Automated Thresholding Algorithm Research Articles

Comparison of Automated Thresholding Algorithms in Optical Coherence Tomography Angiography Image Analysis.

(1) Background: Calculation of vessel density in optical coherence tomography angiography (OCTA) images with thresholding algorithms varies in clinical routine. The ability to discriminate healthy from diseased eyes based on perfusion of the posterior pole is critical and may depend on the algorithm applied. This study assessed comparability, reliability, and ability in the discrimination of commonly used automated thresholding algorithms. (2) Methods: Vessel density in full retina and choriocapillaris slabs were calculated with five previously published automated thresholding algorithms (Default, Huang, ISODATA, Mean, and Otsu) for healthy and diseased eyes. The algorithms were investigated with LD-F2-analysis for intra-algorithm reliability, agreement, and the ability to discriminate between physiological and pathological conditions. (3) Results: LD-F2-analyses revealed significant differences in estimated vessel densities for the algorithms (p < 0.001). For full retina and choriocapillaris slabs, intra-algorithm values range from excellent to poor, depending on the applied algorithm; the inter-algorithm agreement was low. Discrimination was good for the full retina slabs, but poor when applied to the choriocapillaris slabs. The Mean algorithm demonstrated an overall good performance. (4) Conclusions: Automated threshold algorithms are not interchangeable. The ability for discrimination depends on the analyzed layer. Concerning the full retina slab, all of the five evaluated automated algorithms had an overall good ability for discrimination. When analyzing the choriocapillaris, it might be useful to consider another algorithm.

Machine Learning Approach for Automatic Detection of Alzheimers Disease using Resting State fMRI

Alzheimer’s Disease (AD) is a neurological disease that affects memory and the livelihood of the people that are diagnosed with it. Efficient automated techniques for early diagnosis of AD is very important because early diagnosis is used to prevent a patient from death. In this work, we present a novel computer-aided diagnosis (CAD) techniques using machine learning algorithms for the early diagnosis of AD. The input resting state fMRI(rsfMRI) images are taken from Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. The input image is pre-processed using Discrete Wavelet Transform(DWT). Automated thresholding algorithm is used to segment the image. Then, the segmented resting state fMRI images are used to extract useful and informative features. The best features are selected by Fisher’s code feature selection algorithm. Finally, an automated Image classification step is performed using machine learning algorithms Support Vector Machine(SVM), Decision Tree , Random Forest and Multi-Layer Perceptron algorithms to distinguish between normal patients and AD patients.

Contrast-enhanced repacked soil cores as a proxy for soil organic matter spatial arrangement

Soil organic matter (SOM) plays a key role in soil structure formation, although the bidirectional relationship between SOM and the soil pore network is complex and needs further investigation. Despite great advances provided by X-ray computed microtomography (µCT), it has only been used in a few studies to investigate the organic matter 3D-arrangement within the soil matrix. Results are based on the X-ray linear attenuation coefficient (α), and mixtures of organic and mineral soil fractions could imply overlapping of information that makes any segmentation procedure difficult. The aim of this study was to visualise, segment, and quantify the particulate organic matter fraction (POM) within the soil matrix through X-ray µCT in combination with contrast agents (phosphomolybdic acid and silver nitrate). Two series of repacked soil cores, ‘dry’ and ‘wet’, were scanned through X-ray µCT at a 7-µm resolution. Different segmentation approaches were tested to separate POM from other soil phases: manual, global, and local thresholding methods. Reported algorithms were also compared with a supervised grey value-based (GV) approach followed by morphological operations. Results showed contrast agents increased α of POM, simplifying its identification and the following segmentation on dry cores. The POM was discriminated from the mineral fraction and its content correctly estimated. This was particularly accurate when applying manual thresholding or GV approach with respect to indicator kriging, suggesting that operator-based ability to set threshold level is still the best solution for accurate POM segmentation. Beyond single-phase accounting, different thresholding algorithms and morphological operations also affected POM morphological characteristics. In particular, the simpler was an object shape, the easier was its segmentation. Improvements are thus required to increase the efficiency of automated thresholding algorithms. Moreover, wet cores were exposed to washing-out phenomena that compromised any digital image processing and further POM quantification, implying that more effort should be made to find other suitable staining agents.

Morphological analysis of pore size and connectivity in a thick mixed-culture biofilm.

Morphological parameters are commonly used to predict transport and metabolic kinetics in biofilms. Yet, quantification of biofilm morphology remains challenging because of imaging technology limitations and lack of robust analytical approaches. We present a novel set of imaging and image analysis techniques to estimate internal porosity, pore size distributions, and pore network connectivity to a depth of 1 mm at a resolution of 10 µm in a biofilm exhibiting both heterotrophic and nitrifying activities. Optical coherence tomography (OCT) scans revealed an extensive pore network with diameters as large as 110 µm directly connected to the biofilm surface and surrounding fluid. Thin-section fluorescence in situ hybridization microscopy revealed that ammonia-oxidizing bacteria (AOB) distributed through the entire thickness of the biofilm. AOB were particularly concentrated in the biofilm around internal pores. Areal porosity values estimated from OCT scans were consistently lower than those estimated from multiphoton laser scanning microscopy, though the two imaging modalities showed a statistically significant correlation (r = 0.49, p < 0.0001). Estimates of areal porosity were moderately sensitive to gray-level threshold selection, though several automated thresholding algorithms yielded similar values to those obtained by manually thresholding performed by a panel of environmental engineering researchers (±25% relative error). These findings advance our ability to quantitatively describe the geometry of biofilm internal pore networks at length scales relevant to engineered biofilm reactors and suggest that internal pore structures provide crucial habitat for nitrifier growth.

