ABSTRACT
PURPOSE: The purpose of this study was to compare ultra-low dose (ULD) and standard low-dose (SLD) chest computed tomography (CT) in terms of radiation exposure, image quality and diagnostic value for diagnosing pulmonary arteriovenous malformation (AVM) in patients with hereditary hemorrhagic telangiectasia (HHT). MATERIALS AND METHODS: In this prospective board-approved study consecutive patients with HHT referred to a reference center for screening and/or follow-up chest CT examination were prospectively included from December 2020 to January 2022. Patients underwent two consecutive non-contrast chest CTs without dose modulation (i.e., one ULD protocol [80 kVp or 100 kVp, CTDIvol of 0.3 mGy or 0.6 mGy] and one SLD protocol [140 kVp, CTDIvol of 1.3 mGy]). Objective image noises measured at the level of tracheal carina were compared between the two protocols. Overall image quality and diagnostic confidence were scored on a 4-point Likert scale (1 = insufficient to 4 = excellent). Sensitivity, specificity, positive predictive value and negative predictive value of ULD CT for diagnosing pulmonary AVM with a feeding artery of over 2 mm in diameter were calculated along with their 95% confidence intervals (CI) using SLD images as the standard of reference. RESULTS: A total of 44 consecutive patients with HHT (31 women; mean age, 42 ± 16 [standard deviation (SD)] years; body mass index, 23.2 ± 4.5 [SD] kg/m2) were included. Thirty-four pulmonary AVMs with a feeding artery of over 2 mm in diameter were found with SLD images versus 35 with ULD images. Sensitivity, specificity, predictive positive value, and predictive negative value of ULD CT for the diagnosis of PAVM were 100% (34/34; 95% CI: 90-100), 96% (18/19; 95% CI: 74-100), 97% (34/35; 95% CI: 85-100) and 100% (18/18; 95% CI: 81-100), respectively. A significant difference in diagnostic confidence scores was found between ULD (3.8 ± 0.4 [SD]) and SLD (3.9 ± 0.1 [SD]) CT images (P = 0.03). No differences in overall image quality scores were found between ULD CT examinations (3.9 ± 0.2 [SD]) and SLD (4 ± 0 [SD]) CT examinations (P = 0.77). Effective radiation dose decreased significantly by 78.8% with ULD protocol, with no significant differences in noise values between ULD CT images (16.7 ± 5.0 [SD] HU) and SLD images (17.7 ± 6.6 [SD] HU) (P = 0.07). CONCLUSION: ULD chest CT provides 100% sensitivity and 96% specificity for the diagnosis of treatable pulmonary AVM with a feeding artery of over 2 mm in diameter, leading to a 78.8% dose-saving compared with a standard low-dose protocol.
ABSTRACT
BACKGROUND AND OBJECTIVE: Prostate multiparametric magnetic resonance imaging (MRI) shows high sensitivity for International Society of Urological Pathology grade group (GG) ≥2 cancers. Many artificial intelligence algorithms have shown promising results in diagnosing clinically significant prostate cancer on MRI. To assess a region-of-interest-based machine-learning algorithm aimed at characterising GG ≥2 prostate cancer on multiparametric MRI. METHODS: The lesions targeted at biopsy in the MRI-FIRST dataset were retrospectively delineated and assessed using a previously developed algorithm. The Prostate Imaging-Reporting and Data System version 2 (PI-RADSv2) score assigned prospectively before biopsy and the algorithm score calculated retrospectively in the regions of interest were compared for diagnosing GG ≥2 cancer, using the areas under the curve (AUCs), and sensitivities and specificities calculated with predefined thresholds (PIRADSv2 scores ≥3 and ≥4; algorithm scores yielding 90% sensitivity in the training database). Ten predefined biopsy strategies were assessed retrospectively. KEY FINDINGS AND LIMITATIONS: After excluding 19 patients, we analysed 232 patients imaged on 16 different scanners; 85 had GG ≥2 cancer at biopsy. At patient level, AUCs of the algorithm and PI-RADSv2 were 77% (95% confidence interval [CI]: 70-82) and 80% (CI: 74-85; p = 0.36), respectively. The algorithm's sensitivity and specificity were 86% (CI: 76-93) and 65% (CI: 54-73), respectively. PI-RADSv2 sensitivities and specificities were 95% (CI: 89-100) and 38% (CI: 26-47), and 89% (CI: 79-96) and 47% (CI: 35-57) for thresholds of ≥3 and ≥4, respectively. Using the PI-RADSv2 score to trigger a biopsy would have avoided 26-34% of biopsies while missing 5-11% of GG ≥2 cancers. Combining prostate-specific antigen density, the PI-RADSv2 and algorithm's scores would have avoided 44-47% of biopsies while missing 6-9% of GG ≥2 cancers. Limitations include the retrospective nature of the study and a lack of PI-RADS version 2.1 assessment. CONCLUSIONS AND CLINICAL IMPLICATIONS: The algorithm provided robust results in the multicentre multiscanner MRI-FIRST database and could help select patients for biopsy. PATIENT SUMMARY: An artificial intelligence-based algorithm aimed at diagnosing aggressive cancers on prostate magnetic resonance imaging showed results similar to expert human assessment in a prospectively acquired multicentre test database.
