• 한국멀티미디어학회논문지
• /
• 제25권8호
• /
• pp.1212-1223
• /
• 2022

Skin lesions are common diseases that range from skin rashes to skin cancer, which can lead to death. Note that early diagnosis of skin diseases can be important because early diagnosis of skin diseases considerably can reduce the course of treatment and the harmful effect of the disease. Recently, the development of computer-aided diagnosis (CAD) systems based on artificial intelligence has been actively made for the early diagnosis of skin diseases. In a typical CAD system, the accurate classification of skin lesion types is of great importance for improving the diagnosis performance. Motivated by this, we propose a novel deep ensemble classification with multi-scale attention networks. The proposed deep ensemble networks are jointly trained using a single loss function in an end-to-end manner. In addition, the proposed deep ensemble network is equipped with a multi-scale attention mechanism and segmentation information of the original skin input image, which improves the classification performance. To demonstrate our method, the publicly available human skin disease dataset (HAM 10000) and the private animal skin lesion dataset were used for the evaluation. Experiment results showed that the proposed methods can achieve 97.8% and 81% accuracy on each HAM10000 and animal skin lesion dataset. This research work would be useful for developing a more reliable CAD system which helps doctors early diagnose skin diseases.

• 디지털산업정보학회논문지
• /
• 제14권4호
• /
• pp.69-77
• /
• 2018

The malignant melanoma accounts for about 1 to 3% of the total malignant tumor in the West, especially in the US, it is a disease that causes more than 9,000 deaths each year. Generally, skin lesions are difficult to detect the features through photography. In this paper, we propose a computer-aided diagnosis algorithm based on deep learning for classification of malignant melanoma and benign skin tumor in RGB channel skin images. The proposed deep learning model configures the tumor lesion segmentation model and a classification model of malignant melanoma. First, U-Net was used to segment a skin lesion area in the dermoscopic image. We could implement algorithms to classify malignant melanoma and benign tumor using skin lesion image and results of expert's labeling in ResNet. The U-Net model obtained a dice similarity coefficient of 83.45% compared with results of expert's labeling. The classification accuracy of malignant melanoma obtained the 83.06%. As the result, it is expected that the proposed artificial intelligence algorithm will utilize as a computer-aided diagnosis algorithm and help to detect malignant melanoma at an early stage.

• Archives of Plastic Surgery
• /
• 제41권5호
• /
• pp.500-504
• /
• 2014

Background Skin cancer is the most prevalent cancer by organ type and referral accuracy is vital for diagnosis and management. The British Association of Dermatologists (BAD) and literature highlight the importance of accurate skin lesion examination, diagnosis and educationally-relevant studies. Methods We undertook a review of the relevant literature, a national audit of skin lesion description standards and a study of speciality training influences on these descriptions. Questionnaires (n=200), with pictures of a circular and an oval lesion, were distributed to UK dermatology/plastic surgery consultants and speciality trainees (ST), general practitioners (GP), and medical students (MS). The following variables were analysed against a pre-defined 95% inclusion accuracy standard: site, shape, size, skin/colour, and presence of associated scars. Results There were 250 lesion descriptions provided by 125 consultants, STs, GPs, and MSs. Inclusion accuracy was greatest for consultants over STs (80% vs. 68%; P<0.001), GPs (57%) and MSs (46%) (P<0.0001), for STs over GPs (P<0.010) and MSs (P<0.0001) and for GPs over MSs (P<0.010), all falling below audit standard. Size description accuracy sub-analysis according to circular/oval dimensions was as follows: consultants (94%), GPs (80%), STs (73%), MSs (37%), with the most common error implying a quadrilateral shape (66%). Addressing BAD guidelines and published requirements for more empirical performance data to improve teaching methods, we performed a national audit and studied skin lesion descriptions. To improve diagnostic and referral accuracy for patients, healthcare professionals must strive towards accuracy (a circle is not a square). Conclusions We provide supportive evidence that increased speciality training improves this process and propose that greater focus is placed on such training early on during medical training, and maintained throughout clinical practice.

• 동의생리병리학회지
• /
• 제27권2호
• /
• pp.167-172
• /
• 2013

In order to establish the diagnostic indices of skin diseases, some physiological and pathological concepts of interstices(腠理) were researched based on , and western dermatology with etymological analysis. As physiological indices for diagnosis of skin diseases, measurement of epidermis and dermis using ultrasonogram in the zhongwan and dachui's location. And for grading looseness and fineness of interstices with 3 unsound groups, measuring numbers and sizes of sweat pores in each point's 1 cm diameter circular area using comparative method and palpation assessment. Another index is superficial temperature. As pathological indices for diagnosis, validating volumes of dead skin cells and grading degrees of atrophy and degeneration of skin lesion. And as supplementary measures, absorptive degrees of cosmetics on face should be recorded according to 3 grades. These diagnostic indices can contribute to establishment of standard pattern identification and prescription of skin diseases through converting anatomical cognizance into classical concepts of interstices objectively.

• Journal of Multimedia Information System
• /
• 제1권1호
• /
• pp.77-85
• /
• 2014

Usually, the skin pigmentation detection and diagnosis are made by clinicians. In this process it is subjective and non-quantitative. We develop an approach to detect and measure the different pigmentation lesions base on computer vision technology. In the paper we study several usually used skin-detecting color space like HSV, YCbCr and normalized RGB. We compare their performance with illumination influence for detecting the pigmentation lesions better. Base on a relatively stable color space, we propose an approach which is RGB channels vector difference characteristic for the detection. After the object region detection, we also use the difference to measure the difference between the lesion and the surrounding normal skin. From the experiment results, our approach can effectively detect the pigmentation lesion, and perform robustness with different illumination.

• 대한두경부종양학회지
• /
• 제21권2호
• /
• pp.174-177
• /
• 2005

Skin metastasis from papillary thyroid carcinoma is extremely rare. Due to similar histopathologic features, it is difficult to differentiate skin metastatic papillary thyroid carcinoma and some primary skin neoplasms without a clinical history. However, most of metastatic skin lesions showed a strong reactivity to the antithyroglobulin antibodies unlike primary skin neoplasms. Metastatic skin lesions must be completely removed and radiotherapy can be added. Investigators reported that prognosis of skin metastases from thyroid carcinoma is dismal and the average survival after it's diagnosis was only 19 months because distant metastases were often discovered at diagnosis of skin metastasis or during follow-up period. We report a case of skin metastasis from tall cell variant of papillary thyroid carcinoma. In our case, the anti