• Korean Journal of Remote Sensing
• /
• v.39 no.5_3
• /
• pp.979-995
• /
• 2023

As wildfires are difficult to predict, real-time monitoring is crucial for a timely response. Geostationary satellite images are very useful for active fire detection because they can monitor a vast area with high temporal resolution (e.g., 2 min). Existing satellite-based active fire detection algorithms detect thermal outliers using threshold values based on the statistical analysis of brightness temperature. However, the difficulty in establishing suitable thresholds for such threshold-based methods hinders their ability to detect fires with low intensity and achieve generalized performance. In light of these challenges, machine learning has emerged as a potential-solution. Until now, relatively simple techniques such as random forest, Vanilla convolutional neural network (CNN), and U-net have been applied for active fire detection. Therefore, this study proposed an active fire detection algorithm using state-of-the-art (SOTA) deep learning techniques using data from the Advanced Himawari Imager and evaluated it over East Asia and Australia. The SOTA model was developed by applying EfficientNet and lion optimizer, and the results were compared with the model using the Vanilla CNN structure. EfficientNet outperformed CNN with F1-scores of 0.88 and 0.83 in East Asia and Australia, respectively. The performance was better after using weighted loss, equal sampling, and image augmentation techniques to fix data imbalance issues compared to before the techniques were used, resulting in F1-scores of 0.92 in East Asia and 0.84 in Australia. It is anticipated that timely responses facilitated by the SOTA deep learning-based approach for active fire detection will effectively mitigate the damage caused by wildfires.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.13 no.3
• /
• pp.102-109
• /
• 2009

Purpose: Flash 3D (pixon(R) method; 3D OSEM) was developed as a software program to shorten exam time and improve image quality through reconstruction, it is an image processing method that usefully be applied to nuclear medicine tomography. If perfoming brain diamox perfusion scan by reconstructing subtracted images by Flash 3D with shortened image acquisition time, there was a problem that SNR of subtracted image is lower than basal image. To increase SNR of subtracted image, we use LEAP collimators, and we emphasized on sensitivity of vessel dilatation than resolution of brain vessel. In this study, our purpose is to confirm possibility of application of LEAP collimators at brain diamox perfusion tomography, identify proper reconstruction factors by using Flash 3D. Materials and methods: (1) The evaluation of phantom: We used Hoffman 3D Brain Phantom with $^{99m}Tc$. We obtained images by LEAP and LEHR collimators (diamox image) and after 6 hours (the half life of $^{99m}Tc$: 6 hours), we use obtained second image (basal image) by same method. Also, we acquired SNR and ratio of white matters/gray matters of each basal image and subtracted image. (2) The evaluation of patient's image: We quantitatively analyzed patients who were examined by LEAP collimators then was classified as a normal group and who were examined by LEHR collimators then was classified as a normal group from 2008. 05 to 2009. 01. We evaluate the results from phantom by substituting factors. We used one-day protocol and injected $^{99m}Tc$-ECD 925 MBq at both basal image acquisition and diamox image acquisition. Results: (1) The evaluation of phantom: After measuring counts from each detector, at basal image 41~46 kcount, stress image 79~90 kcount, subtraction image 40~47 kcount were detected. LEAP was about 102~113 kcount at basal image, 188~210 kcount at stress image and 94~103 at subtraction image kcount were detected. The SNR of LEHR subtraction image was decreased than LEHR basal image about 37%, the SNR of LEAP subtraction image was decreased than LEAP basal image about 17%. The ratio of gray matter versus white matter is 2.2:1 at LEHR basal image and 1.9:1 at subtraction, and at LEAP basal image was 2.4:1 and subtraction image was 2:1. (2) The evaluation of patient's image: the counts acquired by LEHR collimators are about 40~60 kcounts at basal image, and 80~100 kcount at stress image. It was proper to set FWHM as 7 mm at basal and stress image and 11mm at subtraction image. LEAP was about 80~100 kcount at basal image and 180~200 kcount at stress image. LEAP images could reduce blurring by setting FWHM as 5 mm at basal and stress images and 7 mm at subtraction image. At basal and stress image, LEHR image was superior than LEAP image. But in case of subtraction image like a phantom experiment, it showed rough image because SNR of LEHR image was decreased. On the other hand, in case of subtraction LEAP image was better than LEHR image in SNR and sensitivity. In all LEHR and LEAP collimator images, proper subset and iteration frequency was 8 times. Conclusions: We could archive more clear and high SNR subtraction image by using proper filter with LEAP collimator. In case of applying one day protocol and reconstructing by Flash 3D, we could consider application of LEAP collimator to acquire better subtraction image.