ABSTRACT
Rationale and objectives: To develop a Natural Language Processing (NLP) method based on Bidirectional Encoder Representations from Transformers (BERT) adapted to French CT reports and to evaluate its performance to calculate the diagnostic yield of CT in patients with clinical suspicion of pulmonary embolism (PE). Materials and methods: All the CT reports performed in our institution in 2019 (99,510 reports, training and validation dataset) and 2018 (94,559 reports, testing dataset) were included after anonymization. Two BERT-based NLP sentence classifiers were trained on 27.700, manually labeled, sentences from the training dataset. The first one aimed to classify the reports' sentences into three classes ("Non chest", "Healthy chest", and "Pathological chest" related sentences), the second one to classify the last class into eleven sub classes pathologies including "pulmonary embolism". F1-score was reported on the validation dataset. These NLP classifiers were then applied to requested CT reports for pulmonary embolism from the testing dataset. Sensitivity, specificity, and accuracy for detection of the presence of a pulmonary embolism were reported in comparison to human analysis of the reports. Results: The F1-score for the 3-Classes and 11-SubClasses classifiers was 0.984 and 0.985, respectively. 4,042 examinations from the testing dataset were requested for pulmonary embolism of which 641 (15.8%) were positively evaluated by radiologists. The sensitivity, specificity, and accuracy of the NLP network for identifying pulmonary embolism in these reports were 98.2%, 99.3% and 99.1%, respectively. Conclusion: BERT-based NLP sentences classifier enables the analysis of large databases of radiological reports to accurately determine the diagnostic yield of CT screening.
ABSTRACT
INTRODUCTION: Prostate multiparametric MRI (mpMRI) has shown good sensitivity in detecting cancers with an International Society of Urological Pathology (ISUP) grade of ≥2. However, it lacks specificity, and its inter-reader reproducibility remains moderate. Biomarkers, such as the Prostate Health Index (PHI), may help select patients for prostate biopsy. Computer-aided diagnosis/detection (CAD) systems may also improve mpMRI interpretation. Different prototypes of CAD systems are currently developed under the Recherche Hospitalo-Universitaire en Santé / Personalized Focused Ultrasound Surgery of Localized Prostate Cancer (RHU PERFUSE) research programme, tackling challenging issues such as robustness across imaging protocols and magnetic resonance (MR) vendors, and ability to characterise cancer aggressiveness. The study primary objective is to evaluate the non-inferiority of the area under the receiver operating characteristic curve of the final CAD system as compared with the Prostate Imaging-Reporting and Data System V.2.1 (PI-RADS V.2.1) in predicting the presence of ISUP ≥2 prostate cancer in patients undergoing prostate biopsy. METHODS: This prospective, multicentre, non-inferiority trial will include 420 men with suspected prostate cancer, a prostate-specific antigen level of ≤30 ng/mL and a clinical stage ≤T2 c. Included men will undergo prostate mpMRI that will be interpreted using the PI-RADS V.2.1 score. Then, they will undergo systematic and targeted biopsy. PHI will be assessed before biopsy. At the end of patient inclusion, MR images will be assessed by the final version of the CAD system developed under the RHU PERFUSE programme. Key secondary outcomes include the prediction of ISUP grade ≥2 prostate cancer during a 3-year follow-up, and the number of biopsy procedures saved and ISUP grade ≥2 cancers missed by several diagnostic pathways combining PHI and MRI findings. ETHICS AND DISSEMINATION: Ethical approval was obtained from the Comité de Protection des Personnes Nord Ouest III (ID-RCB: 2020-A02785-34). After publication of the results, access to MR images will be possible for testing other CAD systems. TRIAL REGISTRATION NUMBER: NCT04732156.
