Research in diagnostic and interventional imaging最新文献

{"title":"Additional value of chest CT AI-based quantification of lung involvement in predicting death and ICU admission for COVID-19 patients","authors":"Eloise Galzin ,&nbsp;Laurent Roche ,&nbsp;Anna Vlachomitrou ,&nbsp;Olivier Nempont ,&nbsp;Heike Carolus ,&nbsp;Alexander Schmidt-Richberg ,&nbsp;Peng Jin ,&nbsp;Pedro Rodrigues ,&nbsp;Tobias Klinder ,&nbsp;Jean-Christophe Richard ,&nbsp;Karim Tazarourte ,&nbsp;Marion Douplat ,&nbsp;Alain Sigal ,&nbsp;Maude Bouscambert-Duchamp ,&nbsp;Salim Aymeric Si-Mohamed ,&nbsp;Sylvain Gouttard ,&nbsp;Adeline Mansuy ,&nbsp;François Talbot ,&nbsp;Jean-Baptiste Pialat ,&nbsp;Olivier Rouvière ,&nbsp;Loic Boussel","doi":"10.1016/j.redii.2022.100018","DOIUrl":"10.1016/j.redii.2022.100018","url":null,"abstract":"<div><h3>Objectives</h3><p>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.</p></div><div><h3>Methods</h3><p>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.</p></div><div><h3>Results</h3><p>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&lt;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&lt;0.001).</p></div><div><h3>Conclusions</h3><p>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.</p></div>","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"4 ","pages":"Article 100018"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9716289/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"9583093","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}

{"title":"Performance of ultrasound guidance for vacuum-assisted biopsy of breast microcalcifications without associated mass","authors":"S. Le Cam ,&nbsp;Y. Badachi ,&nbsp;S. Ayadi ,&nbsp;O. Lucidarme","doi":"10.1016/j.redii.2022.100012","DOIUrl":"10.1016/j.redii.2022.100012","url":null,"abstract":"","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"3 ","pages":"Article 100012"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2772652522000126/pdfft?md5=a2a89cdadd9b98514d6da6175e622e41&pid=1-s2.0-S2772652522000126-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47031241","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}

{"title":"Correlation between native CT density of thyroid parenchyma and thyroid function tests","authors":"Dr. Padma Vikram Badhe M.D. (Radiodiagnosis (Professor)) ,&nbsp;Dr. Moinuddin Sultan M.D. (Radiodiagnosis Senior Resident) ,&nbsp;Dr. Sanika Patil DMRD (Former Senior Resident) ,&nbsp;Dr. Ajith Varrior (Junior Resident) ,&nbsp;Dr. Tejas Ghodasara (Junior Resident) ,&nbsp;Dr. Gautham Shankar (Junior Resident) ,&nbsp;Dr. Satyam Barchha (Junior Resident)","doi":"10.1016/j.redii.2022.100016","DOIUrl":"10.1016/j.redii.2022.100016","url":null,"abstract":"","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"3 ","pages":"Article 100016"},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2772652522000163/pdfft?md5=4fcd65118e56d0a742a37b7884b371c3&pid=1-s2.0-S2772652522000163-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47915302","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}

