Layer-selective deep representation to improve esophageal cancer classification.

Even though artificial intelligence and machine learning have demonstrated remarkable performances in medical image computing, their accountability and transparency level must be improved to transfer this success into clinical practice. The reliability of machine learning decisions must be explained and interpreted, especially for supporting the medical diagnosis. For this task, the deep learning techniques' black-box nature must somehow be lightened up to clarify its promising results. Hence, we aim to investigate the impact of the ResNet-50 deep convolutional design for Barrett's esophagus and adenocarcinoma classification. For such a task, and aiming at proposing a two-step learning technique, the output of each convolutional layer that composes the ResNet-50 architecture was trained and classified for further definition of layers that would provide more impact in the architecture. We showed that local information and high-dimensional features are essential to improve the classification for our task. Besides, we observed a significant improvement when the most discriminative layers expressed more impact in the training and classification of ResNet-50 for Barrett's esophagus and adenocarcinoma classification, demonstrating that both human knowledge and computational processing may influence the correct learning of such a problem.

Error-Correcting Mean-Teacher: Corrections instead of consistency-targets applied to semi-supervised medical image segmentation.

Semantic segmentation is an essential task in medical imaging research. Many powerful deep-learning-based approaches can be employed for this problem, but they are dependent on the availability of an expansive labeled dataset. In this work, we augment such supervised segmentation models to be suitable for learning from unlabeled data. Our semi-supervised approach, termed Error-Correcting Mean-Teacher, uses an exponential moving average model like the original Mean Teacher but introduces our new paradigm of error correction. The original segmentation network is augmented to handle this secondary correction task. Both tasks build upon the core feature extraction layers of the model. For the correction task, features detected in the input image are fused with features detected in the predicted segmentation and further processed with task-specific decoder layers. The combination of image and segmentation features allows the model to correct present mistakes in the given input pair. The correction task is trained jointly on the labeled data. On unlabeled data, the exponential moving average of the original network corrects the student's prediction. The combined outputs of the students' prediction with the teachers' correction form the basis for the semi-supervised update. We evaluate our method with the 2017 and 2018 Robotic Scene Segmentation data, the ISIC 2017 and the BraTS 2020 Challenges, a proprietary Endoscopic Submucosal Dissection dataset, Cityscapes, and Pascal VOC 2012. Additionally, we analyze the impact of the individual components and examine the behavior when the amount of labeled data varies, with experiments performed on two distinct segmentation architectures. Our method shows improvements in terms of the mean Intersection over Union over the supervised baseline and competing methods. Code is available at https://github.com/CloneRob/ECMT.

Vessel and tissue recognition during third-space endoscopy using a deep learning algorithm.

In this study, we aimed to develop an artificial intelligence clinical decision support solution to mitigate operator-dependent limitations during complex endoscopic procedures such as endoscopic submucosal dissection and peroral endoscopic myotomy, for example, bleeding and perforation. A DeepLabv3-based model was trained to delineate vessels, tissue structures and instruments on endoscopic still images from such procedures. The mean cross-validated Intersection over Union and Dice Score were 63% and 76%, respectively. Applied to standardised video clips from third-space endoscopic procedures, the algorithm showed a mean vessel detection rate of 85% with a false-positive rate of 0.75/min. These performance statistics suggest a potential clinical benefit for procedure safety, time and also training.

Convolutional Neural Networks for the evaluation of cancer in Barrett's esophagus: Explainable AI to lighten up the black-box.

