A New Method Based on Adaptive Discrete Wavelet Entropy Energy and Neural Network Classifier (ADWEENN) for Recognition of Urine Cells from Microscopic Images Independent of Rotation and Scaling

Microscopic urinary particle detection by different YOLOv5 models with evolutionary genetic algorithm based hyperparameter optimization

The diagnosis of kidney disease often involves analysing urine sediment particles. Traditionally, urinalysis was performed manually by collecting urine samples and using a centrifuge, which was prone to manual errors and relied on labour-intensive processes. Automated urine sediment microscopy, based on machine learning models, requires segmentation and feature extraction, which can hinder model performance due to intrinsic characteristics of microscopic images. Deep learning models based on convolutional neural networks (CNNs) often rely on a large number of manually annotated data, making the system computationally complex. This study propose an advanced deep learning model based on YOLOv5, which offers faster performance and requires comparatively less data. The proposed model used five variants of the YOLOv5 model (YOLOv5n, YOLOv5s, YOLOv5m, YOLOv5l, and YOLOv5x) to detect six categories of urine particles (erythrocyte, leukocyte, crystals, cast, mycete, epithelial cells) from microscopic urine sediment images. The dataset involved 5376 images of urine sediments with 6 particles. There are 30 sets of hyperparamreteres are employed in the YOLOv5 model. To optimize the hyperparameters and fine-tune with the urine sediment dataset and for training each model, the system employed a genetic algorithm (GA) based on evolutionary principles named as Evolutionary Genetic Algorithm (EGA). Among the six categories of detected particles mycete achieved maximum performance with a mAP of 97.6 % and crystals achieved minimum performance with a mAP of 81.7 % with YOLOv5x model compared to other particles. To optimize the hyperparameters for training each model, the system employed a genetic algorithm (GA) based on evolutionary principles named as Evolutionary Genetic Algorithm (EGA). Among all the models, YOLOv5l and YOLOv5x performed the best. YOLOv5l achieved a mean average precision (mAP) of 85.8 % while YOLOv5x achieved a mAP of 85.4 % at an IoU threshold of 0.5. The detection speed per image was 23.4 ms for YOLOv5l and 28.4 ms for YOLOv5x. The proposed method developed a faster and better automated microscopic model using advanced deep learning techniques to detect urinary particles from microscopic urine sediment images for kidney disease identification. The method demonstrated strong performance in urinalysis.

Applications of Artificial Intelligence in Urinalysis: Is the Future Already Here?

BACKGROUND

Artificial intelligence (AI) has emerged as a promising and transformative tool in the field of urinalysis, offering substantial potential for advancements in disease diagnosis and the development of predictive models for monitoring medical treatment responses.

CONTENT

Through an extensive examination of relevant literature, this narrative review illustrates the significance and applicability of AI models across the diverse application area of urinalysis. It encompasses automated urine test strip and sediment analysis, urinary tract infection screening, and the interpretation of complex biochemical signatures in urine, including the utilization of cutting-edge techniques such as mass spectrometry and molecular-based profiles.

SUMMARY

Retrospective studies consistently demonstrate good performance of AI models in urinalysis, showcasing their potential to revolutionize clinical practice. However, to comprehensively evaluate the real clinical value and efficacy of AI models, large-scale prospective studies are essential. Such studies hold the potential to enhance diagnostic accuracy, improve patient outcomes, and optimize medical treatment strategies. By bridging the gap between research and clinical implementation, AI can reshape the landscape of urinalysis, paving the way for more personalized and effective patient care.