• The Korean Journal of Nuclear Medicine
• /
• v.33 no.3
• /
• pp.327-336
• /
• 1999

Purpose: To assess the quantitative accuracy and the clinical utility of 3D volumetric PET imaging with FDG in brain studies, 24 patients with various neurological disorders were studied. Materials and Methods: Each patient was injected with 370 MBq of 2-[$^{18}F$]fluoro-2-deoxy-D-glucose. After a 30 min uptake period, the patients were imaged for 30 min in 2 dimensional acquisition (2D) and subsequently for 10 min in 3 dimensional acquisition imaging (3D) using a GE $Advance^{TM}$ PET system, The scatter corrected 3D (3D SC) and non scatter-corrected 3D images were compared with 2D images by applying ROIs on gray and white matter, lesion and contralateral normal areas. Measured and calculated attenuation correction methods for emission images were compared to get the maximum advantage of high sensitivity of 3D acquisition. Results: When normalized to the contrast of 2D images, the contrasts of gray to white matter were $0.75{\pm}0.13$ (3D) and $0.95{\pm}0.12$ (3D SC). The contrasts of normal area to lesion were $0.83{\pm}0.05$ (3D) and $0.96{\pm}0.05$ (3D SC). Three nuclear medicine physicians judged 3D SC images to be superior to the 2D with regards to resolution and noise. Regional counts of calculated attenuation correction was not significantly different to that of measured attenuation correction. Conclusion: 3D PET images with the scatter correction in FDG brain studies provide quantitatively and qualitatively similar images to 2D and can be utilized in a routine clinical setting to reduce scanning time and patient motion artifacts.

• Journal of Yeungnam Medical Science
• /
• v.6 no.1
• /
• pp.47-57
• /
• 1989

T. verrucosum infection has been reported for the first time in 1986 in Korea and has been increasing progressively. To evaluate the progress of clinical and histopathological change of dermatophytosis caused by T. verrucosum, inoculation study, using T. verrucosum isolated from infected human(human strain) and from infected cattle(cattle strain), was done in 24 male albino Hartley guinea pig. Their clinical and histopathological changes were evaluated. In addition, comparison for the growth rate between human strain and cattle strain on Sabouraud's glucose agar was made. The results were as follows: 1. Growth rate on Sabouraud's glucose agar : Cattle strain showed significantly more rapid growth rate than human strain on Sabouraud's glucose rate at $25^{\circ}C$ and $37^{\circ}C$. And cattle strain showed more rapid growth rate at $37^{\circ}C$ than $25^{\circ}C$. But human strain showed no significant difference of growth rate at both temperature. 2. Clinical findings: Initial erythema, scale and crust were developed about 8th after inoculation. All three findings reached maximum severity about 12th to 16th day and disappeared about 30th to 34th day after inoculation. There was no significant difference in progress of erythema, scale and crust between cattle strain and human strain. 3. Histopathological findings: Although mild acanthosis was noticed on the 3rd day after inoculation, the other findings including parakeratosis, intraepidermal abscess, spongiosis and vascular change, cellular infiltration were found on 9th day after inoculation. They reached maximum severity on the 12th day and lasted to the 25th day after inoculation. After that, all three findings were decreased gradually between 29th day and 33th day. On the PAS stainings, hyphae and spores were found on the 6th day and disappeared on the 21th day after inoculation. 4. In trichophytin skin test, all of the 24 guinea pigs became positive within average $9.83{\pm}1.17$ lays. These findings suggested that dermatophytosis caused by T. verrucosum induced rapid cell mediated immunity and contributed to rapid resolution of the lesion.

• Journal of the Korean Arthroscopy Society
• /
• v.15 no.2
• /
• pp.92-98
• /
• 2011

