• Radiation Oncology Journal
• /
• v.28 no.1
• /
• pp.1-8
• /
• 2010

Purpose: This retrospective study was performed to evaluate the efficacy of radiation therapy (RT) and to investigate the prognostic factors for thymoma when treated with RT. Materials and Methods: We analyzed 21 patients with thymoma and also received RT from March 2002 to January 2008. The median follow-up time was 37 months (range, 3 to 89 months). The median patient age was 57 years (range, 24 to 77 years) and the gender ratio of males to females was 4：3. Of the 21 patients, complete resections (trans-sternal thymectomy) and R2 resections were performed in 14 and 1 patient, respectively. A biopsy was performed in 6 patients (28.7%). The WHO cell types in the 21 patients were as follows: 1 patient (4.8%) had type A, 10 patients (47.6%) had type B1-3, and 10 patients (47.6%) had type C. Based on Masaoka staging, 10 patients (47.6%) were stage II, 7 patients (33.3%) were stage III, and 4 patients (19.1%) were stage IVa. Three-dimensional RT was adminstered to the tumor volume (planned target volume), including the anterior mediastinum and the residual disease. The total RT dose ranged from 52.0 to 70.2 Gy (median dose, 54 Gy). Consistent with the WHO criteria, the response rate was only analyzed for the 6 patients who received a biopsy only. The prognostic factors analyzed for an estimate of survival included age, gender, tumor size, tumor pathology, Masaoka stage, the possibility of treatment by performing surgery, the presence of myasthenia gravis, and RT dose. Results: The 3-year overall survival rate (OS) and the progression free survival rate (PFS) were 80.7% and 78.2%, respectively. Among the 10 patients with WHO cell type C, 3 of 4 patients (75%) who underwent a complete resection and 3 of 6 patients (50%) who underwent a biopsy survived. Distant metastasis developed in 4 patients (19.1%). The overall response rate in the 6 patients who received biopsy only were as follows: partial remission in 4 patients (66.7%), stable disease in 1 patient (16.6%), and progressive disease in 1 patient (16.6%). Acute RTOG radiation pneumonitis occurred in 1 patient (4.8%), grade 2 occurred in 2 patients (9.5%), grade 3 occurred in 1 patient (4.8%), and grade 4 occurred in 1 patient (4.8%). A univariate analysis revealed that the significant prognostic factors for OS were age (${\geq}60$, 58.3%; <60, 100%; p=0.0194), pathology (WHO cell type A-B3, 100%; C, 58.3%; p=0.0194) and, whether the patient underwent surgery (yes, 93.3%; no, 50%; p=0.0096). Conclusion: For the 15 patients who received surgery, there was no local failure within the radiation field. In patients with WHO cell type C, surgical procedures could have resulted in a more favorable outcome than biopsy alone. We report here our clinical experience in 21 patients with thymoma who were treated by radiation therapy.