Detection and classification of hepatocytes and hepatoma cells using atomic force microscopy and machine learning algorithms

Hepatocellular carcinoma is a high-risk malignant tumor. Hepatoma cells are transformed from normal cells and have unique surface nanofeatures in addition to the characteristics of the original cells. In this paper, atomic force microscopy was used to extract the three-dimensional morphology and mechanical information of HL-7702 human hepatocytes and SMMC-7721 and HepG2 hepatoma cells in culture, such as the elastic modulus and viscoelasticity. The characteristics of different cells were compared and analyzed. Finally, the cell morphology and mechanics information were used for training machine learning algorithms. With the trained model, the detection of cells was realized. The classification accuracy was as high as 94.54%, and the area under the receiver operating characteristic (ROC) curve (AUC) was 0.99. Thus, hepatocytes and hepatoma cells were accurately identified and assessed. We also compared the classification effects of other machine learning algorithms, such as support vector machine and logistic regression. Our method extracts cellular nanofeatures directly from the surface of cells of unknown type for cell classification. Compared with microscope image-based analysis and other methods, this approach can avoid the misjudgment that may occur when different doctors have different levels of experience. Thus, the proposed method provides an objective basis for the early diagnosis of hepatocellular carcinoma. RESEARCH HIGHLIGHTS: The 3D appearance and mechanical characteristics of hepatocellular carcinoma cells are very similar to those of hepatocytes. Application of atomic force microscopy with machine learning algorithm. Collect the data set of nano-characteristic parameters of the cell. The machine learning algorithms is trained by data set, and its classification effect is better than that of a single nano-parameter.

Effects of Iron Oxide Nanoparticles on Structural Organization of Hepatocyte Chromatin: Gray Level Co-occurrence Matrix Analysis

Gray level co-occurrence matrix (GLCM) analysis is a contemporary and innovative computer-based algorithm that can be used for the quantification of subtle changes in a cellular structure. In this work, we use this method for the detection of discrete alterations in hepatocyte chromatin distribution after in vivo exposure to iron oxide nanoparticles (IONPs). The study was performed on 40 male, healthy C57BL/6 mice divided into four groups: three experimental groups that received different doses of IONPs and 1 control group. We describe the dose-dependent reduction of chromatin textural uniformity measured as GLCM angular second moment. Similar changes were detected for chromatin textural uniformity expressed as the value of inverse difference moment. To the best of our knowledge, this is the first study to investigate the impact of iron-based nanomaterials on hepatocyte GLCM parameters. Also, this is the first study to apply discrete wavelet transform analysis, as a supplementary method to GLCM, for the assessment of hepatocyte chromatin structure in these conditions. The results may present the useful basis for future research on the application of GLCM and DWT methods in pathology and other medical research areas.

Inspection of visible components in urine based on deep learning

PURPOSE

Urinary particles are particularly important parameters in clinical urinalysis, especially for the diagnosis of nephropathy. Therefore, it is highly important to precisely detect urinary particles in the clinical setting. However, artificial microscopy is subjective and time consuming, and various previous detection algorithms lack the adequate accuracy. In this study, a method is proposed for the analysis of urinary particles based on deep learning.

METHODS

We used seven cellular components (i.e., erythrocytes, leukocytes, epithelial, low-transitional epithelium, casts, crystal, and squamous epithelial cells) in the microscopic imaging of urine as the detection targets. After the extraction of features using Resnet50, feature maps of different sizes are obtained in the last few layers of the feature pyramid net (FPN). The feature maps are then input into the classification subnetwork and regression subnetwork for classification and localization respectively, and detection results are obtained. First, we introduce the basic model (RetinaNet) to detect the cellular components in urinary particles, and the features of the objects can then be extracted more effectively by replacing different basic networks. Lastly, the effects of different weight initialization methods and different anchor scales on the performance of the model are investigated.

RESULTS

We obtained the optimal network structure based on the adjustment of the loss functional parameters, thereby achieving the best results in the test set of urinary particles. The experimental data yielded an accuracy of 88.65% with a processing time of only 0.2 s for each image on a GeForce GTX 1080 graphics processing unit (GPU). Our results demonstrate that this method cannot only achieve the speed of the first-stage target detector, but also the accuracy of the two-stage target algorithm in the analysis of urinary particles.

CONCLUSION

This study developed new automated analysis urinary particles based on deep learning, and this method is expected to be used for the automated analysis and detection of urinary particles. Moreover, our approach will be useful for the detection of other cells in the clinical setting.

