• Journal of the Korea Computer Industry Society
• /
• v.3 no.9
• /
• pp.1279-1286
• /
• 2002

Generally, radiologists can fail to detect pulmonary nodules in up to 30%. If an automatic system can inform the radiologists of thelocations of the doubtful nodules in the X-ray chest images, the frequency of mistakenly observed numbers of the nodules can be potentially reduced. This software is using morphological filtering and two feature-extraction techniques. The morphological filtering is the first process, which subsequently adds the operations of erosion and dilation to the original images so that this process can transform the original X-ray chest images into manageable ones. The false-positives are frequently being mistaken as nodules but actually these are not real nodules. The second process is the two feature-extraction techniques which are used to reduce the false-positives. Therefore, this system will make more effective detection of pulmonary nodules by reducing the false-positives when applied to the X-ray chest images which is difficult to get accurate detection.

• Journal of Chest Surgery
• /
• v.22 no.3
• /
• pp.425-435
• /
• 1989

The solitary pulmonary nodule is considered as a round or ovoid lesion with sharp, circumscribed borders, surrounded by normal appearing lung parenchyme on all sides, and found on a simple chest X-ray without any particular symptoms or signs. There is a wide spectrum of pathologic conditions in the solitary pulmonary nodules prove to be malignant tumors, either primary or metastatic. Most Benign granulomas and other benign conditions can also be seen as solitary nodules. The resection of solitary malignant nodules results in a surprisingly high 5-year survival rate. On the contrary, most benign nodules do not need to be resected and a period of prolonged observation and nonsurgical management is usually indicated. Therefore, the best approach to the controversial management of solitary pulmonary nodules depends on finding factors affecting the probability of malignancy. In this article, clinical records and chest roentgenographies of 60 patients operated on over the past 8 years at the Department of Thoracic and Cardiovascular Surgery, Seoul National University Hospital were reviewed. There were 15 malignant nodules and 45 benign nodules and the prevalence of malignancy was 25%. The most common pathologic entity was tuberculoma [21 cases]. The mean age was 55.5*9.6 years in the malignant group, 45.8>12.5 years in the benign group and there was a significant statistical difference between the two groups [P < 0.05]. The malignant ratio in each age group increased with advancing age. The average smoking amount was 35.6*12.9 cigarettes per day in malignant smokers, 20.9* 12.0 cigarettes per day in benign smokers, and there was a significant statistical difference between the two groups [p< 0.05]. The malignant ratio also increased with the increasing smoking amount. Comparing the appearance of the nodule on chest films, 6 calcifications and 7 cavitations were found only in benign nodules, not in malignant nodules. Therefore, calcification and cavitation can be considered as preferential findings for benignity. Previous cancer history was also a significant factor deciding the prognosis of the nodule [p< 0.05]. The average diameter on chest X-ray was 3.07*0.82 cm in malignant nodules, 3.25*1.04 cm in benign nodules and there was no significant statistical difference between the two groups [p< 0.05]. The author used Bayes theorem to develop a simple method for combining individual clinical or radiological factors of patients with solitary nodules into an overall estimate of the probability that the nodule is malignant. In conclusion, patient age, smoking amount, appearance of nodule on chest film such as calcification and cavitation, and previous cancer history were found to be strongly associated with malignancy, but size of nodule was not associated with malignancy. Since these prognostic factors have been found retrospectively, prospective controlled studies are needed to determine whether these factors have really prognostic significance.

• Journal of the Korean Society of Radiology
• /
• v.83 no.2
• /
• pp.387-393
• /
• 2022

Primary pulmonary malignant melanoma is an extremely rare type of melanoma. The radiologic features of primary pulmonary malignant melanoma are nonspecific; however, it almost always presents as a well-demarcated round or lobulated solitary solid nodule or mass. Herein, we report the case of a 78-year-old male with primary pulmonary malignant melanoma that was mistaken for primary pulmonary adenocarcinoma with lepidic growth and was seen as bilateral multiple subsolid nodules on CT.

