Pub Date : 2024-01-19DOI: 10.1016/j.ejro.2024.100549
Bo Li , Jie Su , Kai Liu, Chunfeng Hu
Purpose
Programmed cell death protein-1 ligand (PD-L1) is an important prognostic predictor for immunotherapy of non-small cell lung cancer (NSCLC). This study aimed to develop a non-invasive deep learning and radiomics model based on positron emission tomography and computed tomography (PET/CT) to predict PD-L1 expression in NSCLC.
Methods
A total of 136 patients with NSCLC between January 2021 and September 2022 were enrolled in this study. The patients were randomly divided into the training dataset and the validation dataset in a ratio of 7:3. Radiomics feature and deep learning feature were extracted from their PET/CT images. The Mann-whitney U-test, Least Absolute Shrinkage and Selection Operator algorithm and Spearman correlation analysis were used to select the top significant features. Then we developed a radiomics model, a deep learning model, and a fusion model based on the selected features. The performance of three models were compared by the area under the curve (AUC), sensitivity, specificity, accuracy, positive predictive value, and negative predictive value.
Results
Of the patients, 42 patients were PD-L1 negative and 94 patients were PD-L1 positive. A total of 2446 radiomics features and 4096 deep learning features were extracted per patient. In the training dataset, the fusion model achieved a highest AUC (0.954, 95% confident internal [CI]: 0.890–0.986) compared with the radiomics model (0.829, 95%CI: 0.738–0.898) and the deep learning model (0.935, 95%CI: 0.865–0.975). In the validation dataset, the AUC of the fusion model (0.910, 95% CI: 0.779–0.977) was also higher than that of the radiomics model (0.785, 95% CI: 0.628–0.897) and the deep learning model (0.867, 95% CI: 0.724–0.952).
Conclusion
The PET/CT-based deep learning radiomics model can predict the PD-L1 expression accurately in NSCLC patients, and provides a non-invasive tool for clinicians to select positive PD-L1 patients.
{"title":"Deep learning radiomics model based on PET/CT predicts PD-L1 expression in non-small cell lung cancer","authors":"Bo Li , Jie Su , Kai Liu, Chunfeng Hu","doi":"10.1016/j.ejro.2024.100549","DOIUrl":"https://doi.org/10.1016/j.ejro.2024.100549","url":null,"abstract":"<div><h3>Purpose</h3><p>Programmed cell death protein-1 ligand (PD-L1) is an important prognostic predictor for immunotherapy of non-small cell lung cancer (NSCLC). This study aimed to develop a non-invasive deep learning and radiomics model based on positron emission tomography and computed tomography (PET/CT) to predict PD-L1 expression in NSCLC.</p></div><div><h3>Methods</h3><p>A total of 136 patients with NSCLC between January 2021 and September 2022 were enrolled in this study. The patients were randomly divided into the training dataset and the validation dataset in a ratio of 7:3. Radiomics feature and deep learning feature were extracted from their PET/CT images. The Mann-whitney U-test, Least Absolute Shrinkage and Selection Operator algorithm and Spearman correlation analysis were used to select the top significant features. Then we developed a radiomics model, a deep learning model, and a fusion model based on the selected features. The performance of three models were compared by the area under the curve (AUC), sensitivity, specificity, accuracy, positive predictive value, and negative predictive value.</p></div><div><h3>Results</h3><p>Of the patients, 42 patients were PD-L1 negative and 94 patients were PD-L1 positive. A total of 2446 radiomics features and 4096 deep learning features were extracted per patient. In the training dataset, the fusion model achieved a highest AUC (0.954, 95% confident internal [CI]: 0.890–0.986) compared with the radiomics model (0.829, 95%CI: 0.738–0.898) and the deep learning model (0.935, 95%CI: 0.865–0.975). In the validation dataset, the AUC of the fusion model (0.910, 95% CI: 0.779–0.977) was also higher than that of the radiomics model (0.785, 95% CI: 0.628–0.897) and the deep learning model (0.867, 95% CI: 0.724–0.952).</p></div><div><h3>Conclusion</h3><p>The PET/CT-based deep learning radiomics model can predict the PD-L1 expression accurately in NSCLC patients, and provides a non-invasive tool for clinicians to select positive PD-L1 patients.</p></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":null,"pages":null},"PeriodicalIF":2.0,"publicationDate":"2024-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352047724000042/pdfft?md5=665b39078d34d10cc3d399f816e580ab&pid=1-s2.0-S2352047724000042-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139494210","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-01-16DOI: 10.1016/j.ejro.2024.100548
Xinna Lv , Ye Li , Bing Wang , Yichuan Wang , Zexuan Xu , Dailun Hou
Background
Kirsten rat sarcoma virus (KRAS) has evolved from a genotype with predictive value to a therapeutic target recently. The study aimed to establish non-invasive radiomics models based on MRI to discriminate KRAS from epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK) mutations in lung cancer patients with brain metastases (BM), then further explore the optimal sequence for prediction.
