy and significantly shortened the time to diagnosis in a setting with a backlog of examinations.Keywords: CT, Computer Applications, Detection, Diagnosis, Embolism, Thorax, ThrombosisSupplemental material is available for this article.© RSNA, 2023See also the commentary by Elicker in this issue.

IoMT-Enabled Computer-Aided Diagnosis of Pulmonary Embolism from Computed Tomography Scans Using Deep Learning

The Internet of Medical Things (IoMT) has revolutionized Ambient Assisted Living (AAL) by interconnecting smart medical devices. These devices generate a large amount of data without human intervention. Learning-based sophisticated models are required to extract meaningful information from this massive surge of data. In this context, Deep Neural Network (DNN) has been proven to be a powerful tool for disease detection. Pulmonary Embolism (PE) is considered the leading cause of death disease, with a death toll of 180,000 per year in the US alone. It appears due to a blood clot in pulmonary arteries, which blocks the blood supply to the lungs or a part of the lung. An early diagnosis and treatment of PE could reduce the mortality rate. Doctors and radiologists prefer Computed Tomography (CT) scans as a first-hand tool, which contain 200 to 300 images of a single study for diagnosis. Most of the time, it becomes difficult for a doctor and radiologist to maintain concentration going through all the scans and giving the correct diagnosis, resulting in a misdiagnosis or false diagnosis. Given this, there is a need for an automatic Computer-Aided Diagnosis (CAD) system to assist doctors and radiologists in decision-making. To develop such a system, in this paper, we proposed a deep learning framework based on DenseNet201 to classify PE into nine classes in CT scans. We utilized DenseNet201 as a feature extractor and customized fully connected decision-making layers. The model was trained on the Radiological Society of North America (RSNA)-Pulmonary Embolism Detection Challenge (2020) Kaggle dataset and achieved promising results of 88%, 88%, 89%, and 90% in terms of the accuracy, sensitivity, specificity, and Area Under the Curve (AUC), respectively.

Automatic detection of pulmonary embolism in computed tomography pulmonary angiography using Scaled-YOLOv4

BACKGROUND

Pulmonary embolism (PE) is a common but fatal clinical condition and the gold standard of diagnosis is computed tomography pulmonary angiography (CTPA). Prompt diagnosis and rapid treatment can dramatically reduce mortality in patients. However, the diagnosis of PE is often delayed and missed.

METHODS

In this study, we identified a deep learning model Scaled-YOLOv4 that enables end-to-end automated detection of PE to help solve these problems. A total of 307 CTPA data (Tianjin 142 cases, Linyi 133 cases, and FUMPE 32 cases) were included in this study. The Tianjin dataset was divided 10 times in the ratio of training set: validation set: test set = 7:2:1 for model tuning, and both the Linyi and FUMPE datasets were used as independent external test sets to evaluate the generalization of the model.

RESULTS

Scaled-YOLOv4 was able to process one patient in average 3.55 s [95% CI: 3.51-3.59 s]. It also achieved an average precision (AP) of 83.04 [95% CI: 79.36-86.72] for PE detection on the Tianjin test set, and 75.86 [95% CI: 75.48-76.24] and 72.74 [95% CI: 72.10-73.38] on Linyi and FUMPE, respectively.

CONCLUSIONS

This deep learning algorithm helps detect PE in real time, providing radiologists with aided diagnostic evidence without increasing their workload, and can effectively reduce the probability of delayed patient diagnosis.

Artificial Intelligence Techniques to Predict the Airway Disorders Illness: A Systematic Review

Airway disease is a major healthcare issue that causes at least 3 million fatalities every year. It is also considered one of the foremost causes of death all around the globe by 2030. Numerous studies have been undertaken to demonstrate the latest advances in artificial intelligence algorithms to assist in identifying and classifying these diseases. This comprehensive review aims to summarise the state-of-the-art machine and deep learning-based systems for detecting airway disorders, envisage the trends of the recent work in this domain, and analyze the difficulties and potential future paths. This systematic literature review includes the study of one hundred fifty-five articles on airway diseases such as cystic fibrosis, emphysema, lung cancer, Mesothelioma, covid-19, pneumoconiosis, asthma, pulmonary edema, tuberculosis, pulmonary embolism as well as highlights the automated learning techniques to predict them. The study concludes with a discussion and challenges about expanding the efficiency and machine and deep learning-assisted airway disease detection applications.

