Search

In experiments with multiple quantitative measurements per subject, for example measurements on multiple lesions per patient, the additional measurements on the same patient provide limited additional information. Treating these measurements as independent observations will produce biased estimators for standard deviations and confidence intervals, and increases the risk of false positives in statistical tests. The problem can be remedied in a simple way by first taking the average of all observations of each specific patient, and then doing all further calculations only on the list of these patient means. A more sophisticated statistical modeling of the experiment, for example in a linear mixed model, is only required if (i) there is a large imbalance in the number of observations per patient or (ii) there is a specific interest in actually identifying the various sources of variation in the experiment. [ABSTRACT FROM AUTHOR]

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.ahj.2006.03.026 Byline: Michael Lindstaedt (a), Aydan Yazar (a), Alfried Germing (a), Markus K. Fritz (b), Tim Holland-Letz (c), Andreas Mugge (a), Waldemar Bojara (a) Abstract: Correct assessment of the significance of left main stem lesions is of pivotal importance to the patient with coronary artery disease. On the basis of clinical and angiographic information alone, this evaluation often cannot be done reliably. Limited data suggest that coronary pressure-derived fractional flow reserve (FFR) supports decision making in equivocal left main disease. Author Affiliation: (a) Medical Clinic II, University Hospital Bergmannsheil, Bochum, Germany (b) Division of Cardiothoracic Surgery, University Hospital Bergmannsheil, Bochum, Germany (c) Department of Medical Informatics, Biometry, and Epidemiology, Ruhr-University Bochum, Germany Article History: Received 14 August 2005; Accepted 20 March 2006

Summary Chromosomal instability is a hallmark of cancer and correlates with the presence of extra centrosomes, which originate from centriole overduplication. Overduplicated centrioles lead to the formation of centriole rosettes, which mature into supernumerary centrosomes in the subsequent cell cycle. While extra centrosomes promote chromosome missegregation by clustering into pseudo-bipolar spindles, the contribution of centriole rosettes to chromosome missegregation is unknown. We used multi-modal imaging of cells with conditional centriole overduplication to show that mitotic rosettes in bipolar spindles frequently harbor unequal centriole numbers, leading to biased chromosome capture that favors binding to the prominent pole. This results in chromosome missegregation and aneuploidy. Rosette mitoses lead to viable offspring and significantly contribute to progeny production. We further show that centrosome abnormalities in primary human malignancies frequently consist of centriole rosettes. As asymmetric centriole rosettes generate mitotic errors that can be propagated, rosette mitoses are sufficient to cause chromosome missegregation in cancer. [ABSTRACT FROM AUTHOR]

Respiratory tract infections are among the deadliest communicable diseases worldwide. Severe cases of viral lung infections are often associated with a cytokine storm and alternating platelet numbers. We report that hematopoietic stem and progenitor cells (HSPCs) sense a non-systemic influenza A virus (IAV) infection via inflammatory cytokines. Irrespective of antiviral treatment or vaccination, at a certain threshold of IAV titer in the lung, CD41-positive hematopoietic stem cells (HSCs) enter the cell cycle while endothelial protein C receptor-positive CD41-negative HSCs remain quiescent. Active CD41-positive HSCs represent the source of megakaryocytes, while their multi-lineage reconstitution potential is reduced. This emergency megakaryopoiesis is thrombopoietin independent and attenuated in IAV-infected interleukin-1 receptor-deficient mice. Newly produced platelets during IAV infection are immature and hyper-reactive. After viral clearance, HSC quiescence is re-established. Our study reveals that non-systemic viral respiratory infection has an acute impact on HSCs via inflammatory cytokines to counteract IAV-induced thrombocytopenia. [Display omitted] • IAV infection transiently activates CD41+ HSCs, while EPCR+ CD41− HSCs remain quiescent • CD41+ HSCs produce megakaryocytes in IAV-induced emergency megakaryopoiesis • Platelets generated by IAV-induced emergency megakaryopoiesis are pro-coagulant • IAV-induced emergency megakaryopoiesis is attenuated in IL1R1-deficient mice Rommel et al. show that non-systemic respiratory influenza infection activates CD41-expressing hematopoietic stem cells (HSCs) and promotes emergency megakaryopoiesis, thereby generating pro-coagulant and immature platelets that contribute to lung pathology. [ABSTRACT FROM AUTHOR]

