Lung Nodule Detection Using Ensemble Classifier in Computed Tomography Images

In diagnostic ultrasound, an ultrasound transducer converts an electrical signal into an ultrasound pulse, which enters the tissue from the body surface. At the surface, an echo appears. The probe senses and receives the echo, and all the echoes are converted back to signals and graphics, which can be analyzed by medical staff. We studied the neural network (NN)-based classification of hepatocellular carcinoma (HCC) and liver abscess using texture features of ultrasound images. From 79 cases of liver diseases (44 liver cancer and 35 liver abscess cases), we extracted 52 features of the gray-level co-occurrence matrix (GLCM) and 44 features of the gray-level run-length matrix (GLRLM), giving a total of 96 features. We used three feature selection models to distinguish these two liver diseases: sequential forward selection (SFS), sequential backward selection (SBS), and F-score. We proved that our developed system can be used to classify liver cancer and liver abscess using an NN with an accuracy of 88.375%, which can provide diagnostic assistance for inexperienced clinicians.

Section: Nnmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Classification of Hepatocellular Carcinoma and Liver Abscess by Applying Neural Network to Ultrasound Images

Xu¹,

Chang²,

Su³

et al. 2020

“…X-rays are extensively utilized in various optical imaging fields, such as computed tomography (1,2) and biomedical imaging. (3,4) To improve the resolution and speed of optical imaging, researchers have been exploring novel X-ray sources.…”

Section: Introductionmentioning

confidence: 99%

Effect of Laser Intensity on an Electron’s Optimal Position and Spatial Radiation Characteristics in a Linearly Polarized Tightly Focused Laser Field

Qiu¹,

Tian²,

Chen³

2023

X-rays are widely used in the field of optical imaging. A novel type of X-ray source is the relativistic nonlinear Thomson scattering (RNTS). Within the framework of RNTS, we investigated in detail the effects of different intensities of linearly polarized tightly focused lasers on the maximum radiation power, optimal position of an electron (initial position of the electron that yields the maximum radiation power in the entire space), as well as the motion and spatial radiation features of the electron at the optimal position. Our results reveal, for the first time, that the optimal position and maximum radiation power of the electron exhibit significant linear and exponential dependences, respectively, on the laser amplitude. After interacting with the laser, the initially stationary electron at the optimal position first undergoes an oscillatory motion and then moves linearly. The entire trajectory of the electron is asymmetric. As the laser intensity increases, the spatial and angular distributions of the radiation become smaller, and the collimation of radiation increases. In the direction of maximum power radiation, the azimuth angle φ remains at 0°, whereas the corresponding polar angle θ decreases from 41° to 20°, indicating that the spatial radiation approaches the z-axis. The time when the maximum radiation power is obtained is approximately 50.7 fs. Additionally, the asymmetry and micro-double-peak structure of the temporal spectra, as well as the modulation characteristics of the frequency spectra, were further investigated in the direction of the maximum power radiation. These findings will assist researchers in achieving high-power and ultrashort X-rays in experiments, thereby increasing the resolution and imaging speed of optical imaging.

“…In recent years, computer-aided technologies have been widely adopted to achieve automated inspection for health care delivery, having a wide range of applications from scalp inspection (1) to lung nodule detection (2) in radiology and lesion classification (3) in histopathology. However, biomedical image analysis is a complex task that relies on highly skilled domain experts, such as radiologists and pathologists.…”

Section: Introductionmentioning

confidence: 99%

Understanding the Mechanism of Deep Learning Frameworks in Lesion Detection for Pathological Images with Breast Cancer

Hsu¹,

Chen²,

Chang³

et al. 2022

With the advances of scanning sensors and deep learning algorithms, computational pathology has drawn much attention in recent years and started to play an important role in the clinical workflow. Computer-aided detection (CADe) systems have been developed to assist pathologists in slide assessment, increasing diagnosis efficiency and reducing misdetections. In this study, we conducted four experiments to demonstrate that the features learned by deep learning models are interpretable from a pathological perspective. In addition, classifiers such as the support vector machine (SVM) and random forests (RF) were used in experiments to replace the fully connected layers and decompose the end-to-end framework, verifying the validity of feature extraction in the convolutional layers. The experimental results reveal that the features learned from the convolutional layers work as morphological descriptors for specific cells or tissues, in agreement with the diagnostic rules in practice. Most of the properties learned by the deep learning models summarized detection rules that agree with those of experienced pathologists. The interpretability of deep features from a clinical viewpoint not only enhances the reliability of AI systems, enabling them to gain acceptance from medical experts, but also facilitates the development of deep learning frameworks for different tasks in pathological analytics.