Methods
This retrospective study involved 317 patients (218 patients in training cohort and 99 patients in testing cohort) who had confirmed of KRAS, EGFR or ALK mutations. Radiomics features were separately extracted from T2WI, T2 fluid-attenuated inversion recovery (T2-FLAIR), diffusion weighted imaging (DWI) and contrast-enhanced T1-weighted imaging (T1-CE) sequences. The maximal information coefficient and recursive feature elimination method were used to select informative features. Then we built four radiomics models for differentiating KRAS from EGFR or ALK using random forest classifier. ROC curves were used to validate the capability of the models.
Results
The four radiomics models for discriminating KRAS from EGFR all worked well, especially DWI and T2WI models (AUCs: 0.942, 0.942 in training cohort, 0.949, 0.954 in testing cohort). When KRAS compared to ALK, DWI and T2-FLAIR models showed excellent performance in two cohorts (AUCs: 0.947, 0.917 in training cohort, 0.850, 0.824 in testing cohort).
Conclusions
Radiomics classifiers integrating MRI have potential to discriminate KRAS from EGFR or ALK, which are helpful to guide treatment and facilitate the discovery of new approaches capable of achieving this long-sought goal of cure in lung cancer patients with KRAS.
{"title":"Multisequence MRI-based radiomics signature as potential biomarkers for differentiating KRAS mutations in non-small cell lung cancer with brain metastases","authors":"Xinna Lv , Ye Li , Bing Wang , Yichuan Wang , Zexuan Xu , Dailun Hou","doi":"10.1016/j.ejro.2024.100548","DOIUrl":"https://doi.org/10.1016/j.ejro.2024.100548","url":null,"abstract":"<div><h3>Background</h3><p>Kirsten rat sarcoma virus (KRAS) has evolved from a genotype with predictive value to a therapeutic target recently. The study aimed to establish non-invasive radiomics models based on MRI to discriminate KRAS from epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK) mutations in lung cancer patients with brain metastases (BM), then further explore the optimal sequence for prediction.</p></div><div><h3>Methods</h3><p>This retrospective study involved 317 patients (218 patients in training cohort and 99 patients in testing cohort) who had confirmed of KRAS, EGFR or ALK mutations. Radiomics features were separately extracted from T2WI, T2 fluid-attenuated inversion recovery (T2-FLAIR), diffusion weighted imaging (DWI) and contrast-enhanced T1-weighted imaging (T1-CE) sequences. The maximal information coefficient and recursive feature elimination method were used to select informative features. Then we built four radiomics models for differentiating KRAS from EGFR or ALK using random forest classifier. ROC curves were used to validate the capability of the models.</p></div><div><h3>Results</h3><p>The four radiomics models for discriminating KRAS from EGFR all worked well, especially DWI and T2WI models (AUCs: 0.942, 0.942 in training cohort, 0.949, 0.954 in testing cohort). When KRAS compared to ALK, DWI and T2-FLAIR models showed excellent performance in two cohorts (AUCs: 0.947, 0.917 in training cohort, 0.850, 0.824 in testing cohort).</p></div><div><h3>Conclusions</h3><p>Radiomics classifiers integrating MRI have potential to discriminate KRAS from EGFR or ALK, which are helpful to guide treatment and facilitate the discovery of new approaches capable of achieving this long-sought goal of cure in lung cancer patients with KRAS.</p></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":null,"pages":null},"PeriodicalIF":2.0,"publicationDate":"2024-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352047724000030/pdfft?md5=a70c992437150ce872f0aba1a3adae00&pid=1-s2.0-S2352047724000030-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139480051","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-01-13DOI: 10.1016/j.ejro.2023.100545
Si Eun Lee , Hanpyo Hong , Eun-Kyung Kim
Purpose
To evaluate artificial intelligence-based computer-aided diagnosis (AI-CAD) for screening mammography, we analyzed the diagnostic performance of radiologists by providing and withholding AI-CAD results alternatively every month.
Methods
This retrospective study was approved by the institutional review board with a waiver for informed consent. Between August 2020 and May 2022, 1819 consecutive women (mean age 50.8 ± 9.4 years) with 2061 screening mammography and ultrasound performed on the same day in a single institution were included. Radiologists interpreted screening mammography in clinical practice with AI-CAD results being provided or withheld alternatively by month. The AI-CAD results were retrospectively obtained for analysis even when withheld from radiologists. The diagnostic performances of radiologists and stand-alone AI-CAD were compared and the performances of radiologists with and without AI-CAD assistance were also compared by cancer detection rate, recall rate, sensitivity, specificity, accuracy and area under the receiver-operating-characteristics curve (AUC).
Results
Twenty-nine breast cancer patients and 1790 women without cancers were included. Diagnostic performances of the radiologists did not significantly differ with and without AI-CAD assistance. Radiologists with AI-CAD assistance showed the same sensitivity (76.5%) and similar specificity (92.3% vs 93.8%), AUC (0.844 vs 0.851), and recall rates (8.8% vs. 7.4%) compared to standalone AI-CAD. Radiologists without AI-CAD assistance showed lower specificity (91.9% vs 94.6%) and accuracy (91.5% vs 94.1%) and higher recall rates (8.6% vs 5.9%, all p < 0.05) compared to stand-alone AI-CAD.