Evaluation of an automated thresholding algorithm for the quantification of paraspinal muscle composition from MRI images

BackgroundThe imaging assessment of paraspinal muscle morphology and fatty infiltration has gained considerable attention in the past decades, with reports suggesting an association between muscle degenerative changes and low back pain (LBP). To date, qualitative and quantitative approaches have been used to assess paraspinal muscle composition. Though highly reliable, manual thresholding techniques are time consuming and not always feasible in a clinical setting. The tedious and rater-dependent nature of such manual thresholding techniques provides the impetus for the development of automated or semi-automated segmentation methods. The purpose of the present study was to develop and evaluate an automated thresholding algorithm for the assessment of paraspinal muscle composition. The reliability and validity of the muscle measurements using the new automated thresholding algorithm were investigated through repeated measurements and comparison with measurements from an established, highly reliable manual thresholding technique.MethodsMagnetic resonance images of 30 patients with LBP were randomly selected cohort of patients participating in a project on commonly diagnosed lumbar pathologies in patients attending spine surgeon clinics. A series of T2-weighted MR images were used to train the algorithm; preprocessing techniques including adaptive histogram equalization method image adjustment scheme were used to enhance the quality and contrast of the images. All muscle measurements were repeated twice using a manual thresholding technique and the novel automated thresholding algorithm, from axial T2-weigthed images, at least 5 days apart. The rater was blinded to all earlier measurements. Inter-method agreement and intra-rater reliability for each measurement method were assessed. The study did not received external funding and the authors have no disclosures.ResultsThere was excellent agreement between the two methods with inter-method reliability coefficients (intraclass correlation coefficients) varying from 0.79 to 0.99. Bland and Altman plots further confirmed the agreement between the two methods. Intra-rater reliability and standard error of measurements were comparable between methods, with reliability coefficient varying between 0.95 and 0.99 for the manual thresholding and 0.97–0.99 for the automated algorithm.ConclusionThe proposed automated thresholding algorithm to assess paraspinal muscle size and composition measurements was highly reliable, with excellent agreement with the reference manual thresholding method.

Integrated Assessment of Cardiac PET/MRI: Co-Registered PET and MRI Polar Plots by Mutual MR-Based Segmentation of the Left Ventricular Myocardium

Background: In the present study, we sought to describe a procedure for the creation of co-registered positron emission tomography (PET) and magnetic resonance imaging (MRI) polar plots of cardiac PET/MRI examinations, validate the resulting plots against available standard methods in patients with myocardial infarction and provide examples that demonstrate the advantage of the novel approach over existing standards. Methods: Co-registered LGE and PET short-axis images were transformed into polar maps based on a radial sampling pattern. LGE was automatically detected using an automated thresholding algorithm (ATA). In 20 PET/MRI examinations in patients with acute myocardial infarction, agreement between manual LGE assessment and the ATA classification was calculated. Also agreement between MRI-segmentation based PET polar plots and standard PET polar plots (created with the Corridor4DM software package) was assessed. Results: No statistically significant difference in infarct sizes between manual and ATA segmentation was found (p = 0.12). Both methods were highly correlated (Pearson’s r = 0.96, p < 0.01). Bland-Altman analysis revealed lower and upper limits of agreement of - 4.7 g and 3.6 g. Agreement between MRI-segmentation based PET polar plots and standard PET polar plots was very high (mean kappa of 0.91 ± 0.10; p < 0.01 in all cases). In three additional patients with myocardial inflammation, the software successfully created polar plots that demonstrate the location and extent of pathologic tracer uptake in the left ventricular myocardium. Conclusion: A straightforward software approach for the creation of co-registered PET and MRI polar plots was described and successfully demonstrated in PET/MRI studies of myocardial infarction and inflammation.

Object-based analysis of astroglial reaction and astrocyte subtype morphology after ischemic brain injury.