Radiographic predictors determined with an objective assessment tool for neonatal patients with necrotizing enterocolitis

OBJECTIVE

The objective of this study was to develop and validate a computational tool to assist radiological decisions on necrotizing enterocolitis.

METHODOLOGY

Patients that exhibited clinical signs and radiographic evidence of Bell's stage 2 or higher were included in the study, resulting in 64 exams. The tool was used to classify localized bowel wall thickening and intestinal pneumatosis using full-width at half-maximum measurements and texture analyses based on wavelet energy decomposition. Radiological findings of suspicious bowel wall thickening and intestinal pneumatosis loops were confirmed by both patient surgery and histopathological analysis. Two experienced radiologists selected an involved bowel and a normal bowel in the same radiography. The full-width at half-maximum and wavelet-based texture feature were then calculated and compared using the Mann-Whitney U test. Specificity, sensibility, positive and negative predictive values were calculated.

RESULTS

The full-width at half-maximum results were significantly different between normal and distended loops (median of 10.30 and 15.13, respectively). Horizontal, vertical, and diagonal wavelet energy measurements were evaluated at eight levels of decomposition. Levels 7 and 8 in the horizontal direction presented significant differences. For level 7, median was 0.034 and 0.088 for normal and intestinal pneumatosis groups, respectively, and for level 8 median was 0.19 and 0.34, respectively.

CONCLUSIONS

The developed tool could detect differences in radiographic findings of bowel wall thickening and IP that are difficult to diagnose, demonstrating the its potential in clinical routine. The tool that was developed in the present study may help physicians to investigate suspicious bowel loops, thereby considerably improving diagnosis and clinical decisions.

An End-to-End System for Automatic Urinary Particle Recognition with Convolutional Neural Network

The urine sediment analysis of particles in microscopic images can assist physicians in evaluating patients with renal and urinary tract diseases. Manual urine sediment examination is labor-intensive, subjective and time-consuming, and the traditional automatic algorithms often extract the hand-crafted features for recognition. Instead of using the hand-crafted features, in this paper we propose to exploit convolutional neural network (CNN) to learn features in an end-to-end manner to recognize the urinary particle. We treat the urinary particle recognition as object detection and exploit two state-of-the-art CNN-based object detection methods, Faster R-CNN and single shot multibox detector (SSD), along with their variants for urinary particle recognition. We further investigate different factors involving these CNN-based methods to improve the performance of urinary particle recognition. We comprehensively evaluate these methods on a dataset consisting of 5,376 annotated images corresponding to 7 categories of urinary particle, i.e., erythrocyte, leukocyte, epithelial cell, crystal, cast, mycete, epithelial nuclei, and obtain a best mean average precision (mAP) of 84.1% while taking only 72 ms per image on a NVIDIA Titan X GPU.

Artificial Neural Network for Total Laboratory Automation to Improve the Management of Sample Dilution

Diluting a sample to obtain a measure within the analytical range is a common task in clinical laboratories. However, for urgent samples, it can cause delays in test reporting, which can put patients' safety at risk. The aim of this work is to show a simple artificial neural network that can be used to make it unnecessary to predilute a sample using the information available through the laboratory information system. Particularly, the Multilayer Perceptron neural network built on a data set of 16,106 cardiac troponin I test records produced a correct inference rate of 100% for samples not requiring predilution and 86.2% for those requiring predilution. With respect to the inference reliability, the most relevant inputs were the presence of a cardiac event or surgery and the result of the previous assay. Therefore, such an artificial neural network can be easily implemented into a total automation framework to sensibly reduce the turnaround time of critical orders delayed by the operation required to retrieve, dilute, and retest the sample.