Conclusion
Radiologists showed no significant difference in diagnostic performance when both screening mammography and ultrasound were performed with or without AI-CAD assistance for mammography. However, without AI-CAD assistance, radiologists showed lower specificity and accuracy and higher recall rates compared to stand-alone AI-CAD.
{"title":"Diagnostic performance with and without artificial intelligence assistance in real-world screening mammography","authors":"Si Eun Lee , Hanpyo Hong , Eun-Kyung Kim","doi":"10.1016/j.ejro.2023.100545","DOIUrl":"https://doi.org/10.1016/j.ejro.2023.100545","url":null,"abstract":"<div><h3>Purpose</h3><p>To evaluate artificial intelligence-based computer-aided diagnosis (AI-CAD) for screening mammography, we analyzed the diagnostic performance of radiologists by providing and withholding AI-CAD results alternatively every month.</p></div><div><h3>Methods</h3><p>This retrospective study was approved by the institutional review board with a waiver for informed consent. Between August 2020 and May 2022, 1819 consecutive women (mean age 50.8 ± 9.4 years) with 2061 screening mammography and ultrasound performed on the same day in a single institution were included. Radiologists interpreted screening mammography in clinical practice with AI-CAD results being provided or withheld alternatively by month. The AI-CAD results were retrospectively obtained for analysis even when withheld from radiologists. The diagnostic performances of radiologists and stand-alone AI-CAD were compared and the performances of radiologists with and without AI-CAD assistance were also compared by cancer detection rate, recall rate, sensitivity, specificity, accuracy and area under the receiver-operating-characteristics curve (AUC).</p></div><div><h3>Results</h3><p>Twenty-nine breast cancer patients and 1790 women without cancers were included. Diagnostic performances of the radiologists did not significantly differ with and without AI-CAD assistance. Radiologists with AI-CAD assistance showed the same sensitivity (76.5%) and similar specificity (92.3% vs 93.8%), AUC (0.844 vs 0.851), and recall rates (8.8% vs. 7.4%) compared to standalone AI-CAD. Radiologists without AI-CAD assistance showed lower specificity (91.9% vs 94.6%) and accuracy (91.5% vs 94.1%) and higher recall rates (8.6% vs 5.9%, all p < 0.05) compared to stand-alone AI-CAD.</p></div><div><h3>Conclusion</h3><p>Radiologists showed no significant difference in diagnostic performance when both screening mammography and ultrasound were performed with or without AI-CAD assistance for mammography. However, without AI-CAD assistance, radiologists showed lower specificity and accuracy and higher recall rates compared to stand-alone AI-CAD.</p></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":null,"pages":null},"PeriodicalIF":2.0,"publicationDate":"2024-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352047723000710/pdfft?md5=c846cac93a8f564a2b410650560d00bd&pid=1-s2.0-S2352047723000710-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139436628","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-01-13DOI: 10.1016/j.ejro.2024.100546
Paulo Savoia , Marcio Valente Yamada Sawamura , Renata Aparecida de Almeida Monteiro , Amaro Nunes Duarte-Neto , Maria da Graça Morais Martin , Marisa Dolhnikoff , Thais Mauad , Paulo Hilário Nascimento Saldiva , Claudia da Costa Leite , Luiz Fernando Ferraz da Silva , Ellison Fernando Cardoso
Objectives
Performing autopsies in a pandemic scenario is challenging, as the need to understand pathophysiology must be balanced with the contamination risk. A minimally invasive autopsy might be a solution. We present a model that combines radiology and pathology to evaluate postmortem CT lung findings and their correlation with histopathology.
Methods
Twenty-nine patients with fatal COVID-19 underwent postmortem chest CT, and multiple lung tissue samples were collected. The chest CT scans were analyzed and quantified according to lung involvement in five categories: normal, ground-glass opacities, crazy-paving, small consolidations, and large or lobar consolidations. The lung tissue samples were examined and quantified in three categories: normal lung, exudative diffuse alveolar damage (DAD), and fibroproliferative DAD. A linear index was used to estimate the global severity of involvement by CT and histopathological analysis.
Results
There was a positive correlation between patient mean CT and histopathological severity score indexes - Pearson correlation coefficient (R) = 0.66 (p = 0.0078). When analyzing the mean lung involvement percentage of each finding, positive correlations were found between the normal lung percentage between postmortem CT and histopathology (R=0.65, p = 0.0082), as well as between ground-glass opacities in postmortem CT and normal lungs in histopathology (R=0.65, p = 0.0086), but negative correlations were observed between ground-glass opacities extension and exudative diffuse alveolar damage in histological slides (R=−0.68, p = 0.005). Additionally, it was found is a trend toward a decrease in the percentage of normal lung tissue on the histological slides as the percentage of consolidations in postmortem CT scans increased (R =−0.51, p = 0.055). The analysis of the other correlations between the percentage of each finding did not show any significant correlation or correlation trends (p ≥ 0.10).