{"title":"CT in non-traumatic acute abdominal emergencies: Comparison of unenhanced acquisitions and single-energy iodine mapping for the characterization of bowel wall enhancement","authors":"Sophie Boyer,&nbsp;Charles Lombard,&nbsp;Ayla Urbaneja,&nbsp;Céline Vogrig,&nbsp;Denis Regent,&nbsp;Alain Blum,&nbsp;Pedro Augusto Gondim Teixeira","doi":"10.1016/j.redii.2022.100010","DOIUrl":"10.1016/j.redii.2022.100010","url":null,"abstract":"<div><h3>Objectives</h3><p>To evaluate the benefit of unenhanced CT and single energy iodine mapping (SIM) to conventional contrast-enhanced CT for bowel wall enhancement characterization in an acute abdomen setting.</p></div><div><h3>Methods</h3><p>CT images from 45 patients with a suspected acute abdomen who underwent abdominopelvic CT from April 2018 to June 2018 were analyzed retrospectively by two independent radiologists. These patients had been referred by emergency department physicians in a context of acute abdominal pain and had a confirmed etiological diagnosis. Three image sets were evaluated separately (portal phase images alone; portal phase images and unenhanced images, portal phase images, and single energy iodine maps). Diagnostic accuracy and confidence were assessed. Quantitative analysis of bowel wall enhancement was also performed.</p></div><div><h3>Results</h3><p>The number of correct diagnoses increased by 8% and 12% with unenhanced images and 6% and 13% with SIM for readers 1 and 2, respectively, compared to the portal phase only. There was an improvement in the confidence of the etiological diagnosis with the number of certain diagnoses increasing from 23% to 100%, which was statistically significant for reader 2 and of borderline significance for reader 1 (<em>P</em> = 0.002 and 0.052, respectively) when unenhanced phase and SIM were added. The inter-rater agreement improved when unenhanced and portal phase images were associated, compared to portal phase images alone (kappa = 0.652 [ICC=0.482–0.822] and 0.42 [ICC=0.241–0.607] respectively).</p></div><div><h3>Conclusion</h3><p>SIM and unenhanced images improve the reproducibility and the diagnostic confidence to diagnose ischemic and inflammatory/infectious bowel wall thickening compared to portal phase images alone</p></div><div><h3>Summary sentence</h3><p>The analysis of unenhanced and SIM images in association with portal phase images improves the reproducibility and the radiologist's confidence in the etiological diagnosis of acute non-traumatic bowel wall thickening in adults.</p></div>","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"2 ","pages":"Article 100010"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2772652522000102/pdfft?md5=3a976b9a4af0f72983e1a9c5962f7058&pid=1-s2.0-S2772652522000102-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42605168","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}

{"title":"MADplots: A methodology for visualizing and characterizing energy-dependent attenuation of tissues in spectral computed tomography","authors":"Matthew A. Lewis PhD ,&nbsp;Todd C. Soesbe PhD ,&nbsp;Xinhui Duan PhD ,&nbsp;Liran Goshen PhD ,&nbsp;Yoad Yagil PhD ,&nbsp;Shlomo Gotman MSc ,&nbsp;Robert E. Lenkinski PhD","doi":"10.1016/j.redii.2022.100011","DOIUrl":"10.1016/j.redii.2022.100011","url":null,"abstract":"<div><h3>Rationale and objectives</h3><p>A method for visualizing and analyzing the complete information contained in spectral CT scans using two-dimensional histograms (i.e. Material Attenuation Decomposition plots – MADplots) of the water-photoelectric attenuation versus water-scatter attenuation at the cohort (combination of multiple studies across patients), examination, series, slice, and organ/ROI levels is described.</p></div><div><h3>Materials and methods</h3><p>The appearance of a MADplot with several standard biological materials was predicted using ideal material properties available from NIST and the ICRU to generate a map for this non-spatial data space. Software tools were developed to generate MADplots as new DICOM series that facilitate spectral analysis. Illustrative examples were selected from an IRB-approved, retrospective cohort of Spectral Basis Images (SBIs) scanned using a pre-release, dual-layer detector spectral CT.</p></div><div><h3>Results</h3><p>By combining all of the voxels for contrast and non-contrast studies, the predicted appearance of the MADplot was confirmed. Locations of several kinds of tissue, the shape of the tissue distributions in normal lung, and the variations in the manner in which organ-specific MADplots change with pathology are demonstrated for the presence of fat in both the liver and pancreas highlighting the potential use for identifying pathologies on spectral CT images.</p></div><div><h3>Conclusions</h3><p>The examples of MADplots shown at cohort (combined studies), examination, series, slice, organ, and ROI levels illustrate their potential utility in analyzing and displaying spectral CT data. Future studies are directed at developing MADplot based organ segmentation and the automated detection and display of organ based pathologies.</p></div>","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"2 ","pages":"Article 100011"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2772652522000114/pdfft?md5=5f8611115bbe7efa1565b4432f1281c8&pid=1-s2.0-S2772652522000114-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47057273","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}