Purpose: We describe a all-arthroscopic technique for decompression of spinoglenoid ganglion cyst and present our clinical results for this procedure. Materials and Methods: From March 2006 to June 2009, eight patients (7 males, 1 female; mean age 40.6 years; range: 21~61) were included who underwent arthroscopic decompression of spinoglenoid ganglion cyst. The most common symptoms were vague shoulder pain and external rotation weakness, which lasted 6.4 months (range: 3~8) on average. Five of eight patients were noted abnormalities on electromyographic (EMG) examination to have suprascapular nerve neuropathy and magnetic resonance imaging (MRI) showed spinoglenoid ganglion cysts in all eight patients and the cyst size was 2.6 cm (range: 1.8~3.6). Labral pathology was identified intraoperatively in all patients and the cysts were decompressed by the posterosuperior capsulotomy under arthroscopic control and labral repair with suture anchors was performed in six patients. Results: The average clinical outcome scores including Constant and Murley, Simple shoulder test (SST) all improved significantly at the time of the final follow-up and there were no complications resulting from the procedures. All patients including the patients with abnormal EMG study recovered strength on isokinetic strength testing. Follow-up MRI scans were performed on all patients at a mean of 5.2 months postoperatively (range: 3~12) revealed complete resolution of the cysts and no evidence of recurrences were seen at an average of 18 months (range: 12~26) of follow-up. Conclusion: Arthroscopic decompression of spinoglenoid ganglion cyst effectively restores patient function and all patients in this study showed improvement in their postoperative MRI findings. Arthroscopic decompression is also useful in the appropriate treatment for labral pathology and may contribute to decreased risk of cyst recurrence.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.9 no.1
• /
• pp.1-16
• /
• 2007

The use of information on natural resources is indispensable to most agricultural activities to avoid disasters, to improve input efficiency, and to increase lam income. Most information is prepared and managed at a spatial scale called the "Hydrologic Unit" (HU), which means watershed or small river basin, because virtually every environmental problem can be handled best within a single HU. South Korea consists of 840 such watersheds and, while other watershed-specific information is routinely managed by government organizations, there are none responsible for agricultural weather and climate. A joint research team of Kyung Hee University and the Agriculture, forestry and Fisheries Information Service has begun a 4-year project funded by the Ministry of Agriculture and forestry to establish a watershed-specific agricultural weather information service based on "high definition" digital climate maps (HD-DCMs) utilizing the state of the art geospatial climatological technology. For example, a daily minimum temperature model simulating the thermodynamic nature of cold air with the aid of raster GIS and microwave temperature profiling will quantify effects of cold air drainage on local temperature. By using these techniques and 30-year (1971-2000) synoptic observations, gridded climate data including temperature, solar irradiance, and precipitation will be prepared for each watershed at a 30m spacing. Together with the climatological normals, there will be 3-hourly near-real time meterological mapping using the Korea Meteorological Administration's digital forecasting products which are prepared at a 5 km by 5 km resolution. Resulting HD-DCM database and operational technology will be transferred to local governments, and they will be responsible for routine operations and applications in their region. This paper describes the project in detail and demonstrates some of the interim results.

• Journal of Bio-Environment Control
• /
• v.31 no.4
• /
• pp.384-392
• /
• 2022

The objective of this study was to estimated nitrogen content and chlorophyll using RGB, Hyperspectral sensors to diagnose of nitrogen nutrition in apple tree leaves. Spectral data were acquired through image processing after shooting with high resolution RGB and hyperspectral sensor for two-year-old 'Hongro/M.9' apple. Growth data measured chlorophyll and leaf nitrogen content (LNC) immediately after shooting. The growth model was developed by using regression analysis (simple, multi, partial least squared) with growth data (chlorophyll, LNC) and spectral data (SPAD meter, color vegetation index, wavelength). As a result, chlorophyll and LNC showed a statistically significant difference according to nitrogen fertilizer level regardless of date. Leaf color became pale as the nutrients in the leaf were transferred to the fruit as over time. RGB sensor showed a statistically significant difference at the red wavelength regardless of the date. Also hyperspectral sensor showed a spectral difference depend on nitrogen fertilizer level for non-visible wavelength than visible wavelength at June 10th and July 14th. The estimation model performance of chlorophyll, LNC showed Partial least squared regression using hyperspectral data better than Simple and multiple linear regression using RGB data (Chlorophyll R²: 81%, LNC: 81%). The reason is that hyperspectral sensor has a narrow Full Half at Width Maximum (FWHM) and broad wavelength range (400-1,000 nm), so it is thought that the spectral analysis of crop was possible due to stress cause by nitrogen deficiency. In future study, it is thought that it will contribute to development of high quality and stable fruit production technology by diagnosis model of physiology and pest for all growth stage of tree using hyperspectral imagery.