• Radiation Oncology Journal
• /
• v.20 no.2
• /
• pp.139-146
• /
• 2002

Purpose : This study evaluated the treatment outcomes, patterns of failure, and treatment related complications of primary lymphoma patients who received definitive radiation therapy. Materials and Methods : A retrospective analysis was undertaken for 31 patients with primary orbital lymphoma at the Asan Medical Center between February 1991 and April 2001. There were 18 males and 13 females with ages ranging from 3 to 73 years (median, 44 years). The involved sites were 9 conjunctivae, 12 eyelids and 10 other orbits. The histological types were 28 MALT lymphomas (low-grade B-cell lymphoma of mucosa-associated lymphoid tissue type), 1 diffuse large B-cell lymphoma, 1 anaplastic large cell lymphoma and 1 lymphoblastic lymphoma. The Ann Arbor stages were all IE $(100\%)$. Ann Arbor stage III or IV patients were excluded from this study, Bilateral orbital involvement occurred in 6 cases. Radiation therapy was given with one anterior port of high energy electrons $(6\~16\;MeV)$ for the lesions located at the anterior structures like the conjunctivae or eyelids. Lesions with a posterior extension or other orbital lesions were treated with 4 or 6 MeV photons with appropriately arranged portals. In particular, lens blocks composed of lead alloy were used in conjunctival or eyelid lesions. Twelve patients received chemotherapy. The median follow-up period was 53 months. Results : The 5-year overall, cause-specific, and disease-free survival was $91\%,\;96\%,\;and\;80\%$, respectively. The complete response rate 6 months after radiation therapy was $100\%$. Local recurrences were observed in 2 patients at 16 and 18 months after completion of radiation treatment. They were salvaged with additional radiation therapy. Two patients developed distant metastases. A MALT lymphoma patient with a lung relapse was successfully salvaged with radiotherapy, but the other lymphoblastic lymphoma patient with bone marrow relapse expired. There were no severe complications but 5 patients developed radiation-induced cataracts and 2 patients developed dry eye. Conclusion : Most primary orbital lymphomas consisted of MALT lymphomas. Radiation therapy was a successful treatment modality for orbital lymphoma without any severe complications. In cases of local relapses, radiation therapy is also a very successful salvage treatment modality.

• Radiation Oncology Journal
• /
• v.24 no.3
• /
• pp.171-178
• /
• 2006

[ $\underline{Purpose}$ ]: To determine the efficacy and safety of concurrent chemotherapy and radiation therapy with high-dose-rate brachytherapy for cervical cancer. $\underline{Materials\;and\;Methods}$: From January 2001 to December 2002, 30 patients with cervical cancer were treated with concurrent chemotherapy (cisplatin and 5-FU) and definitive radiation therapy. The median age was 58 (range $34{\sim}74$) year old. The pathology of the biopsy sections was squamous cell carcinoma in 29 patients and one was adenocarcinoma. The distribution to FIGO staging system was as follows: stage IB, 7 (23%); IIA, 3 (10%); IIB, 12 (40%); IIIA, 3 (10%); IIIB, 5 (17%). All patients received pelvic external beam irradiation (EBRT) to a total dose of $45{\sim}50.4\;Gy$ (median: 50.4 Gy) over $5{\sim}5.5$ weeks. Ir-192 HDR intracavitary brachytherapy (ICBT) was given after a total dose of 41.4 Gy. HDR-ICBT was performed twice a week, with a fraction point A dose of 4 Gy and median dose to point A was 28 Gy (range: $16{\sim}32\;Gy$) in 7 fractions. The median cumulative biologic effective dose (BED) at point A (EBRT+ICBT) was $88\;Gy_{10}$ (range: $77{\sim}94\;Gy_{10}$). The median cumulative BED at ICRU 38 reference point (EBRT+ICBT) was $131\;Gy_3$ (range: $122{\sim}140\;Gy_3$) at point A, $109\;Gy_3$ (range: $88{\sim}125\;Gy_3$) at the rectum and $111\;Gy_3$ (range: $91{\sim}123\;Gy_3$) at the urinary bladder. Cisplatin ($60\;mg/m^2$) and 5-FU ($1,000\;mg/m^2$) was administered intravenously at 3 weeks interval from the first day of radiation for median 5 (range: $2{\sim}6$) cycles. The assessment was performed at 1 month after completion of radiation therapy by clinical examination and CT scan. The median follow-up time was 36 months (range: $8{\sim}50$ months). $\underline{: The complete response rate after concurrent chemoradiation therapy was 93.3%. The 3-yr actuarial pelvic control rate was 87% and 3-yr actuarial overall survival and disease-free survival rate was 93% and 87%, respectively. The local failure rate was 13% and distant metastatic rate was 3.3%. The crude rate of minor hematologic complications (RTOG grade 1-2) occurred in 3 patients (10%) and one patient had suffered from severe leukopenia (RTOG grade 4) during concurrent treatment. Acute minor enterocolitis (RTOG grade 1-2) occurred in 11 patients (37%) and one patient (3%) was suffered from colon perforation during radiation therapy. Late colitis of RTOG grade 1 occurred in 5 patients (15%). Acute cystitis of RTOG grade 1 occurred in 12 patients (40%) and late cystitis of RTOG grade 2 occurred in one patient (3%). No treatment related death was seen. $\underline{Conclusion}$: The results of this study suggest that the concurrent chemoradiation therapy with HDR brachytherapy could be accepted as an effective and safe treatment for cervical cancer.

