• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.04a
• /
• pp.341-344
• /
• 2001

본 논문에서는 학습을 수행하여 뇌 MR 이미지를 자동으로 분류하고 검색하는 시스템을 설계하였다. 이미지로부터 얻을 수 있는 정보는 크게 두 가지 부류로 나눌 수 있다. 이미지 자체로부터 얻을수 있는 크기, 색상, 질감, 윤곽선 등의 하위레벨(low-level) 정보가 있고, 이미지 의미 해석에서 오는 전이, 포함, 방향, 등의 상위레벨(high-level) 정보가 있다. 이 논문은 의료 이미지에 대하여 상위 및 하위 레벨 정보의 각 특징을 살리고 효과적으로 검색하기 위해, 두 부류의 이미지 정보에 대한 결정 트리(Decision Tree) 학습을 2 단계로 적용하여 이미지를 분류하도록 시스템을 설계하였다.

• Journal of radiological science and technology
• /
• v.35 no.2
• /
• pp.165-172
• /
• 2012

We investigated the vascular characteristics of tumors and normal tissue using perfusion CT in the rabbit brain tumor model. The VX2 carcinoma concentration of $1{\times}10^7$ cells/ml(0.1ml) was implanted in the brain of nine New Zealand white rabbits (weight: 2.4kg-3.0kg, mean: 2.6kg). The perfusion CT was scanned when the tumors were grown up to 5mm. The tumor volume and perfusion value were quantitatively analyzed by using commercial workstation (advantage windows workstation, AW, version 4.2, GE, USA). The mean volume of implanted tumors was $316{\pm}181mm^3$, and the biggest and smallest volumes of tumor were 497 $mm^3$ and 195 $mm^3$, respectively. All the implanted tumors in rabbits are single-nodular tumors, and intracranial metastasis was not observed. In the perfusion CT, cerebral blood volume (CBV) were $74.40{\pm}9.63$, $16.08{\pm}0.64$, $15.24{\pm}3.23$ ml/100g in the tumor core, ipsilateral normal brain, and contralateral normal brain, respectively ($p{\leqq}0.05$). In the cerebral blood flow (CBF), there were significant differences between the tumor core and both normal brains ($p{\leqq}0.05$), but no significant differences between ipsilateral and contralateral normal brains ($962.91{\pm}75.96$ vs. $357.82{\pm}12.82$ vs. $323.19{\pm}83.24$ ml/100g/min). In the mean transit time (MTT), there were significant differences between the tumor core and both normal brains ($p{\leqq}0.05$), but no significant differences between ipsilateral and contralateral normal brains ($4.37{\pm}0.19$ vs. $3.02{\pm}0.41$ vs. $2.86{\pm}0.22$ sec). In the permeability surface (PS), there were significant differences among the tumor core, ipsilateral and contralateral normal brains ($47.23{\pm}25.45$ vs. $14.54{\pm}1.60$ vs. $6.81{\pm}4.20$ ml/100g/min)($p{\leqq}0.05$). In the time to peak (TTP) were no significant differences among the tumor core, ipsilateral and contralateral normal brains. In the positive enhancement integral (PEI), there were significant differences among the tumor core, ipsilateral and contralateral brains ($61.56{\pm}16.07$ vs. $12.58{\pm}2.61$ vs. $8.26{\pm}5.55$ ml/100g). ($p{\leqq}0.05$). In the maximum slope of increase (MSI), there were significant differences between the tumor core and both normal brain($p{\leqq}0.05$), but no significant differences between ipsilateral and contralateral normal brains ($13.18{\pm}2.81$ vs. $6.99{\pm}1.73$ vs. $6.41{\pm}1.39$ HU/sec). Additionally, in the maximum slope of decrease (MSD), there were significant differences between the tumor core and contralateral normal brain($p{\leqq}0.05$), but no significant differences between the tumor core and ipsilateral normal brain($4.02{\pm}1.37$ vs. $4.66{\pm}0.83$ vs. $6.47{\pm}1.53$ HU/sec). In conclusion, the VX2 tumors were implanted in the rabbit brain successfully, and stereotactic inoculation method make single-nodular type of tumor that was no metastasis in intracranial, suitable for comparative study between tumors and normal tissues. Therefore, perfusion CT would be a useful diagnostic tool capable of reflecting the vascularity of the tumors.