Conclusions
A minimally invasive autopsy is valid. As the severity of involvement is increased in CT, more advanced disease is seen on histopathology. However, we cannot state that one specific radiological category represents a specific pathological correspondent. Ground-glass opacities, in the postmortem stage, must be interpreted with caution, as expiratory lungs may overestimate disease.
{"title":"Postmortem chest computed tomography in COVID-19: A minimally invasive autopsy method","authors":"Paulo Savoia , Marcio Valente Yamada Sawamura , Renata Aparecida de Almeida Monteiro , Amaro Nunes Duarte-Neto , Maria da Graça Morais Martin , Marisa Dolhnikoff , Thais Mauad , Paulo Hilário Nascimento Saldiva , Claudia da Costa Leite , Luiz Fernando Ferraz da Silva , Ellison Fernando Cardoso","doi":"10.1016/j.ejro.2024.100546","DOIUrl":"https://doi.org/10.1016/j.ejro.2024.100546","url":null,"abstract":"<div><h3>Objectives</h3><p>Performing autopsies in a pandemic scenario is challenging, as the need to understand pathophysiology must be balanced with the contamination risk. A minimally invasive autopsy might be a solution. We present a model that combines radiology and pathology to evaluate postmortem CT lung findings and their correlation with histopathology.</p></div><div><h3>Methods</h3><p>Twenty-nine patients with fatal COVID-19 underwent postmortem chest CT, and multiple lung tissue samples were collected. The chest CT scans were analyzed and quantified according to lung involvement in five categories: normal, ground-glass opacities, crazy-paving, small consolidations, and large or lobar consolidations. The lung tissue samples were examined and quantified in three categories: normal lung, exudative diffuse alveolar damage (DAD), and fibroproliferative DAD. A linear index was used to estimate the global severity of involvement by CT and histopathological analysis.</p></div><div><h3>Results</h3><p>There was a positive correlation between patient mean CT and histopathological severity score indexes - Pearson correlation coefficient (R) = 0.66 (p = 0.0078). When analyzing the mean lung involvement percentage of each finding, positive correlations were found between the normal lung percentage between postmortem CT and histopathology (R=0.65, p = 0.0082), as well as between ground-glass opacities in postmortem CT and normal lungs in histopathology (R=0.65, p = 0.0086), but negative correlations were observed between ground-glass opacities extension and exudative diffuse alveolar damage in histological slides (R=−0.68, p = 0.005). Additionally, it was found is a trend toward a decrease in the percentage of normal lung tissue on the histological slides as the percentage of consolidations in postmortem CT scans increased (R =−0.51, p = 0.055). The analysis of the other correlations between the percentage of each finding did not show any significant correlation or correlation trends (p ≥ 0.10).</p></div><div><h3>Conclusions</h3><p>A minimally invasive autopsy is valid. As the severity of involvement is increased in CT, more advanced disease is seen on histopathology. However, we cannot state that one specific radiological category represents a specific pathological correspondent. Ground-glass opacities, in the postmortem stage, must be interpreted with caution, as expiratory lungs may overestimate disease.</p></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":null,"pages":null},"PeriodicalIF":2.0,"publicationDate":"2024-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352047724000017/pdfft?md5=62b0ba1e43f345bc6fabdea0bce149ea&pid=1-s2.0-S2352047724000017-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139436627","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-01-03DOI: 10.1016/j.ejro.2023.100543
Yuzhen Xi , Hao Dong , Mengze Wang , Shiyu Chen , Jing Han , Miao Liu , Feng Jiang , Zhongxiang Ding
Purpose
The objective is to create a comprehensive model that integrates clinical, semantic, and radiomics features to forecast the 5-year progression-free survival (PFS) of individuals diagnosed with non-distant metastatic Nasopharyngeal Carcinoma (NPC).
Methods
In a retrospective analysis, we included clinical and MRI data from 313 patients diagnosed with primary NPC. Patient classification into progressive and non-progressive categories relied on the occurrence of recurrence or distant metastasis within a 5-year timeframe. Initial screening comprised clinical features and statistically significant image semantic features. Subsequently, MRI radiomics features were extracted from all patients, and optimal features were selected to formulate the Rad-Score.Combining Rad-Score, image semantic features, and clinical features to establish a combined model Evaluation of predictive efficacy was conducted using ROC curves and nomogram specific to NPC progression. Lastly, employing the optimal ROC cutoff value from the combined model, patients were dichotomized into high-risk and low-risk groups, facilitating a comparison of 10-year overall survival (OS) between the groups.
Results
The combined model showcased superior predictive performance for NPC progression, reflected by AUC values of 0.84, an accuracy rate of 81.60%, sensitivity at 0.77, and specificity at 0.81 within the training group. In the test set, the AUC value reached 0.81, with an accuracy of 74.6%, sensitivity at 0.82, and specificity at 0.66.
Conclusion
The amalgamation of Rad-Score, clinical, and imaging semantic features from multi-parameter MRI exhibited significant promise in prognosticating 5-year PFS for non-distant metastatic NPC patients. The combined model provided quantifiable data for informed and personalized diagnosis and treatment planning.