Subject(s)
Multiparametric Magnetic Resonance Imaging , Prostatic Neoplasms , Artificial Intelligence , Humans , Image-Guided Biopsy/methods , Magnetic Resonance Imaging/methods , Male , Prospective Studies , Prostatic Neoplasms/diagnosis , Reproducibility of Results , Retrospective StudiesABSTRACT
Objectives: We evaluated the contribution of lung lesion quantification on chest CT using a clinical Artificial Intelligence (AI) software in predicting death and intensive care units (ICU) admission for COVID-19 patients. Methods: For 349 patients with positive COVID-19-PCR test that underwent a chest CT scan at admittance or during hospitalization, we applied the AI for lung and lung lesion segmentation to obtain lesion volume (LV), and LV/Total Lung Volume (TLV) ratio. ROC analysis was used to extract the best CT criterion in predicting death and ICU admission. Two prognostic models using multivariate logistic regressions were constructed to predict each outcome and were compared using AUC values. The first model ("Clinical") was based on patients' characteristics and clinical symptoms only. The second model ("Clinical+LV/TLV") included also the best CT criterion. Results: LV/TLV ratio demonstrated best performance for both outcomes; AUC of 67.8% (95% CI: 59.5 - 76.1) and 81.1% (95% CI: 75.7 - 86.5) respectively. Regarding death prediction, AUC values were 76.2% (95% CI: 69.9 - 82.6) and 79.9% (95%IC: 74.4 - 85.5) for the "Clinical" and the "Clinical+LV/TLV" models respectively, showing significant performance increase (+ 3.7%; p-value<0.001) when adding LV/TLV ratio. Similarly, for ICU admission prediction, AUC values were 74.9% (IC 95%: 69.2 - 80.6) and 84.8% (IC 95%: 80.4 - 89.2) respectively corresponding to significant performance increase (+ 10%: p-value<0.001). Conclusions: Using a clinical AI software to quantify the COVID-19 lung involvement on chest CT, combined with clinical variables, allows better prediction of death and ICU admission.
ABSTRACT
BACKGROUND: Previous studies have suggested the role of microcalcifications in plaque vulnerability. This exploratory study sought to assess the potential of hybrid positron-emission tomography (PET)/magnetic resonance imaging (MRI) using 18F-sodium fluoride (18F-NaF) to check simultaneously 18F-NaF uptake, a marker of microcalcifications, and morphological criteria of vulnerability. METHODS AND RESULTS: We included 12 patients with either recently symptomatic or asymptomatic carotid stenosis. All patients underwent 18F-NaF PET/MRI. 18F-NaF target-to-background ratio (TBR) was measured in culprit and nonculprit (including contralateral plaques of symptomatic patients) plaques as well as in other arterial walls. Morphological criteria of vulnerability were assessed on MRI. Mineral metabolism markers were also collected. 18F-NaF uptake was higher in culprit compared to nonculprit plaques (median TBR 2.6 [2.2-2.8] vs 1.7 [1.3-2.2]; P = 0.03) but was not associated with morphological criteria of vulnerability on MRI. We found a positive correlation between 18F-NaF uptake and calcium plaque volume and ratio but not with circulating tissue-nonspecific alkaline phosphatase (TNAP) activity and inorganic pyrophosphate (PPi) levels. 18F-NaF uptake in the other arterial walls did not differ between symptomatic and asymptomatic patients. CONCLUSIONS: 18F-NaF PET/MRI may be a promising tool for providing additional insights into the plaque vulnerability.
