JPEG2000 3D compression vs 2D compression: An assessment of artifact amount and computing time in compressing thin-section abdomen CT images : JPEG2000 3D compression vs 2D compression

Kim, Bohyoung,Lee, Kyoung Ho,Kim, Kil Joong,Richter, Thomas,Kang, Heung-Sik,Kim, So Yeon,Kim, Young Hoon,Seo, Jinwook Wiley (John WileySons) 2009 Medical physics Vol.36 No.3

<P>To assess the advantages of the Joint Photographic Experts Group (JPEG)2000 3D (part 2) over JPEG2000 in compressing thin-section abdomen CT data sets, 60 thin-section (0.67 mm) scans from 35 males and 25 females, ranging from 23 to 95 years of age (mean, 58 years), were compressed reversibly (as a negative control) and irreversibly to 4:1, 6:1, 8:1, 10:1, and 12:1 using JPEG2000 3D and JPEG2000 algorithms. Encoding and decoding times and peak signal-to-noise ratios (PSNRs) were measured. For 60 (one image per scan) representative sections containing abnormalities, three radiologists independently compared original and compressed images and graded compression artifacts as 0 (none, indistinguishable), 1 (barely perceptible), 2 (subtle), or 3 (significant). According to pooled radiologists' responses, the range of visually lossless threshold (VLT, the highest compression ratio at which a compressed image is indistinguishable from its original) was determined as one of <4:1, 4:1-6:1, 6:1-8:1, 8:1-10:1, 10:1-12:1, and >12:1. Wilcoxon signed rank tests and exact tests for paired proportions were used for the comparisons between the two compressions. At each irreversible compression ratio, compared to JPEG2000, JPEG2000 3D required two- or threefold greater computing times (p < 0.001) and introduced less artifacts in terms of PSNR (p <0.001) and the grade (p < 0.02 at 6:1 or higher) and the presence of perceived artifacts (p <0.008, at 6:1 for all readers and at 8:1 for two readers). According to PSNR and readers' responses, 6:1 and 8:1 JPEG2000 3D compressions showed more artifacts than 4:1 and 6:1 JPEG2000 compressions, respectively, and 10:1 and 12:1 JPEG2000 3D compressions showed similar artifacts to those of 8:1 and 10:1 JPEG2000 compressions, respectively. The determined VLT range was higher for JPEG2000 3D than for JPEG2000 (p < 0.001): the 3D compression showed the VLT ranges of 4:1-6:1, 6:1-8:1, and 8:1-10:1 for 24 (40%), 30 (50%), and 6 (10%) of the 60 original images, respectively, while the 2D compression showed the VLT ranges of <4:1, 4:1-6:1, and 6:1-8:1 for 1 (1.7%), 40 (66.7%), and 19 (31.6%) images, respectively. Compared to JPEG2000, JPEG2000 3D increased the VLT range in 23 of the 60 original images by one (n=22) or two ranges (n=1), while the remaining 37 images had the same VLT range between the two compressions. In conclusion, compared to JPEG2000 compression, JPEG2000 3D compression yields less artifacts in compressing thin-section abdomen CT images but requires significantly greater computing times. For the tested data set compressed to the range from 4:1 to 12:1, JPEG2000 3D could increase compression level reasonably (by 2 or less in terms of compression ratio) compared to JPEG2000 for the same amount of artifacts.</P>

Artifacts in slab average-intensity-projection images reformatted from JPEG 2000 compressed thin-section abdominal CT data sets.

Kim, Bohyoung,Lee, Kyoung Ho,Kim, Kil Joong,Mantiuk, Rafal,Kim, Hye-ri,Kim, Young Hoon American Roentgen Ray Society, etc.] 2008 American Journal of Roentgenology Vol.190 No.6

