Multi-scale characterizations of colon polyps via computed tomographic colonography

Specular highlight removal for endoscopic images using partial attention network

Objective.Endoscopic imaging is a visualization method widely used in minimally invasive surgery. However, owing to the strong reflection of the mucus layer on the organs, specular highlights often appear to degrade the imaging performance. Thus, it is necessary to develop an effective highlight removal method for endoscopic imaging.Approach.A specular highlight removal method using a partial attention network (PatNet) for endoscopic imaging is proposed to reduce the interference of bright light in endoscopic surgery. The method is designed as two procedures: highlight segmentation and endoscopic image inpainting. Image segmentation uses brightness threshold based on illumination compensation to divide the endoscopic image into the highlighted mask and the non-highlighted area. The image inpainting algorithm uses a partial convolution network that integrates an attention mechanism. A mask dataset with random hopping points is designed to simulate specular highlight in endoscopic imaging for network training. Through the filtering of masks, the method can focus on recovering defective pixels and preserving valid pixels as much as possible.Main results.The PatNet is compared with 3 highlight segmentation methods, 3 imaging inpainting methods and 5 highlight removal methods for effective analysis. Experimental results show that the proposed method provides better performance in terms of both perception and quantification. In addition, surgeons are invited to score the processing results for different highlight removal methods under realistic reflection conditions. The PatNet received the highest score of 4.18. Correspondingly, the kendall's W is 0.757 and the asymptotic significancep= 0.000 < 0.01, revealing that the subjective scores have good consistency and confidence.Significance.Generally, the method can realize irregular shape highlight reflection removal and image restoration close to the ground truth of endoscopic images. This method can improve the quality of endoscopic imaging for accurate image analysis.

MM-GLCM-CNN: A multi-scale and multi-level based GLCM-CNN for polyp classification

Distinguishing malignant from benign lesions has significant clinical impacts on both early detection and optimal management of those early detections. Convolutional neural network (CNN) has shown great potential in medical imaging applications due to its powerful feature learning capability. However, it is very challenging to obtain pathological ground truth, addition to collected in vivo medical images, to construct objective training labels for feature learning, leading to the difficulty of performing lesion diagnosis. This is contrary to the requirement that CNN algorithms need a large number of datasets for the training. To explore the ability to learn features from small pathologically-proven datasets for differentiation of malignant from benign polyps, we propose a Multi-scale and Multi-level based Gray-level Co-occurrence Matrix CNN (MM-GLCM-CNN). Specifically, instead of inputting the lesions' medical images, the GLCM, which characterizes the lesion heterogeneity in terms of image texture characteristics, is fed into the MM-GLCN-CNN model for the training. This aims to improve feature extraction by introducing multi-scale and multi-level analysis into the construction of lesion texture characteristic descriptors (LTCDs). To learn and fuse multiple sets of LTCDs from small datasets for lesion diagnosis, we further propose an adaptive multi-input CNN learning framework. Furthermore, an Adaptive Weight Network is used to highlight important information and suppress redundant information after the fusion of the LTCDs. We evaluated the performance of MM-GLCM-CNN by the area under the receiver operating characteristic curve (AUC) merit on small private lesion datasets of colon polyps. The AUC score reaches 93.99% with a gain of 1.49% over current state-of-the-art lesion classification methods on the same dataset. This gain indicates the importance of incorporating lesion characteristic heterogeneity for the prediction of lesion malignancy using small pathologically-proven datasets.

The role of deep learning in diagnosing colorectal cancer

Colon cancer is a major public health issue, affecting a growing number of individuals worldwide. Proper and early diagnosis of colon cancer is the necessary first step toward effective treatment and/or prevention of future disease relapse. Artificial intelligence and its subtypes, deep learning in particular, tend nowadays to have an expanding role in all fields of medicine, and diagnosing colon cancer is no exception. This report aims to summarize the entire application spectrum of deep learning in all diagnostic tests regarding colon cancer, from endoscopy and histologic examination to medical imaging and screening serologic tests.

Performance sensitivity analysis of brain metastasis stereotactic radiosurgery outcome prediction using MRI radiomics

Recent studies have used T1w contrast-enhanced (T1w-CE) magnetic resonance imaging (MRI) radiomic features and machine learning to predict post-stereotactic radiosurgery (SRS) brain metastasis (BM) progression, but have not examined the effects of combining clinical and radiomic features, BM primary cancer, BM volume effects, and using multiple scanner models. To investigate these effects, a dataset of n = 123 BMs from 99 SRS patients with 12 clinical features, 107 pre-treatment T1w-CE radiomic features, and BM progression determined by follow-up MRI was used with a random decision forest model and 250 bootstrapped repetitions. Repeat experiments assessed the relative accuracy across primary cancer sites, BM volume groups, and scanner model pairings. Correction for accuracy imbalances across volume groups was investigated by removing volume-correlated features. We found that using clinical and radiomic features together produced the most accurate model with a bootstrap-corrected area under the receiver operating characteristic curve of 0.77. Accuracy also varied by primary cancer site, BM volume, and scanner model pairings. The effect of BM volume was eliminated by removing features at a volume-correlation coefficient threshold of 0.25. These results show that feature type, primary cancer, volume, and scanner model are all critical factors in the accuracy of radiomics-based prognostic models for BM SRS that must be characterised and controlled for before clinical translation.