Subject(s)
Calcinosis , Carotid Stenosis , Plaque, Atherosclerotic , Calcinosis/diagnostic imaging , Carotid Stenosis/diagnostic imaging , Humans , Magnetic Resonance Imaging , Plaque, Atherosclerotic/diagnostic imaging , Positron-Emission Tomography/methods , Sodium FluorideABSTRACT
OBJECTIVES: To assess the diagnostic performances of chest CT for triage of patients in multiple emergency departments during COVID-19 epidemic, in comparison with reverse transcription polymerase chain reaction (RT-PCR) test. METHOD: From March 3 to April 4, 2020, 694 consecutive patients from three emergency departments of a large university hospital, for which a hospitalization was planned whatever the reasons, i.e., COVID- or non-COVID-related, underwent a chest CT and one or several RT-PCR tests. Chest CTs were rated as "Surely COVID+," "Possible COVID+," or "COVID-" by experienced radiologists. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated using the final RT-PCR test as standard of reference. The delays for CT reports and RT-PCR results were recorded and compared. RESULTS: Among the 694 patients, 287 were positive on the final RT-PCR exam. Concerning the 694 chest CT, 308 were rated as "Surely COVID+", 34 as "Possible COVID+," and 352 as "COVID-." When considering only the "Surely COVID+" CT as positive, accuracy, sensitivity, specificity, PPV, and NPV reached 88.9%, 90.2%, 88%, 84.1%, and 92.7%, respectively, with respect to final RT-PCR test. The mean delay for CT reports was three times shorter than for RT-PCR results (187 ± 148 min versus 573 ± 327 min, p < 0.0001). CONCLUSION: During COVID-19 epidemic phase, chest CT is a rapid and most probably an adequately reliable tool to refer patients requiring hospitalization to the COVID+ or COVID- hospital units, when response times for virological tests are too long. KEY POINTS: ⢠In a large university hospital in Lyon, France, the accuracy, sensitivity, specificity, PPV, and NPV of chest CT for COVID-19 reached 88.9%, 90.2%, 88%, 84.1%, and 92.7%, respectively, using RT-PCR as standard of reference. ⢠The mean delay for CT reports was three times shorter than for RT-PCR results (187 ± 148 min versus 573 ± 327 min, p < 0.0001). ⢠Due to high accuracy of chest CT for COVID-19 and shorter time for CT reports than RT-PCR results, chest CT can be used to orient patients suspected to be positive towards the COVID+ unit to decrease congestion in the emergency departments.
Subject(s)
COVID-19/diagnostic imaging , Triage , Aged , Aged, 80 and over , COVID-19/epidemiology , Emergency Service, Hospital , Epidemics , Female , France , Hospitals, University , Humans , Male , Predictive Value of Tests , SARS-CoV-2 , Time Factors , Tomography, X-Ray ComputedABSTRACT
OBJECTIVES: The aim of the present study was to assess the association between collateral status and DWI-FLAIR mismatch in patients with acute ischemic stroke within the 4.5â¯h time-window. METHODS: We analysed DWI, FLAIR, and PWI data in patients within 4.5â¯h after symptom onset from the I-KNOW European database. Collateral flow maps were graded by analyzing contrast 'staining' extent over the early, mid and late perfusion phases. ADC values, DWI lesion volume, and normalised perfusion parameters (CBV,Tmax) within DWI lesions were determined. Visibility of parenchymal hyperintensivty on FLAIR was evaluated ("FLAIR positive"), and DWI-FLAIR mismatch was assessed. Spontaneously reperfused regions were defined as voxels with Tmax <6â¯s within the DWI lesion. Final infarct size was assessed on day-30 FLAIR images. RESULTS: Of the 168 patients included in I-KNOW database, 87 were eligible for this study. DWI-FLAIR mismatch was present in 69 patients. There was no difference between poor and good collaterals status according to age, sex, baseline NIHSS score, time to MRI and DWI lesion volume. Collateral status was significantly better in the FLAIR positive group (pâ¯=â¯.001). Patients with poor collaterals had significantly increased Tmax (pâ¯=â¯.005). Baseline DWI lesion volume and final lesion volume were significantly smaller in patients with good collateral status (pâ¯<â¯.001 and 0.01, respectively). CONCLUSIONS: We found that patients with early FLAIR lesion visibility have a better collateral status. This finding has implications for the management of stroke patients with unknown time-of-onset, and more widely should be considered in the current context of extending the therapeutic window.