{"title":"Evaluating bone biopsy quality by technique in an animal model","authors":"Corey K Ho MD ,&nbsp;David Gimarc MD ,&nbsp;Hsieng-Feng Carroll PhD ,&nbsp;Michael Clay MD ,&nbsp;Jeffrey Schowinsky MD ,&nbsp;MK Jesse MD ,&nbsp;Amanda M Crawford MD ,&nbsp;Carrie B Marshall MD","doi":"10.1016/j.redii.2022.100008","DOIUrl":"10.1016/j.redii.2022.100008","url":null,"abstract":"<div><h3>Rationale and Objectives</h3><p>Powered bone biopsy technique is popular due to its ease of use. However, there is conflicting evidence regarding the diagnostic quality of the samples. The purpose of this study is to evaluate the diagnostic adequacy of different bone biopsy devices and techniques as it relates to the frequency of sample artifacts.</p></div><div><h3>Materials and Methods</h3><p>Bone biopsy was performed on same-day processed lamb femora using the following techniques: manual, pulsed powered and full powered. Ten samples were collected using each method by a single musculoskeletal-trained radiologist and were reviewed by 3 blinded pathologists. Samples were compared across multiple categories: length, bone dust, thermal/crush artifact, cellular morphology, fragmentation, and diagnostic acceptability. Bayesian Multilevel Nonlinear Regression models were performed assessing the association between the techniques across the categories.</p></div><div><h3>Results</h3><p>Statistical analysis revealed that the manual technique outperformed any powered technique across all categories: decreased thermal/crush artifact (<em>P</em> = 0.014), decreased bone dust (p&lt;0.001), better cellular morphology (<em>P</em> = 0.005), less fragmentation (<em>P</em> &lt; 0.0001) and better diagnostic acceptability (<em>P</em> &lt; 0.0001).</p></div><div><h3>Conclusion</h3><p>Manually obtained bone biopsy samples generally produce a more diagnostic sample as compared to powered techniques in an animal model. Given these results, manual bone biopsy methods should be encouraged after consideration for lesion composition, difficulty of access and the patient's overall condition.</p></div>","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"2 ","pages":"Article 100008"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2772652522000084/pdfft?md5=c0b43c642a873d8d251dc9ffdb132cfb&pid=1-s2.0-S2772652522000084-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44016643","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}

{"title":"Grading of soft tissues sarcomas using radiomics models: Choice of imaging methods and comparison with conventional visual analysis","authors":"Bailiang Chen ,&nbsp;Olivier Steinberger ,&nbsp;Roman Fenioux ,&nbsp;Quentin Duverger ,&nbsp;Tryphon Lambrou ,&nbsp;Gauthier Dodin ,&nbsp;Alain Blum ,&nbsp;Pedro Augusto Gondim Teixeira","doi":"10.1016/j.redii.2022.100009","DOIUrl":"10.1016/j.redii.2022.100009","url":null,"abstract":"<div><h3>Purpose</h3><p>To determine which combination of imaging modalities/contrast, radiomics models, and how many features provides the best diagnostic performance for the differentiation between low- and high-grade soft tissue sarcomas (STS) using a radiomics approach.</p></div><div><h3>Methods</h3><p>MRI and CT from 39 patients with a histologically confirmed STS were prospectively analyzed. Images were evaluated both quantitatively by radiomics models and qualitatively by visual evaluation (used as reference) for grading (low-grade vs high-grade). In radiomics analysis, 120 radiomic features were extracted and contributed into three models: least absolute shrinkage and selection operator with logistic regression(LASSO-LR), recursive feature elimination and cross-validation (RFECV-SVC) and analysis of variance with SVC (ANOVA-SVC). Those were applied to different combinations of imaging modalities acquisition, with and without contrast medium administration, as well as selected number of features.</p></div><div><h3>Results</h3><p>Fat-saturated T2w (FS-T2w) MR images using RFECV-SVC radiomic models involving five features yielded the best results with mean sensitivity, specificity, and accuracy of 92% ± 10%, 78% ± 30%, and 89% ± 12%, respectively. The performance of radiomics was better than that of conventional analysis (67% accuracy) for STS grading. Combination of multiple contrast or imaging modalities did not increase the diagnostic performance.</p></div><div><h3>Conclusion</h3><p>FS-T2w MR images alone with a five-feature radiomics analysis usingh REFCV-SVC model may be able to provide sufficient diagnositic performance compared to conventional visual evaluation with multiple MRI contrast and CT imaging.</p></div>","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"2 ","pages":"Article 100009"},"PeriodicalIF":0.0,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2772652522000096/pdfft?md5=7b415543fdbd132cdb3a7dc72a83f160&pid=1-s2.0-S2772652522000096-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47915158","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}