Quantitative analysis of high-resolution computed tomography features of idiopathic pulmonary fibrosis: a structure-function correlation study

BACKGROUND

The quantitative analysis of high-resolution computed tomography (HRCT) is increasingly being used to quantify the severity and evaluate the prognosis of disease. Our aim was to quantify the HRCT features of idiopathic pulmonary fibrosis (IPF) and identify their association with pulmonary function tests.

METHODS

This was a retrospective, single-center, clinical research study. Patients with IPF were retrospectively included. Pulmonary segmentation was performed using the deep learning-based method. Radiologists manually segmented 4 findings of IPF, including honeycombing (HC), reticular pattern (RE), traction bronchiectasis (TRBR), and ground glass opacity (GGO). Pulmonary vessels were segmented with the automatic integration segmentation method. All segmentation results were quantified by the corresponding segmentation software. Correlations between the volume of the 4 findings on HRCT, volume of the lesions at different sites, pulmonary vascular-related parameters, and pulmonary function tests were analyzed.

RESULTS

A total of 101 IPF patients (93 males) with a median age of 63 years [interquartile range (IQR), 58 to 68 years] were included in this study. Total lesion extent demonstrated a stronger negative correlation with diffusion capacity for carbon monoxide (DLco) compared to HC, RE, and TRBR [total lesion ratio, correlation coefficient (r) =-0.67, P<0.001; HC, r=-0.45, P<0.001; RE, r=-0.41, P<0.001; TRBR, r=-0.25, P<0.05, respectively]. Correlations with lung function were similar among various lesion sites with r from -0.38 to -0.61 (P<0.001). Pulmonary artery volume (PAV) displayed a slightly increased positive association with the DLco compared to total pulmonary vascular volume (PVV); for PAV, r=0.41 and P<0.001 and for total PVV, r=0.36 and P<0.001. Additionally, total lesion extent, HC, and RE indicated a negative relationship with vascular-related parameters, and the strength of the correlations was independent of lesion site.

CONCLUSIONS

Quantitative analysis of HRCT features of IPF indicated a decline in function and an aggravation of vascular destruction with increasing lesion extent. Furthermore, a positive correlation between vascular-related parameters and pulmonary function was confirmed. This co-linearity indicated the potential of vascular-related parameters as new objective markers for evaluating the severity of IPF.

Quantitative volumetric computed tomography embolic analysis, the Qanadli score, biomarkers, and clinical prognosis in patients with acute pulmonary embolism

Detailed descriptions of acute pulmonary emboli (PE) morphology, total embolic volume (TEV), and their effects upon patients’ clinical presentation and prognosis remain largely unexplored. We studied 201 subjects with acute PE to the emergency department of a single medical center from April 2009 to December 2014. Patient hemodynamics, Troponin I and D-dimer levels, echocardiography, and the 30-day, 90-day and long-term mortality were obtained. Contrast-enhanced computed tomography (CT) of pulmonary structures and 3-dimensional measures of embolic burden were performed. The results showed a linear association between the greater TEV and each of the following 4 variables (increasing incidence of right ventricular (RV) dysfunction, higher systolic pulmonary artery pressure (sPAP), greater RV diameter, and RV/left ventricular (LV) ratio (all p < 0.001)). Among the measures of CT and echocardiography, TEV and RV/LV ratio were significantly associated with impending shock. In backward stepwise logistic regression, TEV, age and respiratory rate remained independent associated with impending shock (OR: 1.58, 1.03, 1.18, respectively and all p < 0.005).Total embolic burden assessed by CT-based quantification serves as a useful index for stressed cardiopulmonary circulation condition and can provide insights into RV dysfunction and the prediction of impending shock.