Even though artificial intelligence and machine learning have demonstrated remarkable performances in medical image computing, their level of accountability and transparency must be provided in such evaluations. The reliability related to machine learning predictions must be explained and interpreted, especially if diagnosis support is addressed. For this task, the black-box nature of deep learning techniques must be lightened up to transfer its promising results into clinical practice. Hence, we aim to investigate the use of explainable artificial intelligence techniques to quantitatively highlight discriminative regions during the classification of early-cancerous tissues in Barrett's esophagus-diagnosed patients. Four Convolutional Neural Network models (AlexNet, SqueezeNet, ResNet50, and VGG16) were analyzed using five different interpretation techniques (saliency, guided backpropagation, integrated gradients, input × gradients, and DeepLIFT) to compare their agreement with experts' previous annotations of cancerous tissue. We could show that saliency attributes match best with the manual experts' delineations. Moreover, there is moderate to high correlation between the sensitivity of a model and the human-and-computer agreement. The results also lightened that the higher the model's sensitivity, the stronger the correlation of human and computational segmentation agreement. We observed a relevant relation between computational learning and experts' insights, demonstrating how human knowledge may influence the correct computational learning.

Endoscopic prediction of submucosal invasion in Barrett's cancer with the use of artificial intelligence: a pilot study.

BACKGROUND: The accurate differentiation between T1a and T1b Barrett's-related cancer has both therapeutic and prognostic implications but is challenging even for experienced physicians. We trained an artificial intelligence (AI) system on the basis of deep artificial neural networks (deep learning) to differentiate between T1a and T1b Barrett's cancer on white-light images. METHODS: Endoscopic images from three tertiary care centers in Germany were collected retrospectively. A deep learning system was trained and tested using the principles of cross validation. A total of 230 white-light endoscopic images (108 T1a and 122 T1b) were evaluated using the AI system. For comparison, the images were also classified by experts specialized in endoscopic diagnosis and treatment of Barrett's cancer. RESULTS: The sensitivity, specificity, F1 score, and accuracy of the AI system in the differentiation between T1a and T1b cancer lesions was 0.77, 0.64, 0.74, and 0.71, respectively. There was no statistically significant difference between the performance of the AI system and that of experts, who showed sensitivity, specificity, F1, and accuracy of 0.63, 0.78, 0.67, and 0.70, respectively. CONCLUSION: This pilot study demonstrates the first multicenter application of an AI-based system in the prediction of submucosal invasion in endoscopic images of Barrett's cancer. AI scored equally to international experts in the field, but more work is necessary to improve the system and apply it to video sequences and real-life settings. Nevertheless, the correct prediction of submucosal invasion in Barrett's cancer remains challenging for both experts and AI.

Assisting Barrett's esophagus identification using endoscopic data augmentation based on Generative Adversarial Networks.

Barrett's esophagus figured a swift rise in the number of cases in the past years. Although traditional diagnosis methods offered a vital role in early-stage treatment, they are generally time- and resource-consuming. In this context, computer-aided approaches for automatic diagnosis emerged in the literature since early detection is intrinsically related to remission probabilities. However, they still suffer from drawbacks because of the lack of available data for machine learning purposes, thus implying reduced recognition rates. This work introduces Generative Adversarial Networks to generate high-quality endoscopic images, thereby identifying Barrett's esophagus and adenocarcinoma more precisely. Further, Convolution Neural Networks are used for feature extraction and classification purposes. The proposed approach is validated over two datasets of endoscopic images, with the experiments conducted over the full and patch-split images. The application of Deep Convolutional Generative Adversarial Networks for the data augmentation step and LeNet-5 and AlexNet for the classification step allowed us to validate the proposed methodology over an extensive set of datasets (based on original and augmented sets), reaching results of 90% of accuracy for the patch-based approach and 85% for the image-based approach. Both results are based on augmented datasets and are statistically different from the ones obtained in the original datasets of the same kind. Moreover, the impact of data augmentation was evaluated in the context of image description and classification, and the results obtained using synthetic images outperformed the ones over the original datasets, as well as other recent approaches from the literature. Such results suggest promising insights related to the importance of proper data for the accurate classification concerning computer-assisted Barrett's esophagus and adenocarcinoma detection.

A technical review of artificial intelligence as applied to gastrointestinal endoscopy: clarifying the terminology.