<P>OBJECTIVE: The objective of our study was to assess the effects of compressing source thin-section abdominal CT images on final transverse average-intensity-projection (AIP) images. MATERIALS AND METHODS: At reversible, 4:1, 6:1, 8:1, 10:1, and 15:1 Joint Photographic Experts Group (JPEG) 2000 compressions, we compared the artifacts in 20 matching compressed thin sections (0.67 mm), compressed thick sections (5 mm), and AIP images (5 mm) reformatted from the compressed thin sections. The artifacts were quantitatively measured with peak signal-to-noise ratio (PSNR) and a perceptual quality metric (High Dynamic Range Visual Difference Predictor [HDR-VDP]). By comparing the compressed and original images, three radiologists independently graded the artifacts as 0 (none, indistinguishable), 1 (barely perceptible), 2 (subtle), or 3 (significant). Friedman tests and exact tests for paired proportions were used. RESULTS: At irreversible compressions, the artifacts tended to increase in the order of AIP, thick-section, and thin-section images in terms of PSNR (p < 0.0001), HDR-VDP (p < 0.0001), and the readers' grading (p < 0.01 at 6:1 or higher compressions). At 6:1 and 8:1, distinguishable pairs (grades 1-3) tended to increase in the order of AIP, thick-section, and thin-section images. Visually lossless threshold for the compression varied between images but decreased in the order of AIP, thick-section, and thin-section images (p < 0.0001). CONCLUSION: Compression artifacts in thin sections are significantly attenuated in AIP images. On the premise that thin sections are typically reviewed using an AIP technique, it is justifiable to compress them to a compression level currently accepted for thick sections.</P>

Comparison of three image comparison methods for the visual assessment of the image fidelity of compressed computed tomography images.

Kim, Bohyoung,Lee, Hyunna,Kim, Kil Joong,Seo, Jinwook,Park, Seongjin,Shin, Yeong-Gil,Kim, Soo Hong,Lee, Kyoung Ho The American Association of Physicists in Medicine 2011 Medical physics Vol.38 No.2

<P>This study aimed to comparatively evaluate three different image comparison methods: alternate display without an intervening blank image (AWOB), alternate display with an intervening blank image (AWB), and side-by-side display (SSD), in terms of the perceptual sensitivity to image differences between Joint Photographic Experts Group 2000 (JPEG2000) compressed body CT images and their originals.</P>

Prediction of perceptible artifacts in JPEG 2000-compressed chest CT images using mathematical and perceptual quality metrics.

Kim, Bohyoung,Lee, Kyoung Ho,Kim, Kil Joong,Mantiuk, Rafal,Hahn, Seokyung,Kim, Tae Jung,Kim, Young Hoon American Roentgen Ray Society, etc.] 2008 American Journal of Roentgenology Vol.190 No.2

<P>OBJECTIVE: The objective of our study was to determine whether peak signal-to-noise ratio (PSNR) and a perceptual quality metric (High-Dynamic Range Visual Difference Predictor [HDR-VDP]) can predict the presence of perceptible artifacts in Joint Photographic Experts Group (JPEG) 2000-compressed chest CT images. MATERIALS AND METHODS: One hundred chest CT images were compressed to 5:1, 8:1, 10:1, and 15:1. Five radiologists determined if the original and compressed images were identical (negative response) or different (positive response). The correlation between the results for each metric and the number of readers with positive responses was evaluated using Spearman's rank correlation test. Using the pooled readers' responses as the reference standard, we performed receiver operating characteristic (ROC) analysis to determine the cutoff values balancing sensitivity and specificity and yielding 100% sensitivity in each metric. These cutoff values were then used to estimate the visually lossless thresholds for the compressions for the 100 original images, and the accuracy of the estimates of two metrics was compared (McNemar test). RESULTS: The correlation coefficients were -0.918 and 0.925 for PSNR and the HDR-VDP, respectively. The areas under the ROC curves for the two metrics were 0.983 and 0.984, respectively (p = 0.11). The PSNR and HDR-VDP accurately predicted the visually lossless threshold for 69% and 72% of the 100 images (p = 0.68), respectively, at the cutoff values balancing sensitivity and specificity and for 43% and 47% (p = 0.22), respectively, at the cutoff values reaching 100% sensitivity. CONCLUSION: Both metrics are promising in predicting the perceptible compression artifacts and therefore can potentially be used to estimate the visually lossless threshold.</P>

Detection of the normal appendix with low-dose unenhanced CT: use of the sliding slab averaging technique.