Developing a Nomogram-Based Scoring Tool to Estimate the Risk of Pulmonary Embolism

Background

Pulmonary embolisms (PEs) are clinically challenging because of their high morbidity and mortality. This study aimed to develop a scoring tool for predicting PEs to improve their clinical management.

Methods

Clinical, laboratory, and imaging parameters were retrospectively collected from suspected PE patients who had cough or chest pain and were hospitalized in West China Hospital of Sichuan University from May 2015 to April 2020. The final diagnosis of PE was defined based on findings from computed tomographic pulmonary angiography (CTPA). In this study, patients were randomly divided 2:1 into derivation and validation cohorts, which were used to create and validate, respectively, a nomogram. Model performance was estimated with the area under the receiver operating characteristic curve and a calibration curve.

Results

Our study incorporated data on more than 100 features from 1480 patients (811 non-PE, 669 PE). The nomogram was constructed using important predictive features including D-dimer, APTT, FDP, platelet count, sodium, albumin and cholesterol and achieved AUC values of 0.692 with the derivation cohort (95% CI 0.688–0.696, P < 0.01) and 0.688 with the validation cohort (95% CI 0.653–0.723, P < 0.01). The calibration curve showed good agreement between the probability predicted by the nomogram and the actual probability.

Conclusion

In this study, we successfully developed a nomogram that can predict the risk of PE, which can not only improve the clinical management of PE patients but also decrease unnecessary CTPA scans and their adverse effects.

Handling of derived imbalanced dataset using XGBoost for identification of pulmonary embolism-a non-cardiac cause of cardiac arrest

Relationship between pulmonary embolism and heart failure is presented in this paper. The proposed research is divided into two phases. The first phase includes the establishment of a novel database with the help of a Cleveland's database for cardiology in order to establish a link between pulmonary embolism and heart failure. The connectivity is based on the relationship between the stroke volume and the pulse pressure (Pp < 25% (ap_hi)). The second phase includes the applicability of machine learning on the novel database. Novel database formed in this work is imbalanced, resulting in the overfitting problem. XGBoost has been used to get rid of overfitting problem. Efficiency has been increased by formulating an ensemble technique by combining extreme learning machines, IB3 tree, logistic regression, and averaged neural network (avNNet) models.

Artificial Intelligence in Diagnostic Radiology: Where Do We Stand, Challenges, and Opportunities

ABSTRACT

Artificial intelligence (AI) is the most revolutionizing development in the health care industry in the current decade, with diagnostic imaging having the greatest share in such development. Machine learning and deep learning (DL) are subclasses of AI that show breakthrough performance in image analysis. They have become the state of the art in the field of image classification and recognition. Machine learning deals with the extraction of the important characteristic features from images, whereas DL uses neural networks to solve such problems with better performance. In this review, we discuss the current applications of machine learning and DL in the field of diagnostic radiology.Deep learning applications can be divided into medical imaging analysis and applications beyond analysis. In the field of medical imaging analysis, deep convolutional neural networks are used for image classification, lesion detection, and segmentation. Also used are recurrent neural networks when extracting information from electronic medical records and to augment the use of convolutional neural networks in the field of image classification. Generative adversarial networks have been explicitly used in generating high-resolution computed tomography and magnetic resonance images and to map computed tomography images from the corresponding magnetic resonance imaging. Beyond image analysis, DL can be used for quality control, workflow organization, and reporting.In this article, we review the most current AI models used in medical imaging research, providing a brief explanation of the various models described in the literature within the past 5 years. Emphasis is placed on the various DL models, as they are the most state-of-art in imaging analysis.