Background and aim The growing number of publications on the application of artificial intelligence (AI) in medicine underlines the enormous importance and potential of this emerging field of research. In gastrointestinal endoscopy, AI has been applied to all segments of the gastrointestinal tract most importantly in the detection and characterization of colorectal polyps. However, AI research has been published also in the stomach and esophagus for both neoplastic and non-neoplastic disorders. The various technical as well as medical aspects of AI, however, remain confusing especially for non-expert physicians. This physician-engineer co-authored review explains the basic technical aspects of AI and provides a comprehensive overview of recent publications on AI in gastrointestinal endoscopy. Finally, a basic insight is offered into understanding publications on AI in gastrointestinal endoscopy.

Automatic frontal sinus recognition in computed tomography images for person identification.

In many cases of person identification the use of biometric features obtained from the hard tissues of the human body, such as teeth and bones, may be the only option. This paper presents a new method of person identification based on frontal sinus features, extracted from computed tomography (CT) images of the skull. In this method, the frontal sinus is automatically segmented in the CT image using an algorithm developed in this work. Next, shape features are extracted from both hemispheres of the segmented frontal sinus by using BAS (Beam Angle Statistics) method. Finally, L2 distance is used in order to recognize the frontal sinus and identify the person. The novel frontal sinus recognition method obtained 77.25% of identification accuracy when applied on a dataset composed of 310 CT images obtained from 31 people, and the automatic frontal sinus segmentation in CT images obtained a mean Cohen Kappa coefficient equal to 0.8852 when compared to the ground truth (manual segmentation).

A survey on Barrett's esophagus analysis using machine learning.

This work presents a systematic review concerning recent studies and technologies of machine learning for Barrett's esophagus (BE) diagnosis and treatment. The use of artificial intelligence is a brand new and promising way to evaluate such disease. We compile some works published at some well-established databases, such as Science Direct, IEEEXplore, PubMed, Plos One, Multidisciplinary Digital Publishing Institute (MDPI), Association for Computing Machinery (ACM), Springer, and Hindawi Publishing Corporation. Each selected work has been analyzed to present its objective, methodology, and results. The BE progression to dysplasia or adenocarcinoma shows a complex pattern to be detected during endoscopic surveillance. Therefore, it is valuable to assist its diagnosis and automatic identification using computer analysis. The evaluation of the BE dysplasia can be performed through manual or automated segmentation through machine learning techniques. Finally, in this survey, we reviewed recent studies focused on the automatic detection of the neoplastic region for classification purposes using machine learning methods.

Activation of Latent Courtship Circuitry in the Brain of Drosophila Females Induces Male-like Behaviors.

Courtship in Drosophila melanogaster offers a powerful experimental paradigm for the study of innate sexually dimorphic behaviors [1, 2]. Fruit fly males exhibit an elaborate courtship display toward a potential mate [1, 2]. Females never actively court males, but their response to the male's display determines whether mating will actually occur. Sex-specific behaviors are hardwired into the nervous system via the actions of the sex determination genes doublesex (dsx) and fruitless (fru) [1]. Activation of male-specific dsx/fru(+) P1 neurons in the brain initiates the male's courtship display [3, 4], suggesting that neurons unique to males trigger this sex-specific behavior. In females, dsx(+) neurons play a pivotal role in sexual receptivity and post-mating behaviors [1, 2, 5-9]. Yet it is still unclear how dsx(+) neurons and dimorphisms in these circuits give rise to the different behaviors displayed by males and females. Here, we manipulated the function of dsx(+) neurons in the female brain to investigate higher-order neurons that drive female behaviors. Surprisingly, we found that activation of female dsx(+) neurons in the brain induces females to behave like males by promoting male-typical courtship behaviors. Activated females display courtship toward conspecific males or females, as well other Drosophila species. We uncovered specific dsx(+) neurons critical for driving male courtship and identified pheromones that trigger such behaviors in activated females. While male courtship behavior was thought to arise from male-specific central neurons, our study shows that the female brain is equipped with latent courtship circuitry capable of inducing this male-specific behavioral program.

Structural basis for both pro- and anti-inflammatory response induced by mannose-specific legume lectin from Cymbosema roseum.