{"title":"Investigating of the role of CT scan for cancer patients during the first wave of COVID-19 pandemic","authors":"Sylvain Bourdoncle ,&nbsp;Thomas Eche ,&nbsp;Jeremy McGale ,&nbsp;Kevin Yiu ,&nbsp;Ephraïm Partouche ,&nbsp;Randy Yeh ,&nbsp;Samy Ammari ,&nbsp;Hervé Rousseau ,&nbsp;Laurent Dercle ,&nbsp;Fatima-Zohra Mokrane","doi":"10.1016/j.redii.2022.100004","DOIUrl":"10.1016/j.redii.2022.100004","url":null,"abstract":"<div><h3>Introduction</h3><p>Amidst this current COVID-19 pandemic, we undertook this systematic review to determine the role of medical imaging, with a special emphasis on computed tomography (CT), on guiding the care and management of oncologic patients.</p></div><div><h3>Material and Methods</h3><p>Study selection focused on articles from 01/02/2020 to 04/23/2020. After removal of irrelevant articles, all systematic or non-systematic reviews, comments, correspondence, editorials, guidelines and meta-analysis and case reports with less than 5 patients were also excluded. Full-text articles of eligible publications were reviewed to select all imaging-based publications, and the existence or not of an oncologic population was reported for each publication. Two independent reviewers collected the following information: ( 1) General publication data; (2) Study design characteristics; (3) Demographic, clinical and pathological variables with percentage of cancer patients if available; (4) Imaging performances. The sensitivity and specificity of chest CT (C-CT) were pooled separately using a random-effects model. The positive predictive value (PPV) and negative predictive value (NPV) of C-CT as a test was estimated for a wide range of disease prevalence rates.</p></div><div><h3>Results</h3><p>A total of 106 publications were fully reviewed. Among them, 96 were identified to have extractable data for a two-by-two contingency table for CT performance. At the end, 53 studies (including 6 that used two different populations) were included in diagnosis accuracy analysis (<em>N</em> = 59). We identified 53 studies totaling 11,352 patients for whom the sensitivity (95CI) was 0.886 (0.880; 0.894), while specificity remained low: in 93% of cases (55/59), specificity was ≤ 0.5. Among all the 106 reviewed studies, only 7 studies included oncologic patients and were included in the final analysis for C-CT performances. The percentage of patients with cancer in these studies was 0.3% (34/11352 patients), lower than the global prevalence of cancer. Among all these studies, only 1 (0.9%, 1/106) reported performance specifically in a cohort of cancer patients, but it however only reported true positives.</p></div><div><h3>Discussion</h3><p>There is a concerning lack of COVID-19 studies involving oncologic patients, showing there is a real need for further investigation and evaluation of the performance of the different medical imaging modalities in this specific patient population.</p></div>","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"1 ","pages":"Article 100004"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8970534/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"9963336","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}