Objective: Recent advances in perioperative care have led to a decrease in mortality of children with hypoplastic left heart syndrome undergoing the Norwood operation. This study aimed to evaluate the outcome of the Norwood operation in a single center over 12 years and to identify clinical and anatomic risk factors for adverse early and longer term outcome. Methods: Full data on all 157 patients treated between 1996 and 2007 were analyzed. Results: Thirty-day mortality of the Norwood operation decreased from 21% in the first 3 years to 2.5% in the last 3 years. The estimated exponentially weighted moving average of early mortality after 157 Norwood operations was 2.3%. Risk factors were an aberrant right subclavian artery, the use and duration of circulatory arrest, and the duration of total support time. The anatomic subgroup mitral stenosis/aortic atresia and female gender tended to show an increased early mortality. In the group of patients who required postoperative cardiopulmonary resuscitation, the ascending aorta was significantly smaller than in the remainder (3.03 ± 1.05 vs 3.63 ± 1.41 mm). Interstage mortality was 15% until the initiation of a home surveillance program in 2005, which has zeroed it so far. It was significantly higher in the mitral stenosis/aortic atresia subgroup and tended to be higher in patients who required cardiopulmonary resuscitation after the Norwood operation. The best actuarial survival was observed in the mitral atresia/aortic atresia subgroup. Conclusion: The Norwood operation can now be performed with low mortality. Patients with mitral stenosis/aortic atresia still constitute the most challenging subgroup. [Copyright &y& Elsevier]

Lymph node status is a prognostic marker and strongly influences therapeutic decisions in colorectal cancer (CRC). The objective of the study is to investigate whether image features extracted by a deep learning model from routine histological slides and/or clinical data can be used to predict CRC lymph node metastasis (LNM). Using histological whole slide images (WSIs) of primary tumours of 2431 patients in the DACHS cohort, we trained a convolutional neural network to predict LNM. In parallel, we used clinical data derived from the same cases in logistic regression analyses. Subsequently, the slide-based artificial intelligence predictor (SBAIP) score was included in the regression. WSIs and data from 582 patients of the TCGA cohort were used as the external test set. On the internal test set, the SBAIP achieved an area under receiver operating characteristic (AUROC) of 71.0%, the clinical classifier achieved an AUROC of 67.0% and a combination of the two classifiers yielded an improvement to 74.1%. Whereas the clinical classifier's performance remained stable on the TCGA set, performance of the SBAIP dropped to an AUROC of 61.2%. Performance of the clinical classifier depended strongly on the T stage. Deep learning–based image analysis may help predict LNM of patients with CRC using routine histological slides. Combination with clinical data such as T stage might be useful. Strategies to increase performance of the SBAIP on external images should be investigated. • We aimed to develop a new biomarker to predict lymph node status of patients with colorectal cancer (CRC). • Convolutional neural network–based image analysis and clinical data were used individually and in combination. • Models were trained and tested on haematoxylin and eosin slides of 3013 primary CRC tumours. • The combined model on the internal test set yielded the best results (area under receiver operating curve of 74.1%). • Studies should address generalisation of the image classifier to external data. [ABSTRACT FROM AUTHOR]

Blood-based protein biomarker signatures might be an alternative or supplement to existing methods for early detection of colorectal cancer (CRC) for population-based screening. The objective of this study was to derive a protein biomarker signature for early detection of CRC and its precursor advanced adenoma (AA). In a two-stage design, 270 protein markers were measured by liquid chromatography/multiple reaction monitoring/mass spectrometry in plasma samples of discovery and validation sets. In the discovery set consisting of 100 newly diagnosed CRC cases and 100 age- and sex-matched controls free of neoplasm at screening colonoscopy, the algorithms predicting the presence of early- or late-stage CRC were derived by Lasso regression and.632 + bootstrap. The prediction algorithms were then externally validated in an independent validation set consisting of participants of screening colonoscopy including 56 participants with CRC, 99 with AA and 99 controls without any colorectal neoplasms. Three different signatures for all-, early- and late-stage CRC consisting of five-, three- and eight-protein markers were obtained in the discovery set with areas under the curves (AUCs) after.632 + bootstrap adjustment of 0.85, 0.83 and 0.96, respectively. External validation in the representative screening population yielded AUCs of 0.79 (95% CI, 0.70–0.86), 0.79 (95% CI, 0.66–0.89) and 0.80 (95% CI, 0.70–0.89) for all-, early- and late-stage CRCs, respectively. The three-marker early-stage algorithm yielded an AUC of 0.65 (95% CI, 0.56–0.73) for detection of AA in the validation set. Although not yet competitive with available stool-based tests for CRC early detection, the identified proteins may contribute to the development of powerful blood-based tests for early detection of CRC and its precursors AAs. A three plasma protein biomarker signature derived in a clinical setting yielded an AUC of 0.83. When independently validated in a true screening setting (participants of screening colonoscopy) AUCs of 0.79 and 0.65 for distinguishing early stage colorectal cancer and advanced adenoma cases from controls free of neoplasms were obtained. These protein biomarkers may contribute to development of a blood based tests for population based screening. Image 1 • Participation rate is often low in endoscopy or stool-based colorectal cancer screening trials. • Blood based test could be an alternative or supplement to existing screening procedures. • With liquid chromatography-multiple reaction monitoring/mass spectrometry three stage specific signatures were identified. • 3-marker signature detected early-stage colorectal cancer and advanced adenomas with area under the curves of 0.79 and 0.65. • These proteins may serve as potential biomarkers for blood-based early detection of colorectal cancer and its precursors. [ABSTRACT FROM AUTHOR]