Legume lectins, despite high sequence homology, express diverse biological activities that vary in potency and efficacy. In studies reported here, the mannose-specific lectin from Cymbosema roseum (CRLI), which binds N-glycoproteins, shows both pro-inflammatory effects when administered by local injection and anti-inflammatory effects when by systemic injection. Protein sequencing was obtained by Tandem Mass Spectrometry and the crystal structure was solved by X-ray crystallography using a Synchrotron radiation source. Molecular replacement and refinement were performed using CCP4 and the carbohydrate binding properties were described by affinity assays and computational docking. Biological assays were performed in order to evaluate the lectin edematogenic activity. The crystal structure of CRLI was established to a 1.8Å resolution in order to determine a structural basis for these differing activities. The structure of CRLI is closely homologous to those of other legume lectins at the monomer level and assembles into tetramers as do many of its homologues. The CRLI carbohydrate binding site was predicted by docking with a specific inhibitory trisaccharide. CRLI possesses a hydrophobic pocket for the binding of α-aminobutyric acid and that pocket is occupied in this structure as are the binding sites for calcium and manganese cations characteristic of legume lectins. CRLI route-dependent effects for acute inflammation are related to its carbohydrate binding domain (due to inhibition caused by the presence of α-methyl-mannoside), and are based on comparative analysis with ConA crystal structure. This may be due to carbohydrate binding site design, which differs at Tyr12 and Glu205 position.

Purification, characterization, and preliminary X-ray diffraction analysis of a lactose-specific lectin from Cymbosema roseum seeds.

The unique carbohydrate-binding property of lectins makes them invaluable tools in biomedical research. Here, we report the purification, partial primary structure, carbohydrate affinity characterization, crystallization, and preliminary X-ray diffraction analysis of a lactose-specific lectin from Cymbosema roseum seeds (CRLII). Isolation and purification of CRLII was performed by a single step using a Sepharose-4B-lactose affinity chromatography column. The carbohydrate affinity characterization was carried using assays for hemagglutination activity and inhibition. CRLII showed hemagglutinating activity toward rabbit erythrocytes. O-glycoproteins from mucine mucopolysaccharides showed the most potent inhibition capacity at a minimum concentration of 1.2 microg mL(-1). Protein sequencing by mass spectrometry was obtained by the digestion of CRLII with trypsin, Glu-C, and AspN. CRLII partial protein sequence exhibits 46% similarity with the ConA-like alpha chain precursor. Suitable protein crystals were obtained with the hanging-drop vapor-diffusion method with 8% ethylene glycol, 0.1 M Tris-HCl pH 8.5, and 11% PEG 8,000. The monoclinic crystals belong to space group P2(1) with unit cell parameters a = 49.4, b = 89.6, and c = 100.8 A.

Morfoanatomia dos órgãos vegetativos de Piper hispidum Sw. (Piperaceae) / Morphology and anatomy of the vegetative organs of Piper hispidum Sw. (Piperaceae)

O presente trabalho teve por objetivo o estudo morfoanatômico dos órgãos vegetativos de Piper hispidum, visando a estabelecer características marcantes para a sua identificação e auxiliar estudos taxonômicos e farmacobotânicos. O material vegetal fresco e fixado foi estudado segundo as técnicas usuais de corte e coloração, incluindo análise em MEV. Piper hispidum é um arbusto com caule cilíndrico, nodoso, verde claro, com folhas alternas, ovadas, de cor verde-escura na face adaxial e verde claro na abaxial. Dentre as características anatômicas importantes para sua identificação destacam-se: parênquima cortical da raiz apresentando grupos de esclereídes. Córtex caulinar com faixas descontínuas de colênquima do tipo angular e tecido vascular organizado em dois círculos descontínuos de feixes colaterais. A folha é dorsiventral, hipoestomática, com estômatos tetracíticos. Hipoderme adaxial descontínua e abaxial frouxa com número variável de camadas; tricomas tectores e glandulares ocorrem nas duas faces. Epiderme uniestratificada e idioblastos oleíferos ocorrem em todos os órgãos.