• Tuberculosis and Respiratory Diseases
• /
• v.72 no.2
• /
• pp.182-186
• /
• 2012

Miliary brain metastasis from the lung is uncommon and has a poor therapeutic response. We report a case of pulmonary adenocarcinoma combined with multiple brain cystic lesions that were initially misdiagnosed as neurocysticercosis. A 53-year-old male who never smoked was admitted to our hospital with complaints of agitation and cognitive impairment. Brain magnetic resonance imaging showed innumerable, small nodular lesions with a central, low signal intensity in whole brain parenchyma. His symptoms were not improved by the empirical praziquantel medication for disseminated neurocysticercosis. After a transbronchial biopsy from the right middle lobe, we could diagnose the primary lung adenocarcinoma with a single nucleotide polymorphism in the epidermal growth factor receptor exon 20 at codon 787 (Q787Q). His neurologic symptoms and imaging findings have been gradually improving with a first-line Gefitinib treatment for five months. We recommend a more active diagnostic approach including biopsy in case of atypical imaging findings.

• Investigative Magnetic Resonance Imaging
• /
• v.2 no.1
• /
• pp.104-112
• /
• 1998

Purpose : To evaluate the clinical utility of diffusion-weighted imaging by analyzing the signal intersity of lesions in patients with various intracranial diseases. Materials and Methods : difusion-weighted MR imaging was prospectively perormed in randomly selected 70 patients with various intracranial idseases. They consisted of 20 patients with acute infarct, 21 patients with chronic infarct of small vessel disease, 14 patients with primary intracranial tumor, three patients with brain metastasis, five patient with brain abscess, five patients with brain abscess, five patients with cerebral hemorrhage, one patient with neurocysticercosis, and one patient with epidermoid cyst. the diffusion-weighted images were obtained immediately after routine T2-weighted imaging on a 1.5T MR unit using single shot spin echo EPI technique with 6500 ms TR, 107ms TE, $128{\times}128$ matrix, 1 number of excitation, $24{\times}24$ field of view, 5-7 mm slice thickness, 2-3 mm inter-slice gap. The diffusion-gradients (b value of ($1000s{\;}/{\;}textrm{mm}^2$)) were applied along three directions(x, y, z). On visual inspection of diffusion-weighted images, the signal intersity of lesions was arbitrarily graded as one of 5 grades. In quantitative assessment, we measured the signal intensity of all the lesions and the contralateral corresponding normal area using round region of interest(ROI), and then calculated the signal intensity ratio of the lesion to the normal brain parenchyma. Results : On visual inspection, markedly hyperintense signals were seen in all cases of acute infarct, brain abscess, epidermoid cyst, and neurocysticercosis in degenerating stage. In all cases of cerebral hematoma, the very high signal internsity was intermingled with low signal intensity. focal very high signal intersity was also seen in a solid portion of the tumor in a patient. the mean signal intensity ratios of all those lesions to the normal brain parenchyma were above 2.5. Gliosis, solid component of brain tumor, brain metastasis, and vasogenic dedma appeared isointense to the normal brain parenchyma in 71%, 64%, 100%, and 67%, respectively ; the mean signal intensity ratios of those lesions to the normal brain parenchyma ranged 1.15 to 1.28 and there was no significant difference among these(p>0.1). Cystic cerebromalacia and necrotic or cystic portions in tumor were markedly or slightly hypointense, and the mean signal intensity ratios were 0.45 and 0.42, respectively. Conclusion : Very high signal intensity of acute infarct, brain abscess, epidermoid cyst, and cystic neurocysticercosis in degenerating stage on diffusion-weighted images may be helpful in differentiating from other diseases that are hypointense or isointense to the normal brain parenchyma. It may be especially useful differentiation of brain abscess from brain tumor with necrotic or cystic portion.

