• Proceedings of the Korean Society of Medical Physics Conference
• /
• 2005.04a
• /
• pp.71-74
• /
• 2005

To introduce the four-dimensional computed tomography (4DCT, Light Speed RT, General Electric, USA) scanner newly installed in our department and evaluate its feasibility for gated radiotherapy. Respiratory signal measured by real-time position management (RPM$^{\circledR}$, Varian Medical, USA) was recorded in synchronization with the 4DCT scanner. 4DCT data were acquired in axial cine mode and sorted retrospective image based on respiratory phase. PTVs delineated from helical CT and 4DCT images were compared. The PTV delineated from conventional helical CT images was 2 cc larger than that from 4DCT images. Dose in PTV of the plan from retrospective CT was 99.3% (minimum=72.0%, maximum=106.5%) and that of helical CT plan was 95.2% (minimum=24.1%, maximum=106.4%) of prescribed dose. Comparing with DVHs of both plan, the coverage for 4CDT plan was 3.7% improved. It is expected that 4DCT could improve tumor control and reduce radiation toxicity for liver cancer.

• Journal of radiological science and technology
• /
• v.43 no.2
• /
• pp.71-77
• /
• 2020

This study aims to explain the process of providing important medical images for the diagnosis of pneumonia caused by coronavirus disease 2019 (COVID-19) through the only mobile hospital computed tomography (CT) in Korea. Since January 28, 2020, medical imaging examinations have been provided to confirmed and suspected COVID-19 patients, and the quality of images was evaluated based on the objective and subjective indicators. In order to prevent the transmission in the hospital that may occur due to exposure to medical staff and general patients, personal protective equipment was worn and the separate route was used blocking human infection factors. For 11 weeks, a total of 185 tests were performed for 98 confirmed patients and 72 suspected patients. The average time to complete the test was 33 minutes. In the course of the test, no cross-infection cases were examined. During the outbreak of the COVID-19, the only mobile hospital CT room of Korea provided medical imaging examinations without infection among medical staff and patients and also provided adequate medical images without significant difference (p >0.05) in determining the degree of pneumonia progression compared to a stationary in-hospital CT.

• Korean Journal of Radiology
• /
• v.23 no.10
• /
• pp.1009-1018
• /
• 2022

Objective: This study aimed to investigate the feasibility of using artificial intelligence (AI) to identify normal chest radiography (CXR) from the worklist of radiologists in a health-screening environment. Materials and Methods: This retrospective simulation study was conducted using the CXRs of 5887 adults (mean age ± standard deviation, 55.4 ± 11.8 years; male, 4329) from three health screening centers in South Korea using a commercial AI (Lunit INSIGHT CXR3, version 3.5.8.8). Three board-certified thoracic radiologists reviewed CXR images for referable thoracic abnormalities and grouped the images into those with visible referable abnormalities (identified as abnormal by at least one reader) and those with clearly visible referable abnormalities (identified as abnormal by at least two readers). With AI-based simulated exclusion of normal CXR images, the percentages of normal images sorted and abnormal images erroneously removed were analyzed. Additionally, in a random subsample of 480 patients, the ability to identify visible referable abnormalities was compared among AI-unassisted reading (i.e., all images read by human readers without AI), AI-assisted reading (i.e., all images read by human readers with AI assistance as concurrent readers), and reading with AI triage (i.e., human reading of only those rendered abnormal by AI). Results: Of 5887 CXR images, 405 (6.9%) and 227 (3.9%) contained visible and clearly visible abnormalities, respectively. With AI-based triage, 42.9% (2354/5482) of normal CXR images were removed at the cost of erroneous removal of 3.5% (14/405) and 1.8% (4/227) of CXR images with visible and clearly visible abnormalities, respectively. In the diagnostic performance study, AI triage removed 41.6% (188/452) of normal images from the worklist without missing visible abnormalities and increased the specificity for some readers without decreasing sensitivity. Conclusion: This study suggests the feasibility of sorting and removing normal CXRs using AI with a tailored cut-off to increase efficiency and reduce the workload of radiologists.

• Journal of Institute of Convergence Technology
• /
• v.11 no.1
• /
• pp.1-5
• /
• 2021

Deep learning based methods achieve state-of-the-art accuracy, however, they typically rely on supervised training with large labeled datasets. It is known in many medical applications that labeling medical images requires significant expertise and much time, and typical hand-tuned approaches for data augmentation fail to capture the complex variations in such images. This paper proposes a 3D image augmentation method to overcome these difficulties. It allows us to enrich diversity of training data samples that is essential in medical image segmentation tasks, thus reducing the data overfitting problem caused by the fact the scale of medical image dataset is typically smaller. Our numerical experiments demonstrate that the proposed approach provides significant improvements over state-of-the-art methods for 3D medical image segmentation.