• Clinical and Experimental Pediatrics
• /
• v.48 no.8
• /
• pp.826-831
• /
• 2005

Purpose : This study is directed to evaluate standardized developmental test performances of infants and children who, as fetuses, had mild isolated cerebral ventriculomegaly diagnosed by ultrasound. Methods : All prenatal sonographic findings from 2001 to 2002 were evaluated. Live isolated mild ventriculomegaly(IMVM) of 10-15 mm were observed in 95 cases(1.1 percent). Standardized developmental testing of 40 cases of IMVM and 36 cases in a comparison group were offered to parents Both groups of children were adjusted to normal antepatum subjects with respect to sex, race, indication for ultrasound and gestational age at the time of ultrasound. Test of cognitive and motor development(Bayley Scales of Infant Development, Second Edition; BSID-II) were administered by developmental examiners. Results : Forty cases and 34 comparison sujects completed the testing. The IMVM and comparison groups were similar with respect to parental age, gestational age, birth weight, familial socioeconomic status. The IMVM subjects scored lower than the comparison group on both the BSID-II, but there was not statistically significant. differences; metal development index(MDI)($92.7{\pm}12.9$ vs $94.7{\pm}14.1$, P=0.47) and psychomotor development index(PDI)($100.3{\pm}14.1$ vs $101.3{\pm}10.7$, P=0.75). Eleven cases(27.5 percent) of IMVM group and five cases(14.7 percent) of the comparison group were developmentally delayed, but most cases in both groups showed mild delays. Resolution or lack of progression, lateral ventricle diameter ${\leq}12mm$ and females were associated with better scores, but there were not statistically significant. Polarity, and head circumference were not related to later development. Conclusion : This study show children with MIVM did not delay performance in the developmental test, but we might suggest a tendency to increase the risk of mild developmental delay.

• Progress in Medical Physics
• /
• v.26 no.1
• /
• pp.42-51
• /
• 2015

In proton therapy, verification of proton dose distribution is important to treat cancer precisely and to enhance patients' safety. To verify proton dose distribution, in a previous study, our team incorporated a vertically-aligned one-dimensional array detection system. We measured 2D prompt-gamma distribution moving the developed detection system in the longitudinal direction and verified similarity between 2D prompt-gamma distribution and 2D proton dose distribution. In the present, we have developed two-dimension prompt-gamma measurement system consisted of a 2D parallel-hole collimator, 2D array-type NaI(Tl) scintillators, and multi-anode PMT (MA-PMT) to measure 2D prompt-gamma distribution in real time. The developed measurement system was tested with $^{22}Na$ (0.511 and 1.275 MeV) and $^{137}Cs$ (0.662 MeV) gamma sources, and the energy resolutions of 0.511, 0.662 and 1.275 MeV were $10.9%{\pm}0.23p%$, $9.8%{\pm}0.18p%$ and $6.4%{\pm}0.24p%$, respectively. Further, the energy resolution of the high gamma energy (3.416 MeV) of double escape peak from Am-Be source was $11.4%{\pm}3.6p%$. To estimate the performance of the developed measurement system, we measured 2D prompt-gamma distribution generated by PMMA phantom irradiated with 45 MeV proton beam of 0.5 nA. As a result of comparing a EBT film result, 2D prompt-gamma distribution measured for $9{\times}10^9$ protons is similar to 2D proton dose distribution. In addition, the 45 MeV estimated beam range by profile distribution of 2D prompt gamma distribution was $17.0{\pm}0.4mm$ and was intimately related with the proton beam range of 17.4 mm.

• Korean Journal of Remote Sensing
• /
• v.31 no.5
• /
• pp.441-448
• /
• 2015

Lack of agricultural information for food supply and demand in Democratic People's republic Korea(DPRK) make people sometimes confused for right and timely decision for policy support. We carried out a study to estimate paddy rice yield in DPRK using MODIS NDVI reflecting rice growth and climate data. Mean of MODIS $NDVI_{max}$ in paddy rice over the country acquired and processed from 2002 to 2014 and accumulated rainfall collected from 27 weather stations in September from 2002 to 2014 were used to estimated paddy rice yield in DPRK. Coefficient of determination of the multiple regression model was 0.44 and Root Mean Square Error(RMSE) was 0.27 ton/ha. Two-way analysis of variance resulted in 3.0983 of F ratio and 0.1008 of p value. Estimated milled rice yield showed the lowest value as 2.71 ton/ha in 2007, which was consistent with RDA rice yield statistics and the highest value as 3.54 ton/ha in 2006, which was not consistent with the statistics. Scatter plot of estimated rice yield and the rice yield statistics implied that estimated rice yield was higher when the rice yield statistics was less than 3.3 ton/ha and lower when the rice yield statistics was greater than 3.3 ton/ha. Limitation of rice yield model was due to lower quality of climate and statistics data, possible cloud contamination of time-series NDVI data, and crop mask for rice paddy, and coarse spatial resolution of MODIS satellite data. Selection of representative areas for paddy rice consisting of homogeneous pixels and utilization of satellite-based weather information can improve the input parameters for rice yield model in DPRK in the future.