Multivariate analysis strategies for processing ToF-SIMS images of biomaterials

Time-of-flight secondary ion mass spectrometry (ToF-SIMS) is a hyperspectral imaging technique. Each pixel in a two-dimensional ToF-SIMS image (or each voxel in a three-dimensional (3-D) ToF-SIMS image) contains a full mass spectrum. Thus, multivariate analysis methods are being increasingly used to process biomaterial ToF-SIMS images so the maximum amount of information can be extracted from the images. This study examines the use of principal component analysis (PCA) and maximum autocorrelation factors (MAF) on four different ToF-SIMS images. These images were selected because they represent significant challenges for biomedical ToF-SIMS image processing (topographical features, low count rates, surface contaminants, etc.). With PCA four different types of scaling methods (auto, root mean, filter, and shift variance scaling) were used. The effect of two preprocessing methods (normalization and mean centering) was also examined for both PCA and MAF. The more computational intense MAF provided the best results for all the images investigated in this study, doing the best job of reducing the number of variables required to describe the image, enhancing image contrast and recovering key spectral features. MAF was particularly good at identifying subtle features that were often lost in PCA and impossible to visualize in single peak images. However, the combination of PCA with either root mean or shift variance scaling provided similar results to MAF. Thus, these combinations offer promising alternatives to MAF for working with large data sets encountered in 3-D imaging. Also, the new method of filter scaling is promising for processing low count rate images with salt and pepper noise. Normalization proved an important tool for deconvoluting chemical effects from topographic and/or matrix effects. Mean centering aided in reducing the dimensionality of the data, but in one case resulted in a loss of information.     

医学图像深度学习处理方法的研究进展

由于医学图像数据爆炸式增长,传统依靠医生人工对医学图像进行分析诊断,不仅工作效率低下,工作量大,还容易误诊、漏诊。随着人工智能(artificial intelligence,AI)技术的发展与应用,机器学习(machine learning,ML),尤其是深度学习(deep learning,DL)在医学图像分析领域发挥着越来越重要的作用。本文对DL在医学图像自动分割和分类识别中的研究进展进行综述,为DL在解决医学图像分析诊断方面提供有益参考。     

Impact of tissue processing,archiving and enrichment techniques on DNA methylation yield in rectal carcinoma

Background

Formalin fixation, duration of tissue storage and tissue enrichment techniques can affect DNA methylation yield but these effects have not been quantitatively measured. The aim is to investigate the relative impact of these conditions on DNA methylation in rectal cancer.

Methods

10 rectal cancers with matched undissected fresh frozen tissues, laser capture microdissected (LCM) formalin-fixed paraffin-embedded (FFPE) tissues, manual macrodissected FFPE tissues, adjacent normal mucosa and stromal tissues were analysed for APC and LINE-1 methylation using bisulphite pyrosequencing.

Results

FFPE cancer tissues, which had been stored for at least 4 years showed similar APC and LINE-1 methylation changes to matched fresh frozen cancer tissues. Laser capture microdissection did not increase the degree of methylation detected compared to manual macrodissection. Analysis of stromal tissues showed that they had undergone significant methylation changes compared to adjacent macroscopically normal mucosa, but not to the same extent as cancer tissues.

Conclusion

Reliable DNA methylation results can be obtained from FFPE rectal cancer tissues, which have been in long-term storage. Because only minor differences in methylation between macrodissected and LCM cancer tissues were found, our results do not support the routine use of LCM to enrich for cancer cells for DNA methylation studies.     

Digital processing of high-energy photon beam images

The potentialities of digital megavoltage radiation field images for image quality improvement and evaluation of patient treatment setup errors are investigated. A treatment verification film is digitized with a densitometer and a treatment planning computer. The 256 X 256 picture matrix is displayed on a viewing console of a computer tomograph. The results of some processed pictures show an improved visibility of anatomical structures, which is of importance for comparison of the treatment beam images with the corresponding localization radiograph made on the simulator. In addition, automated techniques for comparison of irradiation and simulation setup are now possible.     

Medical feature based evaluation of structuring elements for morphological enhancement of ultrasonic images

This paper investigates the use of morphology-based nonlinear filters, and performs deterministic and statistical analysis of the linear combinations of the filters for the image quality enhancement of B-mode ultrasound images. The fact that the structuring element shape greatly influences the output of the filter, is one of the most important features of mathematical morphology. The present reported work comparatively evaluates the structuring elements for morphological liver image enhancement and verifies the hypothesis that the speckles visible in US images are short, slightly ‘banana-shaped’ white lines. Initially, five different liver images were morphologically filtered using 10 different structuring elements and then the filtered images were assessed quantitatively. Image quality parameters such as peak signal-to-noise ratio, mean square error and correlation coefficient have been used to evaluate the performance of the morphological filters with different structuring elements. To endorse the observation of the quantitative analysis, the filtered images were then evaluated qualitatively, based on the image features looked into by the medical fraternity. The evaluation parameters have been taken on the basis of the suggestions made by a group of radiologists. The results of the processed images were then evaluated by a different group of radiologists. A multi-point rank order method has been used to identify small differences or trends in observation. The subjective analysis by radiologists indicates that morphological filter using line shaped structuring element with length 2 performs better than the other structuring elements. These observations were found to be in line with the observations of quantitative analysis.     