{"title":"Early prediction of long-term survival of patients with nasopharyngeal carcinoma by multi-parameter MRI radiomics","authors":"Yuzhen Xi , Hao Dong , Mengze Wang , Shiyu Chen , Jing Han , Miao Liu , Feng Jiang , Zhongxiang Ding","doi":"10.1016/j.ejro.2023.100543","DOIUrl":"https://doi.org/10.1016/j.ejro.2023.100543","url":null,"abstract":"<div><h3>Purpose</h3><p>The objective is to create a comprehensive model that integrates clinical, semantic, and radiomics features to forecast the 5-year progression-free survival (PFS) of individuals diagnosed with non-distant metastatic Nasopharyngeal Carcinoma (NPC).</p></div><div><h3>Methods</h3><p>In a retrospective analysis, we included clinical and MRI data from 313 patients diagnosed with primary NPC. Patient classification into progressive and non-progressive categories relied on the occurrence of recurrence or distant metastasis within a 5-year timeframe. Initial screening comprised clinical features and statistically significant image semantic features. Subsequently, MRI radiomics features were extracted from all patients, and optimal features were selected to formulate the Rad-Score.Combining Rad-Score, image semantic features, and clinical features to establish a combined model Evaluation of predictive efficacy was conducted using ROC curves and nomogram specific to NPC progression. Lastly, employing the optimal ROC cutoff value from the combined model, patients were dichotomized into high-risk and low-risk groups, facilitating a comparison of 10-year overall survival (OS) between the groups.</p></div><div><h3>Results</h3><p>The combined model showcased superior predictive performance for NPC progression, reflected by AUC values of 0.84, an accuracy rate of 81.60%, sensitivity at 0.77, and specificity at 0.81 within the training group. In the test set, the AUC value reached 0.81, with an accuracy of 74.6%, sensitivity at 0.82, and specificity at 0.66.</p></div><div><h3>Conclusion</h3><p>The amalgamation of Rad-Score, clinical, and imaging semantic features from multi-parameter MRI exhibited significant promise in prognosticating 5-year PFS for non-distant metastatic NPC patients. The combined model provided quantifiable data for informed and personalized diagnosis and treatment planning.</p></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":null,"pages":null},"PeriodicalIF":2.0,"publicationDate":"2024-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352047723000692/pdfft?md5=001f7c2fcfb88e25fc45d76bc97b84b5&pid=1-s2.0-S2352047723000692-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139107838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-12-15DOI: 10.1016/j.ejro.2023.100542
Helena Mellander , Amir Hillal , Teresa Ullberg , Johan Wassélius
Objective
To systematically evaluate the ability of the CINA® LVO software to detect large vessel occlusions eligible for mechanical thrombectomy on CTA using conventional neuroradiological assessment as gold standard.
Methods
Retrospectively, two hundred consecutive patients referred for a brain CTA and two hundred patients that had been subject for endovascular thrombectomy, with an accessible preceding CTA, were assessed for large vessel occlusions (LVO) using the CINA® LVO software. The patients were sub-grouped by occlusion site. The original radiology report was used as ground truth and cases with disagreement were reassessed. Two-by-two tables were created and measures for LVO detection were calculated.
Results
A total of four-hundred patients were included; 221 LVOs were present in 215 patients (54 %). The overall specificity was high for LVOs in the anterior circulation (93 %). The overall sensitivity for LVOs in the anterior circulation was 54 % with the highest sensitivity for the M1 segment of the middle cerebral artery (87 %) and T-type internal carotid occlusions (84 %). The sensitivity was low for occlusions in the M2 segment of the middle cerebral artery (13 % and 0 % for proximal and distal M2 occlusions respectively) and in posterior circulation occlusions (0 %, not included in the intended use of the software).
Conclusions
LVO detection sensitivity for the CINA® LVO software differs largely depending on the location of the occlusion, with low sensitivity for detection of some LVOs potentially eligible for mechanical thrombectomy. Further development of the software to increase sensitivity to all LVO locations would increase the clinical usefulness.