• Proceedings of the Korean Society of Medical Physics Conference
• /
• 2004.11a
• /
• pp.96-99
• /
• 2004

The objective of this study is to assess attenuation correction algorithms utilized in a multipurpose whole-body GSO PET scanner. Four different types of phantoms were tested using different types of attenuation correction techniques. FOV (Field of View) of 256mm was used for brain PET imaging. For compensating attenuation, transmission data of a $^{137}$Cs point source were acquired after the F-18 emission source was infused to the phantoms. Scatter correction were peformed. Reconstructed images of the phantoms were assessed. In addition, reconstructed images of a normal subject were compared and assessed by nuclear medicine physicians. As a result, decreased intensity at the central portion of the attenuation map with cylindrical phantom was noticed during use of the measured attenuation correction. On the other hand, segmentation or remapping attenuation correction provided uniform phantom image. the images reconstructed from the clinical brain data explained the attenuation of a skull, at though reconstructed images of the phantoms couldn't explain it. in conclusion, the complicated and improved attenuation correction methods were required to obtain the better accuracy of the quantitative brain PET images. Our study will be useful in improving quantitative brain PET imaging modalities with attenuation correction of $^{137}$Cs transmission source.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.568-571
• /
• 1997

Recently, the number of techniques for analyzing medical images has been increasing in computer vision, employing X-ray CT images, ultrasound images, MR images, moire topographic images, etc. Spinal deformity is a serious problem especially for teenagers and medical doctors inspect moire topographic images of their backs visually for the primary screening. If a subject is normal, the moire image is almost symmetric with respect to the middle line of the subject's back, otherwise it shows asymmetric shape. In this paper, an image analysis technique is described for discriminating suspicious cases from normal in human spinal deformity by recognizing asymmetric moire images of human backs. The principal axes which are sensitive to asymmetry of the moire image are extracted at two parts on a subject's back and their angles are evaluated with respect to the detected middle line of the back. The two angles compose a 2-D feature space and inspected cases are divided into two clusters in the space by a linear discriminant function based on the Mahalanobis distance. Given 120 cases, 60 normal and 60 abnormal, the leave-out method was applied for the recognition and 75% recognition rate was achieved.

• Journal of the Korean Society of Radiology
• /
• v.8 no.1
• /
• pp.27-33
• /
• 2014

Medical imaging has increased rapidly in the increase of interest in health, with the development of computer technology, digitization of medical imaging is rapidly advancing, PACS has been introduced to the medical field. Increase in the production of medical images by these phenomena made increased the workload of radiologist who must read a medical image. in response to the need for secondary diagnosis using a computer, The term of CAD in medical radiology field was introduced. In this study, we have proposed a CAD algorithm for the interpretation of the image obtained by the digital X-ray mammography equipment. The experiments were performed by programmed in Visual C++ for the proposed algorithm. A result of the execution of the CAD algorithm seven sample images, the results of five samples was confirmed in breast cancer and benign tumors, both the images sample was error processing. If you use a program that implements this with the algorithm proposed in this study it is helpful to reading breast images, and it is considered to contribute significantly to the early detection of breast cancer.

• Journal of Biomedical Engineering Research
• /
• v.25 no.3
• /
• pp.183-188
• /
• 2004

Magnetic resonance current density imaging (MRCDI) is a useful method for measuring electrical current density distribution inside an object. To avoid object rotations during the conventional MRCDI scans, we have reconstructed current density component images by applying a spatial filter to the magnetic field data measured both inside and outside the object. To measure the magnetic field outside the object with MRI, we immersed the object in a water tank. To evaluate accuracy of the current density imaging, we have made a conductivity phantom with a corresponding finite element method model. We have compared the experimentally obtained current density images with the ones calculated by the finite element method. The average errors of the reconstructed current density images were 6.6 ∼ 45.4 ％ when the injected currents were 1 ∼ 24 mA. We expect that the current density component imaging technique can be used in diverse biomedical applications such as electrical therapy system developments and biological electrical safety analysis.

• Journal of Korea Multimedia Society
• /
• v.15 no.5
• /
• pp.569-576
• /
• 2012

This paper presents an x-ray medical image panorama system which can overcome the smallness of the images that exist on a source computer during remote medical processing. In the system, after the standard medical image format DICOM is converted to the PC standard image format, a MSR algorithm is used to enhance X-ray images of low quality. Then SURF and Multi-band blending are applied to generate a panoramic image. Also, this paper evaluates the proposed SURF based system through the average gray value error and image quality criterion with X-ray image data by comparing with a SIFT based system. The results show that the proposed system is superior to SIFT based system in image quality.

• Proceedings of the KOSOMBE Conference
• /
• v.1995 no.11
• /
• pp.113-116
• /
• 1995

Medical image compression using the wavelet transform has been tried. Due to the flexibility in representing nonstationary image signal in both time and frequency domains and its ability to adapt human visual characteristics, wavelet transform has unique advantage in images compression. In the proposed wavelet compression original image is decomposed into multi-scale bands. Different scale factors are employed in the quantization of wavelet decomposed images in different bands. For the lowest band, a predictor is designed and error signal is entropy coded. For high scale bands, runlength coding for toro run is used with Huffman coding. From simulation with magnetic resonance images($256\times256$ size, 256 graylevels) the proposed algorithm is superior to the JPEG by more than 2.5 dB in near lossless compression (CR = 8 - 10).