• Journal of Chest Surgery
• /
• v.29 no.11
• /
• pp.1232-1240
• /
• 1996

From January 1990 through June 1995, we operated on 121 patients who were suspected for pulmonary tuberculosis without definite final diagnosis. After operation the final pathologic diagnoses were as follows: 68 pulmonary tuberculosis in which 29 were tuberculoma, 23 lung cancer, 16 bronchiectasis, 6 aspergilloma, 2 lung abscess, 2 benign cyst and 4 others. In 121 cases, 81 were male and 40 were female and the peak age incidence was 4th decade in tuberculosis (39.7%) and 6th and 7th decade in lung cancer (69.6%). The diagnoses in 44 cases presented roentgenographically as pulmonary nodules were pulmonary tuberculosis(29 cases) and lung cancer(15 cases). Tuberculous nodules tended to be smaller in size with calcification and satellite lesions compared to carcinomas. Indications for operation were solitary nodules 44 cases (36.4%); destroyed lobe 31(25.6%); hemoptysis 25 (20.7%); cavitary lesion 11(9.1 %); bronchostenosis 3 (2.5%); destroyed lung 5(4.1 %) and destroyed lung with empyema 2(1.7%). We conclude that preoperatively suspected pulmonary tuberculosis should be distinguished from various pulmonary lesions such as carcinoma, bronchiectasis, aspergilloma, lung abscess and benign cyst. For the possibility of carcinoma, pulmonary nodules of size greater than 3cm, non-calcified, non satellite lesion, newly developed nodule even under the anti-tuberculous medication, negative PPD skin test with elevated CEA level are recommended for an early resectional surgery and follow-up and delayed surgery is recommended in cases such as pulmonary nodules less than 3 cm in size with calcification, satellite lesion, positive PPD skin reaction and elevated ESR, CRP, ALP levels.

• Tuberculosis and Respiratory Diseases
• /
• v.58 no.6
• /
• pp.614-618
• /
• 2005

A benign metastasizing pulmonary leiomyoma (BMPL) is a rare disease that usually occurs in women with a prior or coincident history of uterine leiomyoma. Although leiomyoma is histologically benign, it has the potential to metastasize to a distant site such as the lung. A 35 year old woman who had undergone a hysterectomy due to uterine leiomyoma 5 years prior was admitted for an investigation of multiple pulmonary nodules on a routine chest roentgenogram. An open lung biopsy was taken to make a pathological diagnosis. The microscopic finding of the nodules was leiomyoma and was similar to those of the uterine leiomyoma that had been resected 5 years ago. The woman underwent wedge resections of all pulmonary nodules. This is the first case of BMPL in Korea, which was treated with wedge resections of all multiple pulmonary nodules.

• Journal of Yeungnam Medical Science
• /
• v.7 no.2
• /
• pp.141-149
• /
• 1990

The authors conducted a clinical observation of 55 cases of solitary pulmonary nodules at Yeungnam University Hospital from June 1986 to October 1990, and the following results were obtained : 1. The age distribution was ranged from 18 to 77 years, and the male-to-female ratio was 1.8:1. 2. Among 55 cases of nodules, 28 cases were benign and 27 cases were malignant nodules, and of malignant nodules, the primary lung cancer was 23 cases and of benign nodules, 18 cases were tuberculoma. 3. 23 cases (41.8%) was asymptomatic and the other 32 cases were symptomatic; chest pain 12 cases, hemoptysis; 8 cases, cough; 8 cases and dyspnea; 4 cases. 4. The non-smoker-to-smoker ratio was 1:1.04, but among 23 smoker over 20 pack years, 14 cases were malignant nodules. 5. According to nodular size, there is no striking differences between benign and malignant nodules except 3-4cm sized nodules. 6. The lobar distribution of nodules, 35 cases were in the right lung (upper lobe; 14 cases, middle lobe; 11 cases, and lower lobe; 10 cases) and 20 cases were in the left lung(upper lobe; 9 cases, lower lobe; 11 cases), and the malignant nodules were most commonly observed in the right upper lung.

• The Korean Journal of Cytopathology
• /
• v.11 no.2
• /
• pp.99-102
• /
• 2000

Pulmonary amyloid deposition generally occurs with concurrent primary systemic amyloidosis. Localized forms of pulmonary amyloidosis are rare and appear most frequently as an incidental finding on chest radiographs. We present a case of nodular pulmonary amyloid tumor suggested by fine needle aspiration cytology(FNAC) and confirmed by histology examination with the polarizing light microscopy. A 41-year-old woman presented with ill-defined nodules. In the middle and lower lobes of both lungs. FNAC of the nodules revealed waxy, acellular amorphous fragments. Thoracotomy for diagnosis may be avoided by FNAC diagnosis of this unusual lesion.