Machine Learning-Based Gray-Level Co-Occurrence Matrix (GLCM) Models for Predicting the Depth of Myometrial Invasion in Patients with Stage I Endometrial Cancer

Purpose

Deep myometrial invasion (DMI) is an independent high-risk factor for lymph node metastasis and a prognostic risk factor in early-stage endometrial cancer (EC-I) patients. Thus, we developed a machine learning (ML) assistant model, which can accurately help define the surgical area.

Methods

348 consecutive EC-I patients with the pathological diagnosis were recruited in the tertiary medical centre between January 1, 2012, and October 31, 2021. Five ML-assisted models were developed using two-step estimation methods from the candidate gray level co-occurrence matrix (GLCM). Receiver operating characteristic curve (ROC), decision curve analysis (DCA), and clinical impact curve (CIC) were prepared to evaluate the robustness and clinical practicality of each model.

Results

Our analysis identified several significant differences between the stage IA and IB groups. The top seven-candidate factors included correlation all direction offset1, correlation angle0 offset1, correlation angle45 offset1, correlation angle90 offset1, ID moment all direction offset1, ID moment angle0 offset1, and ID moment angle45 offset1. The areas under the ROC curve (AUCs) of the random forest classifier (RFC) model, support vector machine (SVM), eXtreme gradient boosting (XGBoost), artificial neural network (ANN), and decision tree (DT) ranged from 0.765 to 0.877 in the training set and from 0.716 to 0.862 in the testing set, respectively. Among the five machine algorithms, RFC obtained the optimal prediction efficiency using correlation angle0 offset1, correlation angle45 offset1, correlation angle90 offset1, correlation all direction offset1, ID moment angle0 offset1, and ID moment angle45 offset1, and ID moment angle90 offset1, respectively.

Conclusion

Our ML-based prediction model combined with GLCM parameters assessed the risk of DMI in EC-I patients, especially RFC, which helped distinguish stage IA and IB EC patients. This new predictive model based on supervised learning can be used to establish personalized treatment strategies.

Analysis of Vascular Architecture and Parenchymal Damage Generated by Reduced Blood Perfusion in Decellularized Porcine Kidneys Using a Gray Level Co-occurrence Matrix

There is no cure for kidney failure, but a bioartificial kidney may help address this global problem. Decellularization provides a promising platform to generate transplantable organs. However, maintaining a viable vasculature is a significant challenge to this technology. Even though angiography offers a valuable way to assess scaffold structure/function, subtle changes are overlooked by specialists. In recent years, various image analysis methods in radiology have been suggested to detect and identify subtle changes in tissue architecture. The aim of our research was to apply one of these methods based on a gray level co-occurrence matrix (Topalovic et al.) computational algorithm in the analysis of vascular architecture and parenchymal damage generated by hypoperfusion in decellularized porcine. Perfusion decellularization of the whole porcine kidneys was performed using previously established protocols. We analyzed and compared angiograms of kidneys subjected to pathophysiological arterial perfusion of whole blood. For regions of interest Santos et al. covering kidney medulla and the main elements of the vascular network, five major GLCM features were calculated: angular second moment as an indicator of textural uniformity, inverse difference moment as an indicator of textural homogeneity, GLCM contrast, GLCM correlation, and sum variance of the co-occurrence matrix. In addition to GLCM, we also performed discrete wavelet transform analysis of angiogram ROIs by calculating the respective wavelet coefficient energies using high and low-pass filtering. We report statistically significant changes in GLCM and wavelet features, including the reduction of the angular second moment and inverse difference moment, indicating a substantial rise in angiogram textural heterogeneity. Our findings suggest that the GLCM method can be successfully used as an addition to conventional fluoroscopic angiography analyses of micro/macrovascular integrity following in vitro blood perfusion to investigate scaffold integrity. This approach is the first step toward developing an automated network that can detect changes in the decellularized vasculature.

Adaptive brightness fusion method for intraoperative near-infrared fluorescence and visible images

An adaptive brightness fusion method (ABFM) for near-infrared fluorescence imaging is proposed to adapt to different lighting conditions and make the equipment operation more convenient in clinical applications. The ABFM is designed based on the network structure of Attention Unet, which is an image segmentation technique. Experimental results show that ABFM has the function of adaptive brightness adjustment and has better fusion performance in terms of both perception and quantification. Generally, the proposed method can realize an adaptive brightness fusion of fluorescence and visible images to enhance the usability of fluorescence imaging technology during surgery.

Gray-Level Co-occurrence Matrix Analysis for the Detection of Discrete, Ethanol-Induced, Structural Changes in Cell Nuclei: An Artificial Intelligence Approach

Gray-level co-occurrence matrix (GLCM) analysis is a contemporary and innovative computational method for the assessment of textural patterns, applicable in almost any area of microscopy. The aim of our research was to perform the GLCM analysis of cell nuclei in Saccharomyces cerevisiae yeast cells after the induction of sublethal cell damage with ethyl alcohol, and to evaluate the performance of various machine learning (ML) models regarding their ability to separate damaged from intact cells. For each cell nucleus, five GLCM parameters were calculated: angular second moment, inverse difference moment, GLCM contrast, GLCM correlation, and textural variance. Based on the obtained GLCM data, we applied three ML approaches: neural network, random trees, and binomial logistic regression. Statistically significant differences in GLCM features were observed between treated and untreated cells. The multilayer perceptron neural network had the highest classification accuracy. The model also showed a relatively high level of sensitivity and specificity, as well as an excellent discriminatory power in the separation of treated from untreated cells. To the best of our knowledge, this is the first study to demonstrate that it is possible to create a relatively sensitive GLCM-based ML model for the detection of alcohol-induced damage in Saccharomyces cerevisiae cell nuclei.