{"title":"Evaluation of CINA® LVO artificial intelligence software for detection of large vessel occlusion in brain CT angiography","authors":"Helena Mellander , Amir Hillal , Teresa Ullberg , Johan Wassélius","doi":"10.1016/j.ejro.2023.100542","DOIUrl":"https://doi.org/10.1016/j.ejro.2023.100542","url":null,"abstract":"<div><h3>Objective</h3><p>To systematically evaluate the ability of the CINA® LVO software to detect large vessel occlusions eligible for mechanical thrombectomy on CTA using conventional neuroradiological assessment as gold standard.</p></div><div><h3>Methods</h3><p>Retrospectively, two hundred consecutive patients referred for a brain CTA and two hundred patients that had been subject for endovascular thrombectomy, with an accessible preceding CTA, were assessed for large vessel occlusions (LVO) using the CINA® LVO software. The patients were sub-grouped by occlusion site. The original radiology report was used as ground truth and cases with disagreement were reassessed. Two-by-two tables were created and measures for LVO detection were calculated.</p></div><div><h3>Results</h3><p>A total of four-hundred patients were included; 221 LVOs were present in 215 patients (54 %). The overall specificity was high for LVOs in the anterior circulation (93 %). The overall sensitivity for LVOs in the anterior circulation was 54 % with the highest sensitivity for the M1 segment of the middle cerebral artery (87 %) and T-type internal carotid occlusions (84 %). The sensitivity was low for occlusions in the M2 segment of the middle cerebral artery (13 % and 0 % for proximal and distal M2 occlusions respectively) and in posterior circulation occlusions (0 %, not included in the intended use of the software).</p></div><div><h3>Conclusions</h3><p>LVO detection sensitivity for the CINA® LVO software differs largely depending on the location of the occlusion, with low sensitivity for detection of some LVOs potentially eligible for mechanical thrombectomy. Further development of the software to increase sensitivity to all LVO locations would increase the clinical usefulness.</p></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":null,"pages":null},"PeriodicalIF":2.0,"publicationDate":"2023-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352047723000680/pdfft?md5=c01c43343df5bc80ba1a6999706dc7b0&pid=1-s2.0-S2352047723000680-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138739277","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
To assess the diagnostic performance and calculate the optimal threshold for quantitative biomarkers to differentiate bone metastasis and benign bone marrow lesions using turbo spin echo (TSE) Dixon images with a 3.0 T scanner.
Materials and methods
Each 100 patients diagnosed with bone metastases and variable benign bone marrow lesions on spine MRI were included retrospectively. Images included in-phase (IP), opposed-phase (OP), water images (WI), and fat images (FI) by the TSE Dixon technique with T1WI and T2WI using a 3.0 T scanner. Regions of interest (ROI) of the lesions were manually drawn by two musculoskeletal radiologists independently, and the average signal intensity was recorded. The signal reduction rate from IP to OP (%drop) and a fat fraction (%fat) were calculated.
Results
All biomarkers showed a significant difference between metastatic and benign lesions (P < 0.001). When comparing the AUCs, the %drop of T1WI had the highest AUC (0.934). Although the AUC of %fat from T2WI was significantly lower than that of other biomarkers, the %drop of T2WI was not significantly different from the %drop of T1WI (p = 0.339). The optimal threshold of %drop to differentiate metastatic and benign lesions was 22.0 in T1WI and 15.9 in T2WI. The inter-reader agreement was excellent for all biomarkers (0.82–0.86).
Conclusion
While %drop of T1WI showed the highest diagnostic performance to differentiate bone metastasis from benign lesions, the %drop of T2WI showed a comparable ability using a threshold 15.9.
{"title":"Quantitative biomarkers for distinguishing bone metastasis and benign bone marrow lesions using turbo spin echo T1- and T2-weighted Dixon imaging at 3.0 T","authors":"Sho Ogiwara, Takeshi Fukuda, Takenori Yonenaga, Akira Ogihara, Hiroya Ojiri","doi":"10.1016/j.ejro.2023.100541","DOIUrl":"https://doi.org/10.1016/j.ejro.2023.100541","url":null,"abstract":"<div><h3>Objective</h3><p>To assess the diagnostic performance and calculate the optimal threshold for quantitative biomarkers to differentiate bone metastasis and benign bone marrow lesions using turbo spin echo (TSE) Dixon images with a 3.0 T scanner.</p></div><div><h3>Materials and methods</h3><p>Each 100 patients diagnosed with bone metastases and variable benign bone marrow lesions on spine MRI were included retrospectively. Images included in-phase (IP), opposed-phase (OP), water images (WI), and fat images (FI) by the TSE Dixon technique with T1WI and T2WI using a 3.0 T scanner. Regions of interest (ROI) of the lesions were manually drawn by two musculoskeletal radiologists independently, and the average signal intensity was recorded. The signal reduction rate from IP to OP (%drop) and a fat fraction (%fat) were calculated.</p></div><div><h3>Results</h3><p>All biomarkers showed a significant difference between metastatic and benign lesions (P < 0.001). When comparing the AUCs, the %drop of T1WI had the highest AUC (0.934). Although the AUC of %fat from T2WI was significantly lower than that of other biomarkers, the %drop of T2WI was not significantly different from the %drop of T1WI (p = 0.339). The optimal threshold of %drop to differentiate metastatic and benign lesions was 22.0 in T1WI and 15.9 in T2WI. The inter-reader agreement was excellent for all biomarkers (0.82–0.86).</p></div><div><h3>Conclusion</h3><p>While %drop of T1WI showed the highest diagnostic performance to differentiate bone metastasis from benign lesions, the %drop of T2WI showed a comparable ability using a threshold 15.9.</p></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":null,"pages":null},"PeriodicalIF":2.0,"publicationDate":"2023-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352047723000679/pdfft?md5=6e2e8903363216b187290fcb6966819c&pid=1-s2.0-S2352047723000679-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138467291","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-10DOI: 10.1016/j.ejro.2023.100538
Achala Donuru , Tetsuro Araki , Farouk Dako , Jaydev K. Dave , Raul Porto Perez , Dongming Xu , Arun Nachiappan , Eduardo Mortani Barbosa Jr , Peter Noel , Harold Litt , Friedrich Knollman
Purpose
To investigate if clinical non-contrast chest CT studies obtained with PCD CT using much lower radiation exposure can achieve the same image quality as with the currently established EID protocol.