In recent studies, convolutional neural networks (CNNs) outperformed dermatologists in distinguishing dermoscopic images of melanoma and nevi. In these studies, dermatologists and artificial intelligence were considered as opponents. However, the combination of classifiers frequently yields superior results, both in machine learning and among humans. In this study, we investigated the potential benefit of combining human and artificial intelligence for skin cancer classification. Using 11,444 dermoscopic images, which were divided into five diagnostic categories, novel deep learning techniques were used to train a single CNN. Then, both 112 dermatologists of 13 German university hospitals and the trained CNN independently classified a set of 300 biopsy-verified skin lesions into those five classes. Taking into account the certainty of the decisions, the two independently determined diagnoses were combined to a new classifier with the help of a gradient boosting method. The primary end-point of the study was the correct classification of the images into five designated categories, whereas the secondary end-point was the correct classification of lesions as either benign or malignant (binary classification). Regarding the multiclass task, the combination of man and machine achieved an accuracy of 82.95%. This was 1.36% higher than the best of the two individual classifiers (81.59% achieved by the CNN). Owing to the class imbalance in the binary problem, sensitivity, but not accuracy, was examined and demonstrated to be superior (89%) to the best individual classifier (CNN with 86.1%). The specificity in the combined classifier decreased from 89.2% to 84%. However, at an equal sensitivity of 89%, the CNN achieved a specificity of only 81.5% Our findings indicate that the combination of human and artificial intelligence achieves superior results over the independent results of both of these systems. • This article describes the first experiment on combining human and artificial intelligence for the classification of images suspicious of skin cancer. • The combination achieved a superior accuracy of 82.95% (compared to 81.59%/42.94% achieved by artificial/human intelligence alone). • The combination of human and artificial intelligence indicates superiority over a separated approach. [ABSTRACT FROM AUTHOR]

Melanoma is the most dangerous type of skin cancer but is curable if detected early. Recent publications demonstrated that artificial intelligence is capable in classifying images of benign nevi and melanoma with dermatologist-level precision. However, a statistically significant improvement compared with dermatologist classification has not been reported to date. For this comparative study, 4204 biopsy-proven images of melanoma and nevi (1:1) were used for the training of a convolutional neural network (CNN). New techniques of deep learning were integrated. For the experiment, an additional 804 biopsy-proven dermoscopic images of melanoma and nevi (1:1) were randomly presented to dermatologists of nine German university hospitals, who evaluated the quality of each image and stated their recommended treatment (19,296 recommendations in total). Three McNemar's tests comparing the results of the CNN's test runs in terms of sensitivity, specificity and overall correctness were predefined as the main outcomes. The respective sensitivity and specificity of lesion classification by the dermatologists were 67.2% (95% confidence interval [CI]: 62.6%–71.7%) and 62.2% (95% CI: 57.6%–66.9%). In comparison, the trained CNN achieved a higher sensitivity of 82.3% (95% CI: 78.3%–85.7%) and a higher specificity of 77.9% (95% CI: 73.8%–81.8%). The three McNemar's tests in 2 × 2 tables all reached a significance level of p < 0.001. This significance level was sustained for both subgroups. For the first time, automated dermoscopic melanoma image classification was shown to be significantly superior to both junior and board-certified dermatologists (p < 0.001). • Recent publications demonstrated that deep learning is capable to classify images of benign nevi and melanoma with dermatologist-level precision. • A systematic outperformance of dermatologists was not demonstrated to date. • This study shows the first systematic (p < 0.001) outperformance of board-certified dermatologists in dermoscopic melanoma image classification. [ABSTRACT FROM AUTHOR]