• Radiation Oncology Journal
• /
• v.22 no.1
• /
• pp.17-24
• /
• 2004

Purpose: To evaluate the treatment outcomes after postmastectomy radiotherapy (PMRT) and chemotherapy in patients with breast cancer. Materials and Methods: The PMRT were retrospectively analyzed in 83 patients with stage II-III female breast cancer treated between 1989 and 1995. The median age was 46 years (range, 23-77): Seventy-seven patients had modified radical mastectomies, 5 radical mastectomies and 1 simple mastectomy. Three patients ($4\%$) had pathologically negative axillae, and the remaining 80 ($96\%$) had positive axillae. Eleven, 23, 44 and 5 patients had pathological stages IIA, IIB, IIIA, and IIIB, retrospectively. Eighty ($96\%$) patients were treated with hockey-stick fields. The median dose of PMRT was 50.4 ey, in 1.8 Gy fractions. Adjuvant systemic chemotherapy was given to 74 patients ($89\%$). CMF-based or doxorubicin-containing regimens were given to 54 patients ($55\%$). The median follow-up time was 82 months (range, 8-171) after the mastectomy. Results: The 5 and 10-year overall survival rates for all patients were 65 and $49\%$, respectively. The univariate and multivariate analyses of the factors affecting the overall survival revealed the stage to be the most significant prognostic factor (p=0.002), followed by the combination of chemotherapy. Thirteen patients $16\%$ developed a LRF, at an interval of 4-84 months after radiotherapy, with a median of 20 months. The only significant prognostic factor affecting LRF was the combination of chemotherapy, in both the univariate and multivariate analyses. With respect to the sequence of chemoradiation, the sequence had no saatistical significance (p=0.90). According to the time interval from mastectomy to the onset of radiotherapy, the LRFR of the patients group treated by RT within or after 6 month postmastectomy 6 months were 14 vs. $27\%$ respectively (p=0.24). One third of the pa41en1s (26/83) developed distant metastasis, in 2-92 months, after radiotherapy, with a median of 21 months. The most commonly involved site was bone in 13 cases. The pathological staging was the only significant prognostic factor in both the univariate and multivariate analyses that affected distant failure. Radiological finding of radiation pneumonitis on a simple chest x-ray was shown in $20\%$ (17/83), with a time interval ranging from 2 to 7 months post-radiotherapy, with a median of 3 months. The stable lung fibrosis settled in 11 patients ($65\%$). Conclusion: It was concluded through this analysis that the combination of PMRT with in chemotherapy resulted in better overall survival and local control than PMRT alone in patients needing PMRT.

• Tuberculosis and Respiratory Diseases
• /
• v.44 no.2
• /
• pp.280-289
• /
• 1997