Automatic Identification of Human Erythrocytes in Microscopic Fecal Specimens

Traditional fecal erythrocyte detection is performed via a manual operation that is unsuitable because it depends significantly on the expertise of individual inspectors. To recognize human erythrocytes automatically and precisely, automatic segmentation is very important for extraction of characteristics. In addition, multiple recognition algorithms are also essential. This paper proposes an algorithm based on morphological segmentation and a fuzzy neural network. The morphological segmentation process comprises three operational steps: top-hat transformation, Otsu's method, and image binarization. Following initial screening by area and circularity, fuzzy c-means clustering and the neural network algorithms are used for secondary screening. Subsequently, the erythrocytes are screened by combining the results of five images obtained at different focal lengths. Experimental results show that even when the illumination, noise pollution, and position of the erythrocytes are different, they are all segmented and labeled accurately by the proposed method. Thus, the proposed method is robust even in images with significant amounts of noise.

A new method based for diagnosis of breast cancer cells from microscopic images: DWEE--JHT

In these days, there are many various diseases, whose diagnosis is very hardly. Breast cancer is one of these type diseases. In this paper, accuracy diagnosis of normal, benign, and malign breast cancer cell were found by combining mean success rates Jensen Shannon, Hellinger, and Triangle measure which connected with each other. In this article, an diagnostic method based on feature extraction Discrete Wavelet Entropy Energy (DWEE) and Jensen Shannon, Hellinger, Triangle Measure (JHT) Classifier for diagnosis of breast cancer. This diagnosis method is called as DWEE--JHT this paper. With this diagnosis method have found optimal feature subset using discrete wavelet transform feature extraction. Then these convenient features are given to Jensen Shannon, Hellinger, Triangle Measure (JHT) classifier. Then, between classifiers which are Jensen Shannon, Hellinger, and triangle distance have been validated the measures via relationships. Afterwards, breast cancer cells are classified using Jensen Shannon, Hellinger, and Triangle distance. Mean success rate of 16 feature vector with Jensen Shannon classifier is found % 97.81. Mean success rate of 16 feature vector with Hellinger classifier is found % 97.75. Mean success rate of 16 feature vector with Triangle classifier is found % 97.87. By averaging of results obtained from these 3 classifiers are found as 97.81 % average of accuracy diagnosis.

Brain MR image segmentation with spatial constrained K-mean algorithm and dual-tree complex wavelet transform

In brain MR images, the noise and low-contrast significantly deteriorate the segmentation results. In this paper, we propose an automatic unsupervised segmentation method integrating dual-tree complex wavelet transform (DT-CWT) with K-mean algorithm for brain MR image. Firstly, a multi-dimensional feature vector is constructed based on the intensity, the low-frequency subband of DT-CWT and spatial position information. Then, a spatial constrained K-mean algorithm is presented as the segmentation system. The proposed method is validated by extensive experiments using both simulated and real T1-weighted MR images, and compared with the state-of-the-art algorithms.

A stationary wavelet transform based approach to registration of planning CT and setup cone beam-CT images in radiotherapy

Image registration between planning CT images and cone beam-CT (CBCT) images is one of the key technologies of image guided radiotherapy (IGRT). Current image registration methods fall roughly into two categories: geometric features-based and image grayscale-based. Mutual information (MI) based registration, which belongs to the latter category, has been widely applied to multi-modal and mono-modal image registration. However, the standard mutual information method only focuses on the image intensity information and overlooks spatial information, leading to the instability of intensity interpolation. Due to its use of positional information, wavelet transform has been applied to image registration recently. In this study, we proposed an approach to setup CT and cone beam-CT (CBCT) image registration in radiotherapy based on the combination of mutual information (MI) and stationary wavelet transform (SWT). Firstly, SWT was applied to generate gradient images and low frequency components produced in various levels of image decomposition were eliminated. Then inverse SWT was performed on the remaining frequency components. Lastly, the rigid registration of gradient images and original images was implemented using a weighting function with the normalized mutual information (NMI) being the similarity measure, which compensates for the lack of spatial information in mutual information based image registration. Our experiment results showed that the proposed method was highly accurate and robust, and indicated a significant clinical potential in improving the accuracy of target localization in image guided radiotherapy (IGRT).