ABSTRACT
As an important basis for lesion determination and diagnosis, medical image segmentation has become one of the most important and hot research fields in the biomedical field, among which medical image segmentation algorithms based on full convolutional neural network and U-Net neural network have attracted more and more attention by researchers. At present, there are few reports on the application of medical image segmentation algorithms in the diagnosis of rectal cancer, and the accuracy of the segmentation results of rectal cancer is not high. In this paper, a convolutional network model of encoding and decoding combined with image clipping and pre-processing is proposed. On the basis of U-Net, this model replaced the traditional convolution block with the residual block, which effectively avoided the problem of gradient disappearance. In addition, the image enlargement method is also used to improve the generalization ability of the model. The test results on the data set provided by the "Teddy Cup" Data Mining Challenge showed that the residual block-based improved U-Net model proposed in this paper, combined with image clipping and preprocessing, could greatly improve the segmentation accuracy of rectal cancer, and the Dice coefficient obtained reached 0.97 on the verification set.
Subject(s)
Humans , Algorithms , Delayed Emergence from Anesthesia , Image Processing, Computer-Assisted , Rectal Neoplasms/diagnostic imaging , Tomography, X-Ray ComputedABSTRACT
Recently, the hepatotoxicity issue regarding to Psoraleae Fructus (PF) has attracted remarkable concerns, which highlights the urgent need to explore the toxicity attenuation method for PF. In this study, we proposed an alcohol soaking and water rinsing method for pre-processing PF based on the record in the classics - "Lei Gong Pao Zhi Lun", aiming to attenuate the potential hepatotoxicity of PF. The optimal pre-processing methods and parameters were investigated by U*12(108) uniform design coupled with 3D-cultured human-derived liver organoids model and high-content imaging. The results showed that there were significant variations among the hepatotoxicity intensities of different pre-processed PF products. Four factors, including the concentration of alcohol, the ratio of material and alcohol in alcohol soaking, the time of alcohol soaking and the times of water rinsing, were found as independent significant factors (all P<0.01). The optimal pre-process parameters were further predicted and verified as follows: the alcohol concentration is 80%, the times of alcohol soaking is 3, the ratio of alcohol and material of alcohol soaking is 3, the time for alcohol soaking is 30 h, the ratio of water and material of water rinsing is 2, the times of water rinsing is 3, the time water rinsing is 12 h and the time of steaming is 5 h. This research demonstrated that the alcohol soaking and water rinsing method can effectively reduce the potential hepatotoxicity of PF. This method provides a reference for reducing the risk of PF liver injury from the perspective of Chinese medicinal materials pre-processing.
ABSTRACT
Objective To explore suitable pre-processing methods for the TCM clinical data based on prospective study on insomnia treated by syndrome differentiation. Methods Based on the TCM shared clinical and research information platform and by using man-machine combination method, data cleaning rules, physician review, rule revision, procedural import and batch processing were used to conduct pre-processing for data in prospective study on insomniac treated by syndrome differentiation of 8 TCM doctors. Results Totally 27534 rules for symptoms data of individual treatment of insomnia were made and 1036 rules for diagnostic data, 842 rules for therapeutic ways, 540 rules for formula data, 3785 rules for data of Chinese materia medica. Conclusion Different kinds of terminology concepts were normalized at different levels, at the same time, characteristics of individualized treatment based on syndrome differentiation were reserved. Appropriate pre-processing methods can be used in the reaserch of individualization and standardization of TCM syndrome differentiation clinical data and can provide support for data mining.
ABSTRACT
Objective To investigate the influence of sputum specimens pre-processing on the detection result and its clinical sig-nificance .Methods The sputum samples of 124 inpatients in this hospital from February 2016 to April 2016 were selected .The sputum color ,character ,Gram staining microcopic examinationof sputum smear were observed by adopting the naked-eye observa-tion .Its influence on the sputum specimen isolation results was analyzed .Results Among 124 sputum specimens ,90 cases were qualified ,while 34cases were unqualified ;the qualification rate of colorlesssputum specimens was 30% ,which was lower than 94 .7% of yellow sputum specimens ,100% of rusty sputum specimens ,100% of red sputum specimens and 87 .5% of white sputum specimens ;the qualification rate of frothy or watery sputum specimens was 22 .2% ,which was lower than that of other character sputum specimens ,such as purulent sputum specimens (96 .2% ) ,bloody sputum specimens (100% )and mucous sputum specimens (83 .3% );the cultured results of 34 unqualified specimens were the oral and throat normal flora ,which all had the contaminating bacterial growth ;among 90 qualified specimens ,58 specimens showed pure culture or advantage growth ,and 32 specimens showed contaminating bacterial growth .Conclusion The pre-processing of sputum specimens is an efficient solution for improving the accu-racy of sputum specimens detection results ,and has the important clinical significance .
ABSTRACT
En la última década, la Imagenología por Espectrometría de Masas MALDI (MALDI-MSI) ha demostrado su potencial en el campo de la proteómica. Debido a la alta dimensionalidad de los conjuntos de datos obtenidos en los experimentos de MADI-MSI y las variabilidades inherentes al proceso de adquisición, presentes en los mismos, se hace necesario llevar a cabo una etapa de pre-procesamiento, que reduzca estas distorsiones. La presente investigación propone una metodología de procesamiento de datos de MALDI-MSI sustentada en un conjunto de aplicaciones desarrolladas en MatLab, la Biblioteca Qt4, así como la herramienta de visualización DataCube Explorer. Entre los resultados se pueden destacar la obtención de cambios en las intensidades de los píxeles de las imágenes reconstruidas después de la introducción de ruido, así como el incremento de la Relación Señal-Ruido después de someter los espectros a los métodos de filtrado de Kaiser, Savitzky-Golay y Promedio deslizante, destacándose Kaiser sobre los demás, lo que puede traducirse como una disminución de los niveles de distorsión en los espectros de cada pixel. Se realizó la reconstrucción satisfactoria de la imagen patrón y su visualización con la herramienta DataCube Explorer(AU)
At last decade, MALDI Mass Spectrometry Imaging (MALDI-MSI) has demostrated being a powerful tool in proteomics field. Because of high dimensionality of the acquired datasets in MALDI-MSI experiments and the inherent variabilities to the acquisition process, present in them, it´s necessary carry out a pre-processing stage, which reduces these distortions. This paper proposes a processing methodology of data MALDI-MSI supported in applications developed in MATLAB, Qt4 Library, as well as the visualization tool Datacube Explorer. Among the results obtained, it can be highlighted the changes in the intensities of the pixels of the reconstructed images after introducing noise, as well as the increasing of Signal-Noise Ratio after applying the denoising methods Kaiser, Savitzky-Golay and "sliding average", highlighting Kaiser over the other techniques, which can be interpreted as a decreasing of the distortion levels in each pixel´s spectrum. The reconstruction of example image and its visualization with Datacube Explorer tool were satisfactory(AU)