Predicting pulmonary embolism among hospitalized patients with machine learning algorithms

Background

Objective

Materials and Methods

Results

Conclusions

Clot burden of acute pulmonary thromboembolism: comparison of two deep learning algorithms, Qanadli score, and Mastora score

BACKGROUND

The deep learning convolution neural network (DL-CNN) benefits evaluating clot burden of acute pulmonary thromboembolism (APE). Our objective was to compare the performance of the deep learning convolution neural network trained by the fine-tuning [DL-CNN (ft)] and the deep learning convolution neural network trained from the scratch [DL-CNN (fs)] in the quantitative assessment of APE.

METHODS

We included the data of 680 cases for training DL-CNN by DL-CNN (ft) and DL-CNN (fs), then retrospectively included 410 patients (137 patients with APE, 203 males, mean age 60.3±11.4 years) for testing the models. The distribution and volume of clots were respectively assessed by DL-CNN(ft) and DL-CNN(fs), and sensitivity, specificity, and area under the curve (AUC) were used to evaluate their performances in detecting clots on a per-patient and clot level. Radiologists evaluated the distribution of clots, Qanadli score, and Mastora score and right ventricular metrics, and the correlation of clot volumes with right ventricular metrics were analyzed with Spearman correlation analysis.

RESULTS

On a per-patient level, the two DL-CNN models had high sensitivities and moderate specificities [DL-CNN (ft): 100% and 77.29%; DL-CNN (fs): 100% and 75.82%], and their AUCs were comparable (Z=0.30, P=0.38). On a clot level, DL-CNN (ft) and DL-CNN (fs) sensitivities and specificities in detecting central clots were 99.06% and 72.61%, and 100% and 70.63%, respectively. DL-CNN (ft) sensitivities and specificities in detecting peripheral clots were mostly higher than those of DL-CNN (fs), and their AUCs were comparable. Clot volumes measured with the two models were similar (U=85094.500, P=0.741), and significantly correlated with Qanadli scores [DL-CNN(ft) r=0.825, P<0.001, DL-CNN(fs) r=0.827, P<0.001] and Mastora scores [DL-CNN(ft) r=0.859, P<0.001, DL-CNN(fs) r=0.864, P<0.001]. Clot volumes were also correlated with right ventricular metrics. Clot burdens were increased in the low-risk, moderate-risk, and high-risk patients. Binary logistic regression revealed that only the ratio of right ventricular area/left ventricular area (RVa/LVa) was an independent predictor of in-hospital death (odds ratio 6.73; 95% CI, 2.7-18.12, P<0.001).

CONCLUSIONS

Both DL-CNN (ft) and DL-CNN (fs) have high sensitivities and moderate specificities in detecting clots associated with APE, and their performances are comparable. While clot burdens quantitatively calculated by the two DL-CNN models are correlated with right ventricular function and risk stratification, RVa/LVa is an independent prognostic factor of in-hospital death in patients with APE.

Time efficiency and reliability of established computed tomographic obstruction scores in patients with acute pulmonary embolism

Objective

Acute pulmonary embolism (PE) is a life-threatening disease with a high mortality. Computed tomographic pulmonary angiography (CTPA) is used in clinical routine for diagnosis of PE. Many pulmonary obstruction scores were proposed to aid in stratifying clinical course of PE. The purpose of the present study was to compare common pulmonary obstruction scores in PE in regard of time efficiency and interreader agreement based upon a representative patient sample.

Methods

Overall, 50 patients with acute PE were included in this single center, retrospective analysis. Two readers scored the CT images blinded to each other and assessed the scores proposed by Mastora et al., Qanadli et al., Ghanima et al. and Kirchner et al. The required time was assessed of each reading for scoring.

Results

For reader 1, Mastora score took the longest time duration, followed by Kirchner score, Qanadli score and finally Ghanima score (every test, p<0.0001). The interreader variability was excellent for all scores with no significant differences between them. In the Spearman’s correlation analysis strong correlations were identified between the scores of Mastora, Qanadli and Kirchner, whereas Ghanima score was only moderately correlated with the other scores. There was a weak correlation between time duration and Mastora score (r = 0.35, p = 0.014). For the Ghanima score, a significant inverse correlation was found (r = -0.67, p<0.0001).