• Tuberculosis and Respiratory Diseases
• /
• v.40 no.1
• /
• pp.72-78
• /
• 1993

Background: In malignancies, detection of metastatic foci is of value in making therapeutic plans for treatment of disease and prevention of life-threatening complications. Common sites for metastasis of bronchogenic cancer include lymph nodes, liver, brain, adrenals and bone. Skull, vertebrae, ribs and long bones are common sites for bone metastasis. But in epidermoid carcinoma, the incidence of bone metastasis is relatively low and especially to the distal phalangeal bone is rare. Methods: We experienced a case of epidermoid carcinoma with the first distant metastasis to the fifth distal phalangeal bone, right toe. Results: The initial stage in the diagnosis of epidermoid carcinoma was T4N3MO. During the third round of anticancer chemotherapy, we recognized the distant metastasis to the fifth distal phalangeal bone for the first time. Localized abnormal findings were noted by bone X-ray and scanning. By a histopathologic examination of the amputated toe, we confirmed the metastasis of epidrmoid bronchogenic carcinoma. Conclusion: If localized abnormal finding is discovered at an unusual site for metastasis, we recommend physicians to consider the possibility of metastasis even though it is very low.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.10
• /
• pp.6187-6192
• /
• 2014

The use of WBRT(whole brain radiation therapy) has increased due to the increase in the incidence of metastatic brain tumors. The development of radiation therapy techniques is expected to improve the quality of life. The VMAT (Volumetric Modulated Arc Therapy) is an excellent treatment technique that can distinguish the dose in each volume. Therefore, this study compared conventional WBRT and VMAT for hair loss according to the scalp dose using a head phantom. The CI (Conformity Index), HI (Homogeneity Index) and QOC (Quality of Coverage) were measured brain tissue. A 20 percent and 50 percent dose was measured at the scalp, eyeball, lens, and c-spine. Conventional WBRT is excellent at 10 percent of brain tissue. VMAT is far superior at 1000 percent at the other organs. VMAT at the prescribed dose can be used as radiation therapy of metastatic brain tumors with less hair loss.

• The Journal of Korean Society for Radiation Therapy
• /
• v.30 no.1_2
• /
• pp.73-81
• /
• 2018

Purpose : The aims of this study were to compare and assess the effectiveness of Volumetric Modulated Arc Therapy(VMAT) using jaw tracking(JT) and fixed jaw(FJ) in radiation therapy of multiple brain metastasis. Methode and material : Among the patients with Multiple Brain Metastasis treated with jaw tracking, 10 patients with more than 6 tumors and with the size of radiation field $14{\times}14cm^2$ or more were included. Each Treatment plans with jaw tracking(JT) and fixed jaw(FJ) was established with Eclipse (Ver. 13.6 Varian, USA). Gamma Index (3 mm, 3 % confidence interval - 95 %) and maximum dose difference were measured with an electronic portal imaging device(EPID). The $D_{max}$ and $D_{mean}$ of Organ At Risk(OAR) were assessed and compared, and the Conformity Index(CI) and Homogeneity Index(HI) were evaluated. Result : Evaluating jaw tracking(JT) and fixed jaw(FJ) outcomes, in all cases, Gamma Index met the permissible standard of 3 mm, 3 % confidence intervals of 95 %. The maximum dose difference value from the areas with leaf end transmission was measured at a maximum of 98.4 % and an average of 43.6 % in clockwise(CW), and 67.9 % and 41.0 % for each in Counter-Clockwise(CCW). With jaw tracking, the maximum value of $D_{max}$ for each normal organ in OAR decreased in 15.36 %~74.59 % with the average value decreasing in 2.84 %~39.80 %. The maximum value of $D_{mean}$ in OAR decreased in 27.90 %~65.23 %, with the average value decreasing in 7.70 %~41.71 %. No change has been found in Conformity Index and Homogeneity Index values. Conclusion : When Jaw tracking is used in treating patients with multiple brain metastasis with VMAT, the unnecessary exposure due to leakage and transmission of radiation in unspecified areas was reduced, without affecting the dose distribution of the planning target volume(PTV), and the availability of radiation therapy with lower doses in normal organs is expected.