The astrocytic response to ischemic brain injury is characterized by specific alterations of glial cell morphology and function. Various studies described both beneficial and detrimental aspects of activated astrocytes, suggesting the existence of different subtypes. We investigated this issue using a novel object-based approach to study characteristics of astrogliosis after stroke. Spontaneously hypertensive rats received permanent middle cerebral artery occlusion. After 96 h, brain specimens were removed, fixed and stained for GFAP, glutamine synthetase (GS), S100Beta and Musashi1 (Msh1). Three regions of interest were defined (contralateral hemisphere, ipsilateral remote zone and infarct border zone), and confocal stacks were acquired (n=5 biological with each n=4 technical replicates). The stacks were background-corrected and colocalization between the selected markers and GFAP was determined using an automated thresholding algorithm. The fluorescence and colocalization channels were then converted into 3D-objects using both intensity and volume as filters to ultimately determine the final volumes of marker expression and colocalization, as well as the morphological changes of astrocyte process arborisation. We found that both S100Beta and Msh1 determined the same GFAP-positive astroglial cell population albeit the cellular compartments differed. GFAP stained most of the astrocyte processes and is hence suitable for the analysis of qualitative characteristics of astrogliosis. Due to its peri-nuclear localization, Msh1 is appropriate to estimate the total number of astrocytes even in regions with severe reactive astrogliosis. GS expression in GFAP-positive astrocytes was high in the remote zone and low at the infarct border, indicating the existence of astrocyte subclasses.

Twelve automated thresholding methods for segmentation of PET images: a phantom study

Tumor volume delineation over positron emission tomography (PET) images is of great interest for proper diagnosis and therapy planning. However, standard segmentation techniques (manual or semi-automated) are operator dependent and time consuming while fully automated procedures are cumbersome or require complex mathematical development. The aim of this study was to segment PET images in a fully automated way by implementing a set of 12 automated thresholding algorithms, classical in the fields of optical character recognition, tissue engineering or non-destructive testing images in high-tech structures. Automated thresholding algorithms select a specific threshold for each image without any a priori spatial information of the segmented object or any special calibration of the tomograph, as opposed to usual thresholding methods for PET. Spherical 18F-filled objects of different volumes were acquired on clinical PET/CT and on a small animal PET scanner, with three different signal-to-background ratios. Images were segmented with 12 automatic thresholding algorithms and results were compared with the standard segmentation reference, a threshold at 42% of the maximum uptake. Ridler and Ramesh thresholding algorithms based on clustering and histogram-shape information, respectively, provided better results that the classical 42%-based threshold (p < 0.05). We have herein demonstrated that fully automated thresholding algorithms can provide better results than classical PET segmentation tools.

Selection of the automated thresholding algorithm for the Multi-angle Imaging SpectroRadiometer Radiometric Camera-by-Camera Cloud Mask over land

The Radiometric Camera-by-Camera Cloud Mask (RCCM) is archived at the NASA Langley Distributed Active Archive Center as one of the standard products from the Multi-angle Imaging SpectroRadiometer (MISR) mission. The RCCM algorithm applied over land surfaces uses an Automated Threshold Selection Algorithm (ATSA) to derive thresholds that are applied to a cloud masking test to determine whether a given image pixel is clear or contains cloud. In this article, we established a framework for the selection of ATSA and the cloud masking tests, which is not only suitable for the RCCM over land, but cloud detection for other satellite missions. Using this framework, we have undertaken the largest comparison of existing histogram-based ATSAs (16 in total) and applied them to four cloud masking tests that can be constructed from the MISR radiances, namely the red channel bidirectional reflectance function (BRF), the standard deviation (STDV) of the red channel BRF, the normalized difference vegetation index (NDVI), and a parameter D that is constructed by optimizing the information from NDVI and red channel BRF for cloud detection. The cloud masking tests and ATSAs are applied to 35 MISR scenes from six snow-free land surface types. To evaluate their performance, reference cloud masks are constructed for the 35 scenes using interactive, supervised learning, visualization software. Independent of the ATSA and as a single cloud masking test, D performed the best in terms of having the lowest misclassification rate using the best possible threshold, the highest bimodal rate in the shape of the histograms derived from the 35 scenes, and the least sensitivity to errors in the choice of threshold. Of the 16 ATSAs, the methods of Li and Lee [Li, C.H., and Lee, C.K., (1993). Minimum cross-entropy thresholding. Pattern Recognition, 26(4), 617-625.] and Pal and Bhandari [Pal, N. R., and Bhandari, D., (1993). Image thresholding: some new techniques. Signal Processing, 33, 139–158.] performed the best when applied to D, with essentially unbiased performance and a root mean square of 15% when compared to cloud masks using the best possible thresholds. It is recommended that increased performance of the RCCM-land algorithm can be had through an increase in the space–time sampling used to generate histograms of D and the addition of a STDV cloud masking test to improve the detection of small cumulus clouds.

Recognition of clinker phases by automatic image analysis

Microscopical analysis of clinker phases provides useful informations about its structure and its manufacturing process. This paper deals with automatic segmentation of clinker phases images acquired on polished sections using optical microscopy. The main difficulty lies in the segmentation step. Even sophisticated techniques using textural characterization did not give satisfactory results. Indeed, the concerned textures show variations from an image to another and their spatial extent may vary as well. The most efficient methods we have developed combine automated thresholding algorithms and the use of contrast. This latter notion is defined within the scope of a mathematical model: the logarithmic image processing (LIP) model. Considering the limited reproductibility of the chemical etches used to bring the phases to the fore and the variations in terms of shape and hue of the elements to be segmented, the proposed algorithms have to be adapted to clinkers of various origin. However their efficiency has been proved for one type of clinker.