Conclusion

For the investigated obstruction scores, there are significant differences in regard of time consumption with no relevant differences in regard of interreader variability in patients with acute pulmonary embolism. Mastora score requires the most time effort, whereas the score by Ghanima the least time.

Prediction of chronic thromboembolic pulmonary hypertension with standardised evaluation of initial computed tomography pulmonary angiography performed for suspected acute pulmonary embolism

Objectives

Closer reading of computed tomography pulmonary angiography (CTPA) scans of patients presenting with acute pulmonary embolism (PE) may identify those at high risk of developing chronic thromboembolic pulmonary hypertension (CTEPH). We aimed to validate the predictive value of six radiological predictors that were previously proposed.

Methods

Three hundred forty-one patients with acute PE were prospectively followed for development of CTEPH in six European hospitals. Index CTPAs were analysed post hoc by expert chest radiologists blinded to the final diagnosis. The accuracy of the predictors using a predefined threshold for ‘high risk’ (≥ 3 predictors) and the expert overall judgment on the presence of CTEPH were assessed.

Results

CTEPH was confirmed in nine patients (2.6%) during 2-year follow-up. Any sign of chronic thrombi was already present in 74/341 patients (22%) on the index CTPA, which was associated with CTEPH (OR 7.8, 95%CI 1.9–32); 37 patients (11%) had ≥ 3 of 6 radiological predictors, of whom 4 (11%) were diagnosed with CTEPH (sensitivity 44%, 95%CI 14–79; specificity 90%, 95%CI 86–93). Expert judgment raised suspicion of CTEPH in 27 patients, which was confirmed in 8 (30%; sensitivity 89%, 95%CI 52–100; specificity 94%, 95%CI 91–97).

Conclusions

The presence of ≥ 3 of 6 predefined radiological predictors was highly specific for a future CTEPH diagnosis, comparable to overall expert judgment, while the latter was associated with higher sensitivity. Dedicated CTPA reading for signs of CTEPH may therefore help in early detection of CTEPH after PE, although in our cohort this strategy would not have detected all cases.

Key Points

• Three expert chest radiologists re-assessed CTPA scans performed at the moment of acute pulmonary embolism diagnosis and observed a high prevalence of chronic thrombi and signs of pulmonary hypertension.

• On these index scans, the presence of ≥ 3 of 6 predefined radiological predictors was highly specific for a future diagnosis of chronic thromboembolic pulmonary hypertension (CTEPH), comparable to overall expert judgment.

• Dedicated CTPA reading for signs of CTEPH may help in early detection of CTEPH after acute pulmonary embolism.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00330-021-08364-0.

Role of multidetector CT in predicting patient outcome in cases of pulmonary embolism: correlation between imaging findings, ICU admissions and mortality rate

Background

Pulmonary embolism (PE) is a critical medical condition that requires prompt diagnosis and treatment to avoid serious morbidity and mortality risk. Multidetector CT pulmonary angiography (CTPA) is considered the first-line imaging modality for suspected acute PE. The presence of right heart strain, which supports the diagnosis, requires special attention. The aim of our retrospective study is to assess the reliability of CTPA hemodynamic indices in predicting patients’ outcome in cases of PE.

Results

Sixty patients were included in our study. CTPA parameters including main pulmonary artery (MPA) diameter, left ventricle (LV) diameter, right ventricle (RV)/LV ratio, and septal deviation had a clinical prognostic value for short-term 30-day mortality and ICU admission. Statistically significant relationship between MPA diameter > 29 mm, LV diameter, RV/LV ratio > 1, left-sided septal deviation and contrast reflux into the IVC/distal hepatic veins with ICU admission was observed with p values 0.031, 0.000, 0.000, 0.005 and 0.028 respectively. There was a statistically significant correlation between MPA diameter > 29 mm, LV diameter, RV/LV > 1 ratio and septal deviation with 30-day mortality with p values of < 0.001, 0.001, < 0.001 and 0.015 respectively. No significant correlation was found between 30-day mortality and contrast reflux to IVC with p value of 0.070.