Background : Bronchioloalveolar carcinoma (BAC) has been reported to diveres spectrum of chinical presentations and radiologic patterns. The three representative radiologic patterns are followings ; 1) a solitary nodule or mass, 2) a localized consolidation, and 3) multicentric or diffuse disease. While, the localized consolidation and solitary nodular patterns has favorable prognosis, the multicentric of diffuse pattern has worse prognosis regardless of treatment. BAC presenting as a solitary pulmonary nodule is often misdiagnosed as other benign disease such as tuberculoma. Therefore it is very important to make proper diagnosis of BAC with solitary nodular pattern, since this pattern of BAC is usually curable with a surgical resection. Methods : We reviewed the clinical and radiologic features of patients with pathologically-proven BAC with solitary nodular pattern from January 1995 to September 1996 at Samsung Medical Center. Results : Total 11 patients were identified. 6 were men and 5 were women. Age ranged from 37 to 69. Median age was 60. Most patients with BAC with solitary nodular pattern were asymptomatic and were detected by incidental radiologic abnormality. The chest radiograph showed poorly defined opacity or nodule and computed tomography showed consolidation, ground glass appearance, internal bubble-like lucencies, air bronchogram, open bronchus sign, spiculated margin or pleural tag in most patients. The initial diagnosis on chest X-ray were pulmonary tuberculosis in 4 patients, benign nodule in 2 patients and malignant nodule in 5 patients. The FDG-positron emission tomogram was performed in eight patients. The FDG-PET revealed suggestive findings of malignancy in only 3 patients. The pathologic diagnosis was obtained by transbronchial lung biopsy in 1 patient, by CT guided percutaneous needle aspiration in 2 patients, and by lung biopsy via video-assited thoracocopy in 8 patients. Lobectomy was performed in all patients and postoperative pathologic staging were $T_1N_0N_0$ in 8 patients and $T_2N_0M_0$ in 3 patients. Conclusion : Patients of BAC presenting with solitary nodular pattern were most often asymptomatic and incidentally detected by radiologic abnormality. The chest X-ray showed poorly defined nodule or opacity and these findings were often regarded as benign lesion. If poorly nodule or opacity does not disappear on follow up chest X-ray, computed tomography should be performed. If consolidation, ground glass appearance, open bronchus sign, air bronchogram, internal bubble like lucency, pleural tag or spiculated margin are found on computed tomography, further diagnostic procedures, including open thoracotomy, should be performed to exclude the possiblity of BAC with solitary nodular pattern.

• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.105-129
• /
• 2020

This study uses corporate data from 2012 to 2018 when K-IFRS was applied in earnest to predict default risks. The data used in the analysis totaled 10,545 rows, consisting of 160 columns including 38 in the statement of financial position, 26 in the statement of comprehensive income, 11 in the statement of cash flows, and 76 in the index of financial ratios. Unlike most previous prior studies used the default event as the basis for learning about default risk, this study calculated default risk using the market capitalization and stock price volatility of each company based on the Merton model. Through this, it was able to solve the problem of data imbalance due to the scarcity of default events, which had been pointed out as the limitation of the existing methodology, and the problem of reflecting the difference in default risk that exists within ordinary companies. Because learning was conducted only by using corporate information available to unlisted companies, default risks of unlisted companies without stock price information can be appropriately derived. Through this, it can provide stable default risk assessment services to unlisted companies that are difficult to determine proper default risk with traditional credit rating models such as small and medium-sized companies and startups. Although there has been an active study of predicting corporate default risks using machine learning recently, model bias issues exist because most studies are making predictions based on a single model. Stable and reliable valuation methodology is required for the calculation of default risk, given that the entity's default risk information is very widely utilized in the market and the sensitivity to the difference in default risk is high. Also, Strict standards are also required for methods of calculation. The credit rating method stipulated by the Financial Services Commission in the Financial Investment Regulations calls for the preparation of evaluation methods, including verification of the adequacy of evaluation methods, in consideration of past statistical data and experiences on credit ratings and changes in future market conditions. This study allowed the reduction of individual models' bias by utilizing stacking ensemble techniques that synthesize various machine learning models. This allows us to capture complex nonlinear relationships between default risk and various corporate information and maximize the advantages of machine learning-based default risk prediction models that take less time to calculate. To calculate forecasts by sub model to be used as input data for the Stacking Ensemble model, training data were divided into seven pieces, and sub-models were trained in a divided set to produce forecasts. To compare the predictive power of the Stacking Ensemble model, Random Forest, MLP, and CNN models were trained with full training data, then the predictive power of each model was verified on the test set. The analysis showed that the Stacking Ensemble model exceeded the predictive power of the Random Forest model, which had the best performance on a single model. Next, to check for statistically significant differences between the Stacking Ensemble model and the forecasts for each individual model, the Pair between the Stacking Ensemble model and each individual model was constructed. Because the results of the Shapiro-wilk normality test also showed that all Pair did not follow normality, Using the nonparametric method wilcoxon rank sum test, we checked whether the two model forecasts that make up the Pair showed statistically significant differences. The analysis showed that the forecasts of the Staging Ensemble model showed statistically significant differences from those of the MLP model and CNN model. In addition, this study can provide a methodology that allows existing credit rating agencies to apply machine learning-based bankruptcy risk prediction methodologies, given that traditional credit rating models can also be reflected as sub-models to calculate the final default probability. Also, the Stacking Ensemble techniques proposed in this study can help design to meet the requirements of the Financial Investment Business Regulations through the combination of various sub-models. We hope that this research will be used as a resource to increase practical use by overcoming and improving the limitations of existing machine learning-based models.