Joo, Seung-Moon,Lee, Kyoung Ho,Kim, Young Hoon,Kim, So Yeon,Kim, Kyuseok,Kim, Kil Joong,Kim, Bohyoung Radiological Society of North America 2009 Radiology Vol.251 No.3

<P>PURPOSE: To determine the frequency of normal appendix visualization at low-dose (LD) unenhanced computed tomography (CT) performed with a 16- or 64-detector row scanner when images are reviewed by using the sliding slab averaging technique. MATERIALS AND METHODS: The institutional review board approved the study and waived the informed consent requirement. A total of 259 patients, 37 (14.3%) of whom had previously undergone appendectomy, underwent LD unenhanced CT (mean effective dose, 1.7 mSv) performed with a 16- or 64-detector row scanner to assess urinary colic. Three readers used the sliding slab averaging technique to retrospectively review the thin-section (0.67- or 2.00-mm section thickness) images and grade the appendix as absent, unsurely or partly visualized, or clearly and entirely visualized. Interobserver agreement was measured with weighted kappa statistics. McNemar tests were used to compare sensitivity between the readers. Logistic regression analysis was performed to assess the effects of body mass index, patient sex, and type of CT scanner on appendiceal visualization. RESULTS: The kappa statistics for each reader pair were as follows: 0.97 for agreement between readers 1 and 2, 0.93 for agreement between readers 2 and 3, and 0.92 for agreement between readers 1 and 3. Each reader clearly identified the entire appendix in 213 (96.0%), 209 (94.1%), and 205 (92.3%) of the 222 patients without a history of appendectomy. When unsurely or partly visualized appendices were included, the frequencies increased to 99.1% (n = 220), 98.7% (n = 219), and 97.3% (n = 216), respectively, for readers 1, 2, and 3. These frequencies rarely differed between the readers. (P values ranged from .021 to greater than .99.) The three readers consistently reported that the appendix was not visualized in the 37 patients who had undergone appendectomy. None of the tested variables significantly affected appendix visualization. CONCLUSION: Most normal appendices are visualized on thin-section LD unenhanced CT images reviewed with the sliding slab averaging technique.</P>

Automatic Extraction of Inferior Alveolar Nerve Canal Using Feature-Enhancing Panoramic Volume Rendering

Kim, Gyehyun,Lee, Jeongjin,Lee, Ho,Seo, Jinwook,Koo, Yun-Mo,Shin, Yeong-Gil,Kim, Bohyoung IEEE 2011 IEEE Transactions on Biomedical Engineering Vol.58 No.2

<P>Dental implant surgery, which involves the surgical insertion of a dental implant into the jawbone as an artificial root, has become one of the most successful applications of computed tomography (CT) in dental implantology. For successful implant surgery, it is essential to identify vital anatomic structures such as the inferior alveolar nerve (IAN), which should be avoided during the surgical procedure. Due to the ambiguity of its structure, the IAN is very elusive to extract in dental CT images. As a result, the IAN canal is typically identified in most previous studies. This paper presents a novel method of automatically extracting the IAN canal. Mental and mandibular foramens, which are regarded as the ends of the IAN canal in the mandible, are detected automatically using 3-D panoramic volume rendering (VR) and texture analysis techniques. In the 3-D panoramic VR, novel color shading and compositing methods are proposed to emphasize the foramens and isolate them from other fine structures. Subsequently, the path of the IAN canal is computed using a line-tracking algorithm. Finally, the IAN canal is extracted by expanding the region of the path using a fast marching method with a new speed function exploiting the anatomical information about the canal radius. In experimental results using ten clinical datasets, the proposed method identified the IAN canal accurately, demonstrating that this approach assists dentists substantially during dental implant surgery.</P>

Automatic Teeth Axes Calculation for Well-Aligned Teeth Using Cost Profile Analysis Along Teeth Center Arch

Kim, Gyehyun,Lee, Jeongjin,Seo, Jinwook,Lee, Wooshik,Shin, Yeong-Gil,Kim, Bohyoung IEEE 2012 IEEE Transactions on Biomedical Engineering Vol.59 No.4

<P>In dental implantology and virtual dental surgery planning using computed tomography (CT) images, the examination of the axes of neighboring and/or biting teeth is important to improve the performance of the masticatory system as well as the aesthetic beauty. However, due to its high connectivity to neighboring teeth and jawbones, a tooth and/or its axis is very elusive to automatically identify in dental CT images. This paper presents a novel method of automatically calculating individual teeth axes. The planes separating the individual teeth are automatically calculated using cost profile analysis along the teeth center arch. In this calculation, a novel plane cost function, which considers the intensity and the gradient, is proposed to favor the teeth separation planes crossing the teeth interstice and suppress the possible inappropriately detected separation planes crossing the soft pulp. The soft pulp and dentine of each individually separated tooth are then segmented by a fast marching method with two newly proposed speed functions considering their own specific anatomical characteristics. The axis of each tooth is finally calculated using principal component analysis on the segmented soft pulp and dentine. In experimental results using 20 clinical datasets, the average angle and minimum distance differences between the teeth axes manually specified by two dentists and automatically calculated by the proposed method were 1.94° ± 0.61° and 1.13 ± 0.56 mm, respectively. The proposed method identified the individual teeth axes accurately, demonstrating that it can give dentists substantial assistance during dental surgery such as dental implant placement and orthognathic surgery.</P>