Conclusions

CTPA measurements including MPA diameter, RV/LV ratio and septal deviation were found to be significantly correlated to ICU admission and 30-day mortality as predictors for PE severity. CT contrast reflux was found to be correlated to ICU admission; however, it was not significantly correlated to 30-day mortality.

Accuracy of deep learning-based computed tomography diagnostic system for COVID-19: A consecutive sampling external validation cohort study

Ali-M3, an artificial intelligence program, analyzes chest computed tomography (CT) and detects the likelihood of coronavirus disease (COVID-19) based on scores ranging from 0 to 1. However, Ali-M3 has not been externally validated. Our aim was to evaluate the accuracy of Ali-M3 for detecting COVID-19 and discuss its clinical value. We evaluated the external validity of Ali-M3 using sequential Japanese sampling data. In this retrospective cohort study, COVID-19 infection probabilities for 617 symptomatic patients were determined using Ali-M3. In 11 Japanese tertiary care facilities, these patients underwent reverse transcription-polymerase chain reaction (RT-PCR) testing. They also underwent chest CT to confirm a diagnosis of COVID-19. Of the 617 patients, 289 (46.8%) were RT-PCR-positive. The area under the curve (AUC) of Ali-M3 for predicting a COVID-19 diagnosis was 0.797 (95% confidence interval: 0.762‒0.833) and the goodness-of-fit was P = 0.156. With a cut-off probability of a diagnosis of COVID-19 by Ali-M3 set at 0.5, the sensitivity and specificity were 80.6% and 68.3%, respectively. A cut-off of 0.2 yielded a sensitivity and specificity of 89.2% and 43.2%, respectively. Among the 223 patients who required oxygen, the AUC was 0.825. Sensitivity at a cut-off of 0.5% and 0.2% was 88.7% and 97.9%, respectively. Although the sensitivity was lower when the days from symptom onset were fewer, the sensitivity increased for both cut-off values after 5 days. We evaluated Ali-M3 using external validation with symptomatic patient data from Japanese tertiary care facilities. As Ali-M3 showed sufficient sensitivity performance, despite a lower specificity performance, Ali-M3 could be useful in excluding a diagnosis of COVID-19.

Identification of chronic thromboembolic pulmonary hypertension on CTPAs performed for diagnosing acute pulmonary embolism depending on level of expertise

BACKGROUND

Expert reading often reveals radiological signs of chronic thromboembolic pulmonary hypertension (CTEPH) or chronic PE on computed tomography pulmonary angiography (CTPA) performed at the time of acute pulmonary embolism (PE) presentation preceding CTEPH. Little is known about the accuracy and reproducibility of CTPA reading by radiologists in training in this setting.

OBJECTIVES

To evaluate 1) whether signs of CTEPH or chronic PE are routinely reported on CTPA for suspected PE; and 2) whether CTEPH-non-expert readers achieve comparable predictive accuracy to CTEPH-expert radiologists after dedicated instruction.

METHODS

Original reports of CTPAs demonstrating acute PE in 50 patients whom ultimately developed CTEPH, and those of 50 PE who did not, were screened for documented signs of CTEPH. All scans were re-assessed by three CTEPH-expert readers and two CTEPH-non-expert readers (blinded and independently) for predefined signs and overall presence of CTEPH.

RESULTS

Signs of chronic PE were mentioned in the original reports of 14/50 cases (28%), while CTEPH-expert radiologists had recognized 44/50 (88%). Using a standardized definition (≥3 predefined radiological signs), moderate-to-good agreement was reached between CTEPH-non-expert readers and the experts' consensus (k-statistics 0.46; 0.61) at slightly lower sensitivities. The CTEPH-non-expert readers had moderate agreement on the presence of CTEPH (κ-statistic 0.38), but both correctly identified most cases (80% and 88%, respectively).