Materials/methods
A total of seventy-one patients were identified who had a non-contrast chest computed tomography (CT) done on PCD CT and EID CT scanners within a 4-month interval. Five fellowship trained chest radiologists, blinded to the scanner details were asked to review the cases side-by-side and record their preference for images from either the photon-counting-detector (PCD) CT or the energy-integrating detector (EID) CT scanner.
Results
The median CTDIvol for PCD-CT system was 4.710 mGy and EID system was 7.80 mGy (p < 0.001). The median DLP with the PCD-CT was 182.0 mGy.cm and EID system was 262.60 mGy.cm (p < 0.001). The contrast to noise ratio (CNR) was superior on the PCD-CT system 59.2 compared to the EID-CT 53.3; (p < 0.001). Kappa-statistic showed that there was poor agreement between the readers over the image quality from the PCD and EID scanners (κ = 0.19; 95 % CI: 0.12 – 0.27; p < 0.001). Chi-square analysis revealed that 3 out of 5 readers showed a significant preference for images from the PCDCT (p ≤ 0.012). There was no significant difference in the preferences of two readers between EID-CT and PCD-CT images.
Conclusion
The first clinical PCD-CT system allows a significant reduction in radiation exposure while maintaining image quality and image noise using a standardized non-contrast chest CT protocol.
{"title":"Photon-counting detector CT allows significant reduction in radiation dose while maintaining image quality and noise on non-contrast chest CT","authors":"Achala Donuru , Tetsuro Araki , Farouk Dako , Jaydev K. Dave , Raul Porto Perez , Dongming Xu , Arun Nachiappan , Eduardo Mortani Barbosa Jr , Peter Noel , Harold Litt , Friedrich Knollman","doi":"10.1016/j.ejro.2023.100538","DOIUrl":"https://doi.org/10.1016/j.ejro.2023.100538","url":null,"abstract":"<div><h3>Purpose</h3><p>To investigate if clinical non-contrast chest CT studies obtained with PCD CT using much lower radiation exposure can achieve the same image quality as with the currently established EID protocol.</p></div><div><h3>Materials/methods</h3><p>A total of seventy-one patients were identified who had a non-contrast chest computed tomography (CT) done on PCD CT and EID CT scanners within a 4-month interval. Five fellowship trained chest radiologists, blinded to the scanner details were asked to review the cases side-by-side and record their preference for images from either the photon-counting-detector (PCD) CT or the energy-integrating detector (EID) CT scanner.</p></div><div><h3>Results</h3><p>The median CTDIvol for PCD-CT system was 4.710 mGy and EID system was 7.80 mGy (p < 0.001). The median DLP with the PCD-CT was 182.0 mGy.cm and EID system was 262.60 mGy.cm (p < 0.001). The contrast to noise ratio (CNR) was superior on the PCD-CT system 59.2 compared to the EID-CT 53.3; (p < 0.001). Kappa-statistic showed that there was poor agreement between the readers over the image quality from the PCD and EID scanners (κ = 0.19; 95 % CI: 0.12 – 0.27; p < 0.001). Chi-square analysis revealed that 3 out of 5 readers showed a significant preference for images from the PCDCT (p ≤ 0.012). There was no significant difference in the preferences of two readers between EID-CT and PCD-CT images.</p></div><div><h3>Conclusion</h3><p>The first clinical PCD-CT system allows a significant reduction in radiation exposure while maintaining image quality and image noise using a standardized non-contrast chest CT protocol.</p></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":null,"pages":null},"PeriodicalIF":2.0,"publicationDate":"2023-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352047723000643/pdfft?md5=e2e7964845d89b78014e8b130634a874&pid=1-s2.0-S2352047723000643-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"92071539","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The etiology of aortic stenosis (AS) significantly impacts transcatheter heart valve (THV) implantation, with rheumatic etiology posing challenges. The concept of valve anchoring during transcatheter aortic valve replacement (TAVR) for patients with aortic regurgitation (AR) remains unclear.
Objective
This study aims to investigate the clinical and CT anatomical characteristics of various aortic valve diseases.
Methods
A retrospective analysis was conducted on consecutive patients who underwent CT for severe aortic diseases between April 2019 and February 2023. CT analysis was performed in eight anatomical landmarks: left ventricular outflow tract (LVOT), aortic annulus, sinus of Valsalva (SOV), sinotubular junction (STJ), ascending aorta (AAO), coronary height, aortic angle, and aortic valve calcification volume.
