• 대한의용생체공학회:의공학회지
• /
• 제10권3호
• /
• pp.331-342
• /
• 1989

This paper presents a new algorithm that extracts lung region in X-Rays and enhance.j the region. Comparing to prior algorithms that enhance whole X-Ray image, this algorithm leads more effective results. For this algorithm extracts lung region first, and enhances the lung region excluding parameters of other region. For choosing optimal threshold, we compare OTSU's mothod with the proposed method. We obtain lung boundary using contour following algorithm and Rray level searching method in gray level rescaled image. We Process histogram equalization in lung region and obtain enhanced lung image. By using the proposed algorithm, we obtain lung region effectively in chest X-Ray that need in medical image diagnostic system.

• 대한의용생체공학회:학술대회논문집
• /
• 대한의용생체공학회 1997년도 춘계학술대회
• /
• pp.465-468
• /
• 1997

A Study on Computer Assisted Diagnosis (CAD) system extract ing lung cancer part from Digital X-ray Computerized Tomography(CT) image is discussed in this paper. It is very crucial to segment the image of lung into the three organ area such as inside, outside and the hilum so that the variant image processing algorithm can be applied an each area respectively. In this paper, the efficient algorithm extracting lung cancer part is proposed with characterizing lung hilum part and its associated vessel patterns.

• 대한방사선기술학회지:방사선기술과학
• /
• 제45권1호
• /
• pp.41-47
• /
• 2022

The purpose of this study is to apply a deep learning model that can distinguish lung perfusion and lung ventilation images in nuclear medicine, and to evaluate the image classification ability. Image data pre-processing was performed in the following order: image matrix size adjustment, min-max normalization, image center position adjustment, train/validation/test data set classification, and data augmentation. The convolutional neural network(CNN) structures of VGG-16, ResNet-18, Inception-ResNet-v2, and SE-ResNeXt-101 were used. For classification model evaluation, performance evaluation index of classification model, class activation map(CAM), and statistical image evaluation method were applied. As for the performance evaluation index of the classification model, SE-ResNeXt-101 and Inception-ResNet-v2 showed the highest performance with the same results. As a result of CAM, cardiac and right lung regions were highly activated in lung perfusion, and upper lung and neck regions were highly activated in lung ventilation. Statistical image evaluation showed a meaningful difference between SE-ResNeXt-101 and Inception-ResNet-v2. As a result of the study, the applicability of the CNN model for lung scintigraphy classification was confirmed. In the future, it is expected that it will be used as basic data for research on new artificial intelligence models and will help stable image management in clinical practice.

• 대한의용생체공학회:의공학회지
• /
• 제11권2호
• /
• pp.217-226
• /
• 1990

This paper presents a new algorithm that extracted lung region in X-ray and enhanced the region. With a lung region that was extracted by histogram threshold value, it was diffi cult to detect perfect lung boundary. Therefore we presented perfect lung boundary detection method using apex detection and apex region restoration. Also, by applying modified equalization algorithm and presented function to inside of lung region, we want to give help to automatic diagnosis In X-ray chest image. Presented main line trace algorithm gave good result in detection of lung boundary And, as apex detection method using lung row and column gray level average value found more correct place of lung than the rpethod of prior algorithm, we succeeded perfect lung region detection, Also, presented function that had lung region's gray level distribution characteristic was very effective to image enhancement.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권7호
• /
• pp.2359-2376
• /
• 2022

With the exponential growth of medical image big data represented by high-resolution CT images(CTI), the high-resolution CTI data is of great importance for clinical research and diagnosis. The paper takes lung CTI as an example to study. Retrieving answer CTIs similar to the input one from the large-scale lung CTI database can effectively assist physicians to diagnose. Compared with the conventional content-based image retrieval(CBIR) methods, the CBIR for lung CTIs demands higher retrieval accuracy in both the contour shape and the internal details of the organ. In traditional supervised deep learning networks, the learning of the network relies on the labeling of CTIs which is a very time-consuming task. To address this issue, the paper proposes a Weakly Supervised Similarity Evaluation Network (WSSENet) for efficiently support similarity analysis of lung CTIs. We conducted extensive experiments to verify the effectiveness of the WSSENet based on which the CBIR is performed.

• 한국의학물리학회:학술대회논문집
• /
• 한국의학물리학회 2002년도 Proceedings
• /
• pp.454-456
• /
• 2002

The evaluation of interval changes between temporally sequential chest radiographs is necessary for the detection of new abnormalities or interval changes, such as pulmonary nodules and interstitial disease. For interstitial lung disease, the interval changes are very important for diagnosis and treatment. Especially, interstitial lung disease may show rapid changes in the radiographs, show changes in the entire lung field in minute detail, or show changes in multiple parts depending on the type. It is therefore difficult to have an accurate grasp of the condition of the disease only with conventional radiographs. The temporal subtraction technique which was developed at the University of Chicago, provides a subtraction image of the current warped image and the previous image. A temporal subtraction image, shows only differences and changes between the two images, can be very useful for a diagnosis of interstitial lung disease. However, the evaluation of the temporal subtraction technique for interstitial lung disease using receiver operating characteristic(ROC) studies has not been reported yet. Therefore, we have evaluated the clinical usefulness of a temporal subtraction technique for detection of interval changes of interstitial lung disease by ROC analysis.