Subject(s)
Collateral Circulation/physiology , Diffusion Magnetic Resonance Imaging/methods , Magnetic Resonance Angiography/methods , Stroke/diagnostic imaging , Aged , Aged, 80 and over , Brain Ischemia/complications , Female , Humans , Male , Middle Aged , Prospective Studies , Stroke/etiology , Thrombolytic Therapy , Time FactorsABSTRACT
BACKGROUND: Brain atrophy has shown a protective effect on the risk of early neurological deterioration (END) related to malignant edema in patients with hemispheric infarction but could be deleterious on the outcome. AIMS: We aimed to assess whether brain atrophy has an impact on the risk of END and on the outcome in severe ischemic strokes after intravenous (IV) thrombolysis. METHODS: From a prospective thrombolysis registry, 137 patients who had a National Institutes of Health Stroke Scale (NIHSS) ≥15, MRI at admission, and IV thrombolysis were included. Relative cerebral volume was calculated. END was defined as a ≥2-points deterioration 72-h NIHSS and a good outcome as a modified Rankin Scale (mRS) ≤2 at 3 months. A multiple logistic regression analysis with a stepwise backward procedure was performed. RESULTS: END and a good outcome were observed, respectively, in 20 (14.6%) and 48 (37.5%) patients. In univariate analysis, predictors of END included age (p = 0.049), diabetes (p = 0.041), and parenchymal hemorrhage (p = 0.039). In multivariate analysis, age (p = 0.018) was significantly associated with END. Brain atrophy was not associated with END even in subgroup analysis according to the baseline infarct size. In univariate analysis, age (p = 0.003), prestroke mRS (p = 0.002), hypertension (p = 0.006), baseline NIHSS (p = 0.002), END (p = 0.002), proximal occlusion (p = 0.006), and recanalization at 24 h (p < 0.001) were associated with a good outcome. Only baseline NIHSS (p = 0.006) was associated with a good outcome after adjustment. CONCLUSIONS: We did not find any impact of brain atrophy on the risk of END and the outcome at 3 months in severe ischemic strokes after IV thrombolysis.
Subject(s)
Atrophy/pathology , Recovery of Function/physiology , Stroke/drug therapy , Stroke/pathology , Thrombolytic Therapy/methods , Administration, Intravenous , Aged , Aged, 80 and over , Brain/pathology , Brain Ischemia/drug therapy , Brain Ischemia/pathology , Female , Humans , Magnetic Resonance Imaging , Male , Middle Aged , Prospective Studies , Treatment OutcomeABSTRACT
OBJECTIVES: Endovascular treatment of thoracic aortic lesions appears to be advantageous. However, long-term outcomes remain poorly reported. This retrospective study reported 6-year outcomes of thoracic endovascular aortic repair. METHODS: A total of 74 patients underwent endovascular thoracic aorta treatments between 1999 and 2007; 13 had thoracic aortic dissections, 19 had traumatic aortic injuries, 35 had aneurysms, 6 had pseudoaneurysms and 1 had a penetrating ulcer. The mean follow-up was 66 months after 30 perioperative days. Yearly follow-ups included computed tomography angiography or magnetic resonance angiography. Patient demographics, mortality, complications and reinterventions were analysed. RESULTS: The early 30-day mortality and the overall late mortality were 9.5 (7/74) and 37.8% (28/74), respectively. Late mortality was higher in patients with aneurysms than in the other groups (20/35; 57% vs 8/39; 20.5%; P = 0.002). Aortic-related mortality occurred in 5/35 (14%) patients with aneurysms, but not in other groups (P = 0.02). No relationships among late complications were found for traumatic aortic injuries. The most common complication was an endoleak (21/74, 28.4%), which occurred more frequently with aneurysms than other disorders (18/35, 51.4% vs 3/39, 7.7%; P < 0.001). Endoleaks also occurred most frequently in aortic-related deaths (16/69 vs 5/5; P = 0.001). Type 1 endoleaks occurred significantly more often with aneurysms (13/35) than with other disorders (P = 0.004). Reintervention was required in 9 patients (12%); 8 with atherosclerotic aneurysms (8/35; 23%). A false lumen was thrombosed in 54% of dissections (7/13), and shrank in 39% (5/13). CONCLUSIONS: Long-term outcomes depended on aortic pathology. Aortic aneurysms were the most complicated and caused the highest mortality, probably due to atherosclerotic disease evolution. Patients with traumatic aortic injuries appeared to have the best long-term outcomes.