The morphology and anatomy of the vegetative organs of Piper hispidum are described, detaching remarkable strutural aspects and contributing to taxonomical and pharmacobotanical studies. The material was studied according to the usual techniques, including SEM (Scaning Eletron Microscopy). Piper hispidum is a shrub with cylindrical and green stem, which has alternate leaves. The main anatomical characteristics that can be used in its identification are: root with sclereids on cortical parenchyma, stem cortex with discontinuous strands of angular collenchyma, and vascular tissue constituted by two discontinous circles of collateral vascular bundles. The leaf is dorsiventral and hypostomatic with tetracytic stomata. The hypodermis is discontinuous in adaxial face, loose in abaxial one and presents a variable number of layers. Uniseriate epidermis and oil idioblasts occur in all organs.

Purification, partial characterization and preliminary X-ray diffraction analysis of a mannose-specific lectin from Cymbosema roseum seeds.

A lectin from Cymbosema roseum seeds (CRL) was purified, characterized and crystallized. The best crystals grew in a month and were obtained by the vapour-diffusion method using a precipitant solution consisting of 0.1 M Tris-HCl pH 7.8, 8%(w/v) PEG 3350 and 0.2 M proline at a constant temperature of 293 K. A data set was collected to 1.77 A resolution at a synchrotron-radiation source. CRL crystals are orthorhombic, belonging to space group P2(1)2(1)2(1). Crystallographic refinement and full amino-acid sequence determination are in progress.

Morfo-anatomia do desenvolvimento do fruto de Ocotea puberula (Rich) nees e de Nectandra megapotamica (Spreng) mez (lauracea) / Morphology and anatomy of the fruit development of Ocotea puberla (Rich) nees and Nectandra megapotamica (Spreng) mez (Lauraceae)

El estudio del desarrollo morfo-anatómico de los frutos de Ocotea puberula (Rich.) Nees y de Nectandra megapotamica (Spreng.) Mez, se realizó en flores y frutos en diferentes etapas de desarrollo recolectados en matorrales secundarios, en Maringá, Estado de Paraná, Brasil. Los frutos son drupas, con epicarpio epidérmico, mesocarpio parenquimático y endocarpio esclerenquimático (macroesclereidas). El endocarpio se origina de la epidermis interna del ovario. Las semillas exalbuminosas se diferencian de óvulos anátropos, poseen testa y tegmen, variadamente comprimidos. En la región hilar, estos tegumentos presentan esclereidas dispuestas en filas radiales. El embrión es recto, con cotiledónes gruesos ricos en almidón y aceite. La plúmula y el eje hipocótilo-radicular están reducidos.

Population genetics in the state of Nuevo León, México: VII. Frequencies of ABO, Rh(D), MN, Kell and other genetic traits in patients with allergy

Con el propósito de contribuir al conocimiento de la composición genética de las poblaciones humanas en el Estado de Nuevo León y basándose en el hecho de que los análisis de las frecuencias relativas de diferentes características genéticas en grupos de personas que sufren de varios estados de enfermedad han ayudado para indicar el componente genético que, a un mayor o menor grado están presentes en la mayoría de las enfermedades humanas, fueron analizadas las frecuencias para los grupos sanguíneos ABO, Rh(D), MN y Jell, así como la habiliddad para gustar la feniltiocarbamida (FTC), la presencia de vellos en la falange media, uso de las manos y la preferencia para el cruzado de manos y brazos, en 350 pacientes que padecían asma, rinitis alérgica (fiebre de heno) y diferentes alergias de piel, y 172 personas testígos comparadas por edad y sexo. Pacientes con rinitis mostraron frecuencias más altas de los grupos sanguíneos M y K(-), sensibilidad positiva a la FTC, presencia de vello en la falange digital media, cruzado izquierdo de mano y brazo. Pacientes con asma tenían alta frecuencia de grupo sanguíneos Rh(+) y K(-) y personas con alergias de piel mostraron altos porcentajes de grupo sanguíneo M