A non-linear morphometric feature selection approach for breast tumor contour from ultrasonic images

Ultrasound breast images have been used to improve diagnostics and decrease the number of unneeded biopsies. Malignant breast tumors tend to present irregular and blurred contours while benign ones are usually round, smooth and well-defined. Accordingly, investigating the tumor contour may help in establishing diagnosis. Herein, Mutual Information and Linear Discriminant Analysis were implemented to rank morphometric features in discriminating breast tumors in ultrasound images. Seven features were extracted from Convex Polygon and the Normalized Radial Length techniques. By applying a Mutual Information based approach, it was possible to identity features with possibly non-linear contributions to the outcome.     

医学CT图像的几种后处理技术——VB6编程对CT影像的几种后处理

本文说明后处理的原理，并用VB6对五种主要的CT后处理技术编程，再应用于数字图形和医学CT影像，表现CT影像经后处理的特殊效果。     

Selective medical image compression techniques for telemedical and archiving applications

Selective Image Compression (SeLIC) is a compression technique where explicitly defined regions of interest (RoI) are compressed in a lossless way whereas image regions containing unimportant information are compressed in a lossy manner. Such techniques are of great interest in telemedicine or medical imaging applications with large storage requirements. In this paper we introduce and compare techniques with different functionalities. Moreover, we investigate the impact of using wavelet transforms and JPEG as underlying lossy compression algorithm.     

Planning for a multi-imaging center picture archiving and communications system

The goal of the project was to create a method by which an in-house digital teaching file could be constructed that was simple, inexpensive, independent of hypertext markup language (HTML) restrictions, and appears identical on multiple platforms. To accomplish this, Microsoft PowerPoint and Adobe Acrobat were used in succession to assemble digital teaching files in the Acrobat portable document file format. They were then verified to appear identically on computers running Windows, Macintosh Operating Systems (OS), and the Silicon Graphics Unix-based OS as either a free-standing file using Acrobat Reader software or from within a browser window using the Acrobat browser plug-in. This latter display method yields a file viewed through a browser window, yet remains independent of underlying HTML restrictions, which may confer an advantage over simple HTML teaching file construction. Thus, a hybrid of HTML-distributed Adobe Acrobat generated WWW documents may be a viable alternative for digital teaching file construction and distribution.     

Image processing techniques for noise removal, enhancement and segmentation of cartilage OCT images

Osteoarthritis, whose hallmark is the progressive loss of joint cartilage, is a major cause of morbidity worldwide. Recently, optical coherence tomography (OCT) has demonstrated considerable promise for the assessment of articular cartilage. Among the most important parameters to be assessed is cartilage width. However, detection of the bone cartilage interface is critical for the assessment of cartilage width. At present, the quantitative evaluations of cartilage thickness are being done using manual tracing of cartilage-bone borders. Since data is being obtained near video rate with OCT, automated identification of the bone-cartilage interface is critical. In order to automate the process of boundary detection on OCT images, there is a need for developing new image processing techniques. In this paper we describe the image processing techniques for speckle removal, image enhancement and segmentation of cartilage OCT images. In particular, this paper focuses on rabbit cartilage since this is an important animal model for testing both chondroprotective agents and cartilage repair techniques. In this study, a variety of techniques were examined. Ultimately, by combining an adaptive filtering technique with edge detection (vertical gradient, Sobel edge detection), cartilage edges can be detected. The procedure requires several steps and can be automated. Once the cartilage edges are outlined, the cartilage thickness can be measured.     

Digital processing of microradiographic and fluorochromic images from bone biopsy

Bone biopsy is an essential technique for the diagnosis of metabolic bone disease and the skeletal response to remedial therapeutic maneuvers. Among many approaches in analyzing the biopsy specimen, the imaging technique offers the greatest potential. The objective of this paper is to describe the methodologies developed for digital processing of microradiographic and fluorochromic images from bone biopsy. This procedure requires a microscope with conventional and ultraviolet illuminations, a vidicon camera with three color filters, a general purpose image processing system with three image memories, a digital computer with some specially tailored software programs. A good specimen preparation from bone biopsy is a prerequisite in assuring the success of the method. The methodology for processing the microradiographs is based on a completely developed procedure of black and white image processing. Production run is in progress. The methodology for processing the fluorochromic images is based on true color imaging, although the procedure is still preliminary, early results obtained so far are promising.     

Puncture-proof picture archiving and communication system

As we become increasingly dependent on our picture archiving and communications system (PACS) for the clinical practice of medicine, the demand for improved reliability becomes urgent. Borrowing principles from the discipline of Reliability Engineering, we have identified components of our system that constitute single points of failure and have endeavored to eliminate these through redundant components and manual work-around procedures. To assess the adequacy of our preparations, we have identified a set of plausible events that could interfere with the function of one or more of our PACS components. These events could be as simple as the loss of the network connection to a single component or as broad as the loss of our central data center. We have identified the need to continue to operate during adverse conditions, as well as the requirement to recover rapidly from major disruptions in service. This assessment led us to modify the physical locations of central PACS components within our physical plant. We are also taking advantage of actual disruptive events coincident with a major expansion of our facility to test our recovery procedures. Based on our recognition of the vital nature of our electronic images for patient care, we are now recording electronic images in two copies on disparate media. The image database is critical to both continued operations and recovery. Restoration of the database from periodic tape backups with a 24-hour cycle time may not support our clinical scenario: acquisition modalities have a limited local storage capacity, some of which will not contain the daily workload. Restoration of the database from the archived media is an exceedingly slow process, that will likely not meet our requirement to restore clinical operations without significant delay. Our PACS vendor is working on concurrent image databases that would be capable of nearly immediate switchover and recovery.