Use of Image Features in Predicting Visually Lossless Thresholds of JPEG2000 Compressed Body CT Images: Initial Trial

Kim, Kil Joong,Kim, Bohyoung,Lee, Kyoung Ho,Mantiuk, Rafal,Richter, Thomas,Kang, Heung Sik RSNA 2013 Radiology Vol.268 No.3

<P>Among the five tested image features, variation in high-frequency energy and visual complexity were promising in predicting the visually lossless thresholds of body CT images for JPEG2000 compression.</P>

Automated Versus Handheld Breast Ultrasound for Evaluating Axillary Lymph Nodes in Patients With Breast Cancer

Kim Sun Mi,Jang Mijung,Yun Bo La,Shin Sung Ui,Rim Jiwon,Kang Eunyoung,Kim Eun-Kyu,Shin Hee-Chul,Park So Yeon,Kim Bohyoung 대한영상의학회 2024 Korean Journal of Radiology Vol.25 No.2

Objective: Automated breast ultrasound (ABUS) is a relevant imaging technique for early breast cancer diagnosis and is increasingly being used as a supplementary tool for mammography. This study compared the performance of ABUS and handheld ultrasound (HHUS) in detecting and characterizing the axillary lymph nodes (LNs) in patients with breast cancer. Materials and Methods: We retrospectively reviewed the medical records of women with recently diagnosed early breast cancer (≤ T2) who underwent both ABUS and HHUS examinations for axilla (September 2017–May 2018). ABUS and HHUS findings were compared using pathological outcomes as reference standards. Diagnostic performance in predicting any axillary LN metastasis and heavy nodal-burden metastases (i.e., ≥ 3 LNs) was evaluated. The ABUS-HHUS agreement for visibility and US findings was calculated. Results: The study included 377 women (53.1 ± 11.1 years). Among 385 breast cancers in 377 patients, 101 had axillary LN metastases and 30 had heavy nodal burden metastases. ABUS identified benign-looking or suspicious axillary LNs (average, 1.4 ± 0.8) in 246 axillae (63.9%, 246/385). According to the per-breast analysis, the sensitivity, specificity, positive and negative predictive values, and accuracy of ABUS in predicting axillary LN metastases were 43.6% (44/101), 95.1% (270/284), 75.9% (44/58), 82.6% (270/327), and 81.6% (314/385), respectively. The corresponding results for HHUS were 41.6% (42/101), 95.1% (270/284), 75.0% (42/56), 82.1% (270/329), and 81.0% (312/385), respectively, which were not significantly different from those of ABUS (P ≥ 0.53). The performance results for heavy nodal-burden metastases were 70.0% (21/30), 89.6% (318/355), 36.2% (21/58), 97.3% (318/327), and 88.1% (339/385), respectively, for ABUS and 66.7% (20/30), 89.9% (319/355), 35.7% (20/56), 97.0% (319/329), and 88.1% (339/385), respectively, for HHUS, also not showing significant difference (P ≥ 0.57). The ABUS–HHUS agreement was 95.9% (236/246; Cohen’s kappa = 0.883). Conclusion: Although ABUS showed limited sensitivity in diagnosing axillary LN metastasis in early breast cancer, it was still useful as the performance was comparable to that of HHUS.

딥러닝/머신러닝 모델을 활용한 뇌파 기반 치매 조기진단에 관한 연구

김재영(Jaeyoung Kim),한구현(GuHyun Han),안단태(Dantae An),이시현(Sihyeon Lee),김진욱(Jinwook Kim),임승의(Seungeui Lim),문기욱(Kiwook Moon),이기원(Kiwon Lee),김보형(Bohyoung Kim),서진욱(Jinwook Seo) 한국정보과학회 2022 한국정보과학회 학술발표논문집 Vol.2022 No.12