Results
121 patients with severe aortic valve disease were included, divided into AS (71 cases, 59%) and AR (50 cases, 41%) groups. In patients with AR, the absolute diameters of the annulus, LVOT, SOV, STJ, and AAO, as well as the heights of SOV and STJ and the cardiac angle, are larger than those in patients with AS (all P < 0.05). In normalized aortic root dimensions, the AR group had a higher SOV and STJ diameter-to-annulus ratio than the AS group (STJ-SOV-annulus: 1.51–1.44–1.00 vs 1.33–1.28–1.00). The bicuspid and rheumatic AS groups had smaller sinuses (STJ-SOV-annulus:1.27–1.35–1.00, 1.17–1.30–1.00, respectively), necessitating the downsizing of the THV. For 74% of AR patients, the sinotubular junction could not be used as a second anchoring zone, and anchoring relied primarily on the annulus.
Conclusions
Patients with rheumatic etiology require smaller valves, and anchoring in AR patients depends on the valve annulus. These structural characteristics will influence TAVR selection.
主动脉瓣狭窄(AS)的病因学对经导管心脏瓣膜(THV)植入术有重要影响，其中风湿病病因学提出了挑战。主动脉瓣反流(AR)患者经导管主动脉瓣置换术(TAVR)中瓣膜锚定的概念尚不清楚。目的探讨各种主动脉瓣病变的临床及CT解剖特点。方法回顾性分析2019年4月至2023年2月连续行CT检查的重症主动脉病变患者。CT分析左室流出道(LVOT)、主动脉环、Valsalva窦(SOV)、窦管交界处(STJ)、升主动脉(AAO)、冠状动脉高度、主动脉角、主动脉瓣钙化体积等8个解剖标志。结果121例重度主动脉瓣病变患者分为AS组(71例，59%)和AR组(50例，41%)。AR患者的环、LVOT、SOV、STJ、AAO的绝对直径以及SOV、STJ的高度和心角均大于as患者(P <0.05)。在标准化主动脉根部尺寸方面，AR组SOV和STJ直径与环空比高于AS组(STJ-SOV-环空:1.51-1.44-1.00 vs 1.33-1.28-1.00)。二尖瓣AS组和风湿性AS组鼻窦较小(STJ-SOV-annulus分别为1.27-1.35-1.00、1.17-1.30-1.00)，需要缩小THV。对于74%的AR患者，窦小管交界处不能作为第二锚定区，锚定主要依赖于环空。结论风湿病患者需要更小的瓣膜，AR患者的锚定取决于瓣膜环。这些结构特征将影响TAVR的选择。
{"title":"Anatomical characteristics of aortic valve diseases: Implications for transcatheter aortic valve replacement","authors":"Yanren Peng, Xiaorong Shu, Yongqing Lin, Weibin Huang, Shuwan Xu, Jianming Zheng, Ruqiong Nie","doi":"10.1016/j.ejro.2023.100532","DOIUrl":"https://doi.org/10.1016/j.ejro.2023.100532","url":null,"abstract":"<div><h3>Background</h3><p>The etiology of aortic stenosis (AS) significantly impacts transcatheter heart valve (THV) implantation, with rheumatic etiology posing challenges. The concept of valve anchoring during transcatheter aortic valve replacement (TAVR) for patients with aortic regurgitation (AR) remains unclear.</p></div><div><h3>Objective</h3><p>This study aims to investigate the clinical and CT anatomical characteristics of various aortic valve diseases.</p></div><div><h3>Methods</h3><p>A retrospective analysis was conducted on consecutive patients who underwent CT for severe aortic diseases between April 2019 and February 2023. CT analysis was performed in eight anatomical landmarks: left ventricular outflow tract (LVOT), aortic annulus, sinus of Valsalva (SOV), sinotubular junction (STJ), ascending aorta (AAO), coronary height, aortic angle, and aortic valve calcification volume.</p></div><div><h3>Results</h3><p>121 patients with severe aortic valve disease were included, divided into AS (71 cases, 59%) and AR (50 cases, 41%) groups. In patients with AR, the absolute diameters of the annulus, LVOT, SOV, STJ, and AAO, as well as the heights of SOV and STJ and the cardiac angle, are larger than those in patients with AS (all <em>P</em> < 0.05). In normalized aortic root dimensions, the AR group had a higher SOV and STJ diameter-to-annulus ratio than the AS group (STJ-SOV-annulus: 1.51–1.44–1.00 vs 1.33–1.28–1.00). The bicuspid and rheumatic AS groups had smaller sinuses (STJ-SOV-annulus:1.27–1.35–1.00, 1.17–1.30–1.00, respectively), necessitating the downsizing of the THV. For 74% of AR patients, the sinotubular junction could not be used as a second anchoring zone, and anchoring relied primarily on the annulus.</p></div><div><h3>Conclusions</h3><p>Patients with rheumatic etiology require smaller valves, and anchoring in AR patients depends on the valve annulus. These structural characteristics will influence TAVR selection.</p></div>","PeriodicalId":38076,"journal":{"name":"European Journal of Radiology Open","volume":null,"pages":null},"PeriodicalIF":2.0,"publicationDate":"2023-11-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2352047723000588/pdfft?md5=b68b6ca920ffe5fc5db2b20b2bb2b412&pid=1-s2.0-S2352047723000588-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"92014664","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号