Subject(s)
Aorta, Thoracic/surgery , Aortic Diseases/surgery , Endovascular Procedures , Adult , Aortic Dissection/mortality , Aortic Dissection/surgery , Aneurysm, False/mortality , Aneurysm, False/surgery , Aorta, Thoracic/injuries , Aortic Aneurysm, Thoracic/mortality , Aortic Aneurysm, Thoracic/surgery , Aortic Diseases/diagnostic imaging , Aortic Diseases/mortality , Endovascular Procedures/adverse effects , Endovascular Procedures/mortality , Female , Humans , Male , Middle Aged , Survival Analysis , Tomography, X-Ray Computed , Treatment OutcomeABSTRACT
This prospective study aimed to assess the incidence of silent hypersensitivity to Escherichia coli asparaginase in the treatment of acute lymphoblastic leukemia (ALL). Thirty-three children with newly diagnosed ALL were included in the study and treated according to the FRALLE 2000 protocol. The 'A group' (n = 18) differed from the 'B-T group' (n = 15) by a less intensive chemotherapy, the absence of concurrent prednisone therapy, and different asparaginase administration modalities during the second intensification. Asparagine, asparaginase activity, and anti-asparaginase antibodies were measured in each phase before the next injection of asparaginase. Eighteen percent of children presented a silent hypersensitivity. Most of them were in the 'B-T group' (p = 0.07), and maintained low antibody titers throughout the treatment. Clinical hypersensitivity was statistically more frequent in group A (p = 0.002), and allergy occurred mainly during the second intensification when antibody concentrations were significantly increased. We did not find any significant difference between asparaginase activity or asparagine depletion between the silent hypersensitivity and clinical allergy groups. In all, the results of this study suggest that chemotherapy and corticosteroid therapy associated with asparaginase treatment can lower antibody production and contribute to maintaining a silent hypersensitivity state.
Subject(s)
Antineoplastic Agents/therapeutic use , Asparaginase/adverse effects , Drug Hypersensitivity , Escherichia coli/enzymology , Precursor Cell Lymphoblastic Leukemia-Lymphoma/drug therapy , Adolescent , Antibodies, Anti-Idiotypic/immunology , Antibodies, Anti-Idiotypic/metabolism , Asparaginase/immunology , Asparagine/metabolism , Child , Child, Preschool , Enzyme-Linked Immunosorbent Assay , Female , Humans , Infant , Male , Precursor Cell Lymphoblastic Leukemia-Lymphoma/blood , Precursor Cell Lymphoblastic Leukemia-Lymphoma/diagnosis , Prognosis , Prospective Studies , Survival RateABSTRACT
BACKGROUND: Production of high levels of IL-6 is often correlated with resistance to cytotoxics or ionizing radiations, in cancer cell lines as in various cancer patients. We investigated whether monoclonal antibodies directed against IL-6 may enable to reverse resistance of cancer cell lines. METHODOLOGY/PRINCIPAL FINDINGS: We exposed ten haematological cancer cells from lymphoma, myeloma, or leukemia origins to cytotoxics or ionizing radiations and assessed the effects of anti-IL-6 antibody addition on cell proliferation, apoptosis, or IL-6 signaling. A strong correlation between IL-6 secretion, measured by ELISA, and resistance to doxorubicin as ionizing radiations was observed in the multiple myeloma U266 and the Burkitt's lymphoma Daudi and Namalwa cells. Although an anti-IL-6 antibody combined to both treatments efficiently blocked IL-6 signaling in U266 cells, expressing the IL-6 receptor gp80, it did not increase treatment-induced anti-proliferative and pro-apoptotic effects on these cells, as well as on Daudi and Namalwa cells. This lack of effect could be related to diverse factors: 1) a higher release of the soluble form of IL-6 receptor gp80 in response to doxorubicin and irradiation from all cell lines, 2) an impaired level of the IL-6 pathway inhibitor SOCS3 in Daudi cells, and 3) an increased release of IL-10 and TNFalpha, two cytokines involved in cell radio- and chemoresistance. CONCLUSIONS/SIGNIFICANCE: These data support the fact that IL-6 is not the preponderant actor of cell resistance to cytotoxics and ionizing radiations, which seems to be regulated by a complex network of proteins.