• Tuberculosis and Respiratory Diseases
• /
• v.56 no.2
• /
• pp.151-158
• /
• 2004

Background : When a non small cell lung caner patient at the $_cT_{1-2}N_0M_0$ stage is diagnosed with intrapulmonary nodule(s), the treatment plan and prognosis of the patient largely depend on whether the nodule is benign or malignant. In most cases, however, it is hard to conduct a biopsy on such a nodule, due to its small size. Furthermore, the predictive factors that may imply benignancy or malignancy of the nodules remain unknown. As such, the purpose of our study was to validate the incidence of malignant nodules in such cases, and find if there are any predictive factors. Methods : Chest computed tomography(CT) scans and the medical records of 444 patients, who had undergone non small cell lung cancer surgery, between July, 2001 and September, 2003, at Seoul National University Hospital, were retrospectively reviewed. Among $_cT_{1-2}N_0M_0$ non small cell lung cancer patients, with intrapulmonary nodule(s), only those cases where a CT scan or a biopsy of the nodules had been conducted, and had been followed up at intervals of more than 6 months were included. However, patients who had received chemotherapy or radiation therapy, pre- or post-operatively, or with calcified nodules, were excluded. Results : Our study group consisted of 39 patients, divided into two groups. The first group, 33 patients, had benign nodules, and the second group, 6 patients, had malignant nodules. The two groups were compared with regard to gender, age, cell type, pathologic stage, shape, size, location and number of nodules and presence of calcification around the nodules. There was no statistically significant difference between the two groups. Conclusion : The intrapulmonary nodules in non small cell lung cancer patients at the $_cT_{1-2}N_0M_0$ stage were mostly benign. Therefore, surgical treatment for such patients can be considered. Moreover, without predictive factors, pathological confirmation of the diagnosed nodules should be sought in all patients.

• Journal of Korea Multimedia Society
• /
• v.23 no.6
• /
• pp.729-737
• /
• 2020

Screening with low-dose spiral computed tomography (LDCT) has been shown to reduce lung cancer mortality by about 20% when compared to standard chest radiography. One of the problems arising from screening programs is that large amounts of CT image data must be interpreted by radiologists. To solve this problem, automated detection of pulmonary nodules is necessary; however, this is a challenging task because of the high number of false positive results. Here we demonstrate detection of pulmonary nodules using six off-the-shelf convolutional neural network (CNN) models after modification of the input/output layers and end-to-end training based on publicly databases for comparative evaluation. We used the well-known CNN models, LeNet-5, VGG-16, GoogLeNet Inception V3, ResNet-152, DensNet-201, and NASNet. Most of the CNN models provided superior results to those of obtained using customized CNN models. It is more desirable to modify the proven off-the-shelf network model than to customize the network model to detect the pulmonary nodules.

• The Korean Journal of Cytopathology
• /
• v.6 no.2
• /
• pp.133-139
• /
• 1995

To evaluate the sensitivity and specificity of transthoracic fine needle aspiration cytology(TFNAC) in the preoperative diagnosis of pulmonary nodules, a retrospective analysis was carried out on a consecutive series of 200 TFNACs. They included 186 primary malignant tumors, 66 squamous cell carcinomas, 65 adenocarcinomas, 36 small cell carcinomas, 7 large ceil carcinomas, 4 carcinomas, 8 others, 9 metastatic tumors, and 5 benign tumors. On cytohistologic correlation of malignant pulmonary tumors, the procedure had a sensitivity of 97.3% and a specificity of 100%. A 86.6% correct correlation between the cytologic and histologic diagnoses was achieved. Five out of the 7 undifferentiated large cell carcinomas, 10 out of the 65 adenocarcinomas, 2 out of the 36 small cell carcinomas, and 2 out of the 66 squamous cell carcinomas were turned out to be mistyped in cytologic diagnosis. We concluded that TFNAC is a highly sensitive and specific preoperative diagnostic procedure in the investigation of patients with discrete pulmonary nodules in whom the specific ceil type of the malignant neoplasm has important implications in treatment modality and prognosis.