• Radiation Oncology Journal
• /
• v.18 no.2
• /
• pp.92-100
• /
• 2000

Purpose :This study was performed to determine the optimal treatment velum of Patients treating with radiation therapy for intracranial germ cell tumor. Materials and Methods : From 1993 to 1998, 19 patients with intracranial germ cell tumors treated by gamma knife radiosurgery were analyzed. The location of tumor was as follows; 9 cases on pineal region, 1 case on suprasellar region, and 9 cases of multiple lesion. 7 patients were pathologically verified; 5 cases of germ cell tumor and 2 cases of non germinomatous germ cell tumor. Tumor volume was ranged from 2.4 cm$^{3}$ to 74 cm$^{3}$. Irradiation dose was 10 Gy to 20 Gy with 50% isodose curve. Follow up period was 10 months to 54 months. Results : Recurrences were observed in 14 cases among 19 (74%) patients. Complete remission and partial remission were achieved in 2 (11%) and 10 (53%) respectively. No response was observed in 7 (36%). 2 cases were recurred within original tumor bed. 6 cases were recurred beyond but contiguous with tumor bed. Ventricular relapses separated from pretreatment tumor bed were 3. Spinal recurrences were 4. Among 8 recurred cases of which tumor volume is smaller than 20 cm$^{3}$, 2 were recurred within original tumor bed, 4 were recurred beyond but contiguous with tumor bed, and 1 spinal recurrence. Meanwhile, 6 cases of which tumor volume larger than 20 cm3, 1 case was recurred beyond but contiguous with tumorbed, 2 ventricular recurrences separated with original tumor bed, and 3 spinal recurrences. 5 cases which did not show any recurrence sign showed characteristics of single lesion, tumor volume smaller than 20 cm$^{3}$ and normal tumor marker. All of 4 cases of spinal recurrences happened in the case having ventricular invasion or lesion. Among 9 cases having multiple lesion, only 3 cases recurred within original tumor bed or around tumor bed, the other 6 cases recurred separated from pretreatment tumor bed. Conclusion : Gamma knife radiosurgery is not recommended for the treatment of intracranial germ cell tumor. It is because of small treatment volume and inadequate radiation dose that are characteristics of gamma knife radiosurgery. Tumor volume, ventricular invasion or ventricular lesion in multiple lesion are important factors to be considered for the wide field radiation therapy Tumor volume smaller than 20 cm$^{3}$, single lesion, no ventricular lesion or invasion, and normal tumor marker are ideal indications for small involved field radiation therapy. Prophylactic spinal irradiation seems to be necessary when there is ventricular lesion, ventricular invasion, and multiple lesions. When the tumor volume is larger than 20 cm$^{3}$, multiple lesions, abnormal tumor marker, and whole ventricular irradiation or partial brain irradiation would be possible and neoadjuvant chemotherapy would be most beneficial in these group.

• Journal of Ginseng Research
• /
• v.33 no.1
• /
• pp.1-7
• /
• 2009

Clinical and animal studies have shown that free radical overload is an important cause for a variety of diseases. Although ginseng has been recognized as antioxidant, how it modulates anti-oxidative process at the molecular level remains unknown. Free radical production is induced by tumor necrosis factor-$\alpha$ (TNF-$\alpha$) under the stress condition, and (TNF-$\alpha$) release is activated by TNF-$\alpha$-converting enzyme (TACE). Since TACE inhibitor is also well known for anti-inflammatory agent, ginseng seems to show anti-oxidative activity by repressing TACE pathway. Further studies on signal transduction would be warranted to elucidate molecular action mechanisms of ginseng on anti-oxidation and anti-inflammation.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.7
• /
• pp.37-44
• /
• 2021

Automatic classification of brain MRI images play an important role in early diagnosis of brain tumors. In this work, we present a deep learning-based brain tumor classification model in MRI images using ensemble of deep features. In our proposed framework, three different deep features from brain MR image are extracted using three different pre-trained models. After that, the extracted deep features are fed to the classification module. In the classification module, the three different deep features are first fed into the fully-connected layers individually to reduce the dimension of the features. After that, the output features from the fully-connected layers are concatenated and fed into the fully-connected layer to predict the final output. To evaluate our proposed model, we use openly accessible brain MRI dataset from web. Experimental results show that our proposed model outperforms other machine learning-based models.