{"title":"Quantitative and qualitative evaluation of liver metastases with intraprocedural cone beam CT prior to transarterial radioembolization as a predictor of treatment response","authors":"Florian Messmer MD ,&nbsp;Juliana Zgraggen ,&nbsp;Adrian Kobe MD ,&nbsp;Lyubov Chaykovska MD ,&nbsp;Gilbert Puippe MD ,&nbsp;Caecilia S. Reiner MD ,&nbsp;Thomas Pfammatter MD","doi":"10.1016/j.redii.2022.100005","DOIUrl":"10.1016/j.redii.2022.100005","url":null,"abstract":"<div><h3>Purpose</h3><p>To investigate, by quantitative and qualitative enhancement measurements, the correlation between tumor enhancement on cone beam computed tomography (CBCT) images and treatment response at 6 months in patients undergoing transarterial radioembolization (TARE) for liver metastases.</p></div><div><h3>Materials and Methods</h3><p>36 patients (56% male; median age 62.5 years) with 104 metastases were retrospectively included. Quantitative and qualitative enhancement of liver metastases were evaluated on CBCT images before TARE. Quantitative analysis consisted of lesion enhancement measurements (ROI HU lesion – ROI HU relative to inferior vena cava). Qualitative analysis consisted of subjective enhancement pattern analysis (diffuse, sparse, rim-like or non-enhancing). Morphologic tumor response was evaluated according to RECIST 1.1 criteria on follow-up CT or MR imaging.</p></div><div><h3>Results</h3><p>At a mean follow up of 6.5 ± 3.7 months, progressive disease (PD) was found in 4 patients, partial response (PR) in 11 and stable disease (SD) in 21. Relative lesion enhancement was significantly different between these groups (-37.5±154.2 HU vs. 103.8±93.4 vs. 181±144 HU in PD vs. SD vs. PR group, respectively; p&lt;0.01). ROC analysis of relative lesion enhancement to predict progressive disease showed an area under the curve of 0.86 (p&lt;0.01). For qualitative lesion enhancement analysis, no difference between groups was found.</p></div><div><h3>Conclusion</h3><p>Quantitative enhancement measurements derived from intraprocedural contrast enhanced CBCT may identify responders to TARE in patients with liver metastases.</p></div>","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"1 ","pages":"Article 100005"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2772652522000059/pdfft?md5=22d0d2b8eddf9bfdc73351bc0e2cf2ee&pid=1-s2.0-S2772652522000059-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47726773","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}

{"title":"Value and prognostic impact of a deep learning segmentation model of COVID-19 lung lesions on low-dose chest CT","authors":"Axel Bartoli MD ,&nbsp;Joris Fournel ,&nbsp;Arnaud Maurin MD ,&nbsp;Baptiste Marchi MD ,&nbsp;Paul Habert MD ,&nbsp;Maxime Castelli MD ,&nbsp;Jean-Yves Gaubert MD ,&nbsp;Sebastien Cortaredona MD ,&nbsp;Jean-Christophe Lagier MD, PhD ,&nbsp;Matthieu Million MD, PhD ,&nbsp;Didier Raoult MD, PhD ,&nbsp;Badih Ghattas MCU ,&nbsp;Alexis Jacquier MD, PhD","doi":"10.1016/j.redii.2022.100003","DOIUrl":"10.1016/j.redii.2022.100003","url":null,"abstract":"<div><h3>Objectives</h3><p>1) To develop a deep learning (DL) pipeline allowing quantification of COVID-19 pulmonary lesions on low-dose computed tomography (LDCT). 2) To assess the prognostic value of DL-driven lesion quantification.</p></div><div><h3>Methods</h3><p>This monocentric retrospective study included training and test datasets taken from 144 and 30 patients, respectively. The reference was the manual segmentation of 3 labels: normal lung, ground-glass opacity(GGO) and consolidation(Cons). Model performance was evaluated with technical metrics, disease volume and extent. Intra- and interobserver agreement were recorded. The prognostic value of DL-driven disease extent was assessed in 1621 distinct patients using C-statistics. The end point was a combined outcome defined as death, hospitalization&gt;10 days, intensive care unit hospitalization or oxygen therapy.</p></div><div><h3>Results</h3><p>The Dice coefficients for lesion (GGO+Cons) segmentations were 0.75±0.08, exceeding the values for human interobserver (0.70±0.08; 0.70±0.10) and intraobserver measures (0.72±0.09). DL-driven lesion quantification had a stronger correlation with the reference than inter- or intraobserver measures. After stepwise selection and adjustment for clinical characteristics, quantification significantly increased the prognostic accuracy of the model (0.82 vs. 0.90; <em>p</em>&lt;0.0001).</p></div><div><h3>Conclusions</h3><p>A DL-driven model can provide reproducible and accurate segmentation of COVID-19 lesions on LDCT. Automatic lesion quantification has independent prognostic value for the identification of high-risk patients.</p></div>","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"1 ","pages":"Article 100003"},"PeriodicalIF":0.0,"publicationDate":"2022-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8939894/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"9909529","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}