CONCLUSIONS

Concomitant signs of CTEPH were poorly documented in daily practice, while most CTEPH patients were identified by CTEPH-non-expert readers after dedicated instruction. These findings underline the feasibility of achieving earlier CTEPH diagnosis by assessing CTPAs more attentively.

Machine learning with D-dimer in the risk stratification for pulmonary embolism: a derivation and internal validation study

AIM

To develop a machine learning model to predict the diagnosis of pulmonary embolism (PE).

METHODS AND RESULTS

We undertook a derivation and internal validation study to develop a risk prediction model for use in patients being investigated for possible PE. The machine learning technique, generalized logistic regression using elastic net, was chosen following an assessment of seven machine learning techniques and on the basis that it optimized the area under the receiver operator characteristic curve (AUC) and Brier score. Models were developed both with and without the addition of D-dimer. A total of 3347 patients were included in the study of whom, 219 (6.5%) had PE. Four clinical variables (O2 saturation, previous deep venous thrombosis or PE, immobilization or surgery, and alternative diagnosis equal or more likely than PE) plus D-dimer contributed to the machine learning models. The addition of D-dimer improved the AUC by 0.16 (95% confidence interval 0.13-0.19), from 0.73 to 0.89 (0.87-0.91) and decreased the Brier score by 14% (10-18%). More could be ruled out with a higher positive likelihood ratio than by the Wells score combined with D-dimer, revised Geneva score combined with D-dimer, or the Pulmonary Embolism Rule-out Criteria score. Machine learning with D-dimer maintained a low-false-negative rate at a true-negative rate of nearly 53%, which was better performance than any of the other alternatives.

CONCLUSION

A machine learning model outperformed traditional risk scores for the risk stratification of PE in the emergency department. However, external validation is needed.

Performance of a 3D convolutional neural network in the detection of hypoperfusion at CT pulmonary angiography in patients with chronic pulmonary embolism: a feasibility study

BACKGROUND

Chronic pulmonary embolism (CPE) is a life-threatening disease easily misdiagnosed on computed tomography. We investigated a three-dimensional convolutional neural network (CNN) algorithm for detecting hypoperfusion in CPE from computed tomography pulmonary angiography (CTPA).

METHODS

Preoperative CTPA of 25 patients with CPE and 25 without pulmonary embolism were selected. We applied a 48%-12%-40% training-validation-testing split (12 positive and 12 negative CTPA volumes for training, 3 positives and 3 negatives for validation, 10 positives and 10 negatives for testing). The median number of axial images per CTPA was 335 (min-max, 111-570). Expert manual segmentations were used as training and testing targets. The CNN output was compared to a method in which a Hounsfield unit (HU) threshold was used to detect hypoperfusion. Receiver operating characteristic area under the curve (AUC) and Matthew correlation coefficient (MCC) were calculated with their 95% confidence interval (CI).

RESULTS

The predicted segmentations of CNN showed AUC 0.87 (95% CI 0.82-0.91), those of HU-threshold method 0.79 (95% CI 0.74-0.84). The optimal global threshold values were CNN output probability ≥ 0.37 and ≤ -850 HU. Using these values, MCC was 0.46 (95% CI 0.29-0.59) for CNN and 0.35 (95% CI 0.18-0.48) for HU-threshold method (average difference in MCC in the bootstrap samples 0.11 (95% CI 0.05-0.16). A high CNN prediction probability was a strong predictor of CPE.

CONCLUSIONS

We proposed a deep learning method for detecting hypoperfusion in CPE from CTPA. This model may help evaluating disease extent and supporting treatment planning.

Dual-Energy CT for Pulmonary Embolism: Current and Evolving Clinical Applications

Pulmonary embolism (PE) is a potentially fatal disease if the diagnosis or treatment is delayed. Currently, multidetector computed tomography (MDCT) is considered the standard imaging method for diagnosing PE. Dual-energy CT (DECT) has the advantages of MDCT and can provide functional information for patients with PE. The aim of this review is to present the potential clinical applications of DECT in PE, focusing on the diagnosis and risk stratification of PE.