The diagnosis of most cancers is made by a board-certified pathologist based on a tissue biopsy under the microscope. Recent research reveals a high discordance between individual pathologists. For melanoma, the literature reports 25–26% of discordance for classifying a benign nevus versus malignant melanoma. Deep learning was successfully implemented to enhance the precision of lung and breast cancer diagnoses. The aim of this study is to illustrate the potential of deep learning to assist human assessment for a histopathologic melanoma diagnosis. Six hundred ninety-five lesions were classified by an expert histopathologist in accordance with current guidelines (350 nevi and 345 melanomas). Only the haematoxylin and eosin stained (H&E) slides of these lesions were digitalised using a slide scanner and then randomly cropped. Five hundred ninety-five of the resulting images were used for the training of a convolutional neural network (CNN). The additional 100 H&E image sections were used to test the results of the CNN in comparison with the original class labels. The total discordance with the histopathologist was 18% for melanoma (95% confidence interval [CI]: 7.4–28.6%), 20% for nevi (95% CI: 8.9–31.1%) and 19% for the full set of images (95% CI: 11.3–26.7%). Even in the worst case, the discordance of the CNN was about the same compared with the discordance between human pathologists as reported in the literature. Despite the vastly reduced amount of data, time necessary for diagnosis and cost compared with the pathologist, our CNN archived on-par performance. Conclusively, CNNs indicate to be a valuable tool to assist human melanoma diagnoses. • A convolutional neural network (CNN) was trained with 595 histopathologic images of melanomas and nevi that were classified by an expert dermatohistopathologist. • The CNN was then tested with 100 additional images (melanoma/nevi = 1:1) and revealed a discordance of only 19% to the histopathologist. • Thus, even in the worst case, the discordance of the CNN is about the same compared with the discordance between human pathologists as reported in the literature (25–26%). [ABSTRACT FROM AUTHOR]

Recent studies have successfully demonstrated the use of deep-learning algorithms for dermatologist-level classification of suspicious lesions by the use of excessive proprietary image databases and limited numbers of dermatologists. For the first time, the performance of a deep-learning algorithm trained by open-source images exclusively is compared to a large number of dermatologists covering all levels within the clinical hierarchy. We used methods from enhanced deep learning to train a convolutional neural network (CNN) with 12,378 open-source dermoscopic images. We used 100 images to compare the performance of the CNN to that of the 157 dermatologists from 12 university hospitals in Germany. Outperformance of dermatologists by the deep neural network was measured in terms of sensitivity, specificity and receiver operating characteristics. The mean sensitivity and specificity achieved by the dermatologists with dermoscopic images was 74.1% (range 40.0%–100%) and 60% (range 21.3%–91.3%), respectively. At a mean sensitivity of 74.1%, the CNN exhibited a mean specificity of 86.5% (range 70.8%–91.3%). At a mean specificity of 60%, a mean sensitivity of 87.5% (range 80%–95%) was achieved by our algorithm. Among the dermatologists, the chief physicians showed the highest mean specificity of 69.2% at a mean sensitivity of 73.3%. With the same high specificity of 69.2%, the CNN had a mean sensitivity of 84.5%. A CNN trained by open-source images exclusively outperformed 136 of the 157 dermatologists and all the different levels of experience (from junior to chief physicians) in terms of average specificity and sensitivity. • A convolutional neural network (CNN) received enhanced training with 12,378 open-source dermoscopic images. • In a head-to-head comparison, the CNN outperformed 136 of 157 participating dermatologists. • The CNN was capable to outperform dermatologists of all hierarchical subgroups (from junior to chief physicians) in dermoscopic melanoma image classification. [ABSTRACT FROM AUTHOR]

Abstract: Background: Decisions for coronary revascularisation are frequently based on visual assessment of the severity of a stenosis. In patients with intermediate left main stem lesions clinical decision making based on FFR is safe and feasible. This study was performed to assess the accuracy of visual angiographic assessment of intermediate or equivocal left main coronary artery (LMCA) stenoses by experienced interventional cardiologists when taking fractional flow reserve (FFR) measurements as the gold standard. Methods: Fifty-one patients with intermediate (40–80% diameter stenosis by angiography) or equivocal LMCA disease were evaluated by FFR. Angiograms were then reviewed by 4 experienced interventionalists from different university hospitals blinded to FFR results. Lesions were visually assessed and their significance classified as ‘significant’, ‘not significant’, or ‘unsure’ if the observer was unable to make a decision regarding lesion significance based on the angiogram. Results: Results were compared with two different FFR cutoff values (<0.75 and ≤0.80) for hemodynamic significance. The 4 reviewers achieved correct lesion classification in no more than approximately 50% of cases each, regardless of FFR threshold. The interobserver agreement between two reviewers in excess of the agreement expected due to chance was outperformed on average by only 16%. Furthermore, interobserver variability was large resulting in unanimously correct lesion classification in only 29% of all cases. Conclusions: The functional significance of intermediate and equivocal LMCA stenoses should not be based solely on angiographic assessment even by experienced interventional cardiologists. [Copyright &y& Elsevier]