• Journal of Chest Surgery
• /
• v.38 no.1 s.246
• /
• pp.44-49
• /
• 2005

Although thymomas are relatively common mediastinal tumors, to date not only has a universal system of pathologic classification not been established but neither has a clearly defined predictable relationship between treatment and prognosis been made. Recently, a new guideline for classification was reported by WHO, and efforts, based on this work, have been made to better define the relationship between treatment and pro­gnostic outcome. In the present study a comparative analysis between the WHO classification and Masaoka stage system with the clinical disease pattern was conducted. Material and Method: A total of 98 patients undergoing complete resection for mediastinal thymoma between Juanuary 1993 and June 2003 were included in the present study. The male female ratio was 48 : 50 and the mean age at operation was $49.6{\pm}13.9\;years.$ A retrospective analytic comparison studying the relationship between the WHO classification and the Masaoka stage system with the clinical disease pattern of thymoma was conducted. Pathologic slide specimens were carefully examined, details of postoperative treatment were documented, and a relationship with the prognostic outcome and recurrence was studied. Result: There were 7 patients in type A according to the WHO system of classification, 14 in AB, 28 in B 1, 23 in B2, 18 in B3, and 9 in type C. The study of the relationship between the Masaoka stage and WHO classification system showed 4 patients to be in WHO system type A, 7 in type AB, 22 in B 1, 17 in B2, and 3 in type B3 among 53 $(54{\%})$ patients shown to be in Masaoka stage I. Among 28 $(28.5{\%})$ patients in Masaoka stage II system, there were 2 patients in type A, 7 in AB, 4 in B 1, 2 in B2, 8 in B3, and 5 in type C. Among 15 $(15.3{\%})$ in Masaoka stage III, there were 1 patient in type B1, 3 in B2, 7 in B3, and 4 in type C. Finally, among 2 $(2{\%})$ patients found to be in Masaoka stage IV there was 1 patient in type B1, and 1 in type B2. The mean follow up duration was $28{\pm}6.8$ months. There were 3 deaths in the entire series of which 2 were in type B2 (Masaoka stages III and IV), and 1 was in type C (Masaoka stage II). Of the patients that experienced relapse, 6 patients remain alive of which 2 were in type B2 (Masaoka III), 2 in type B3 (Masaoka I and III) and 2 in type C (Masaoka stage II). The 5 year survival rate by the Kaplan-Meier method was $90{\%}$ for those in type B2 WHO classification system, $87.5{\%}$ for type C. The 5 year freedom from recurrence rate was $80.7{\%}$ for those in WHO type B2, $81.6{\%}$ for those in type B3, and $50{\%}$ for those in type C. By the Log-Rank method, a statistically significant correlation between survival and recurrence was found with the WHO system of classification (p<0.05). An analysis of the relationship between the WHO classification and Masaoka stage system using the Spearman correction method, showed a slope=0.401 (p=0.023), showing a close correlation. Conclusion: As type C of the WHO classification system is associated with a high postoperative mortality and recurrence rate, aggressive treatment postoperatively and meticulous follow up are warranted. The WHO classification and Masaoka stage system were found to have a close relationship with each other and either the WHO classification method or the Masaoka stage system may be used as a predict prognostic outcome of Thymoma.