• Investigative Magnetic Resonance Imaging
• /
• 제26권1호
• /
• pp.32-42
• /
• 2022

Purpose: To evaluate the clinical benefit of 2D contrast-enhanced T2 fluid-attenuated inversion recovery (CE-T2 FLAIR) image for detecting leptomeningeal metastasis (LM) in the brain metastasis work-up for lung cancer. Materials and Methods: From June 2017 to July 2019, we collected all consecutive patients with lung cancer who underwent brain magnetic resonance image (MRI), including contrast-enhanced 3D fast spin echo T1 black-blood image (CE-T1WI) and CE-T2 FLAIR; we recruited clinico-radiologically suspected LM cases. Two independent readers analyzed the images for LM in three sessions: CE-T1WI, CE-T2 FLAIR, and their combination. Results: We recruited 526 patients with suspected lung cancer who underwent brain MRI; of these, we excluded 77 (insufficient image protocol, unclear pathology, different contrast media, poor image quality). Of the 449 patients, 34 were clinico-radiologically suspected to have LM; among them, 23 were diagnosed with true LM. The calculated detection performance of CE-T1WI, CE-T2 FLAIR, and combined analysis obtained from the 34 suspected LM were highest in the combined analysis (AUC: 0.80, 0.82, and 0.89, respectively). The inter-observer agreement was also the highest in the combined analysis (0.68, 0.72, and 0.86, respectively). In quantitative analyses, CNR of CE-T2 FLAIR was significantly higher than that of CE-T1WI (Wilcoxon signed rank test, P < 0.05). Conclusion: Adding CE-T2 FLAIR might provide better detection for LM in the brain-metastasis screening for lung cancer.

• 대한방사선기술학회지:방사선기술과학
• /
• 제41권1호
• /
• pp.1-6
• /
• 2018

The purpose of this study is to provide baseline data on lung field size measured radiological method by chest PA image in normal Korean. The subject of this study is 496 normal persons who performed chest PA examination using x-ray digital radiography system. The measurement method is from the apex of right and left lung to the costophrenic angle of both lung, from the top of the image to the lowest costophrenic angle of both lung and transverse line of the largest lung area. As a result of this study, the following conclusions were obtained. A lung field size of male is larger than the female(p<0.05). The younger the age, the longer both lung length and total lung height statistically significant. As a increase height and length, A lung field size was increased(p<0.05). But, BMI is not associated with a lung field size. This study will be data of reference data when radiological technologists perform chest PA examination.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권3호
• /
• pp.1104-1120
• /
• 2020

The chest X-rays are a common way to diagnose lung cancer or pneumonia. In particular, the finding of a lung nodule is the most important problem in the early detection of lung cancer. Recently, a lot of automatic diagnosis algorithms have been studied to find the lung nodules missed by doctors. The algorithms are typically based on segmentation network like U-Net. However, the occurrence of false positives that similar to lung nodules present outside the lungs can severely degrade performance. In this study, we propose a multi-task learning method that simultaneously learns the lung region and nodule-labeled data based on the prior knowledge that lung nodules exist only in the lung. The proposed method significantly reduces false positives outside the lung and improves the recognition rate of lung nodules to 83.8 F1 score compared to 66.6 F1 score of single task learning with U-net model. The experimental results on the JSRT public dataset demonstrate the effectiveness of the proposed method compared with other baseline methods.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권1호
• /
• pp.68-80
• /
• 2013

Lung cancer is considered to be the leading cause of cancer death worldwide. A technique commonly used consists of analyzing sputum images for detecting lung cancer cells. However, the analysis of sputum is time consuming and requires highly trained personnel to avoid errors. The manual screening of sputum samples has to be improved by using image processing techniques. In this paper we present a Computer Aided Diagnosis (CAD) system for early detection and diagnosis of lung cancer based on the analysis of the sputum color image with the aim to attain a high accuracy rate and to reduce the time consumed to analyze such sputum samples. In order to form general diagnostic rules, we present a framework for segmentation and extraction of sputum cells in sputum images using respectively, a Bayesian classification method followed by region detection and feature extraction techniques to determine the shape of the nuclei inside the sputum cells. The final results will be used for a (CAD) system for early detection of lung cancer. We analyzed the performance of a Bayesian classification with respect to the color space representation and quantification. Our methods were validated via a series of experimentation conducted with a data set of 100 images. Our evaluation criteria were based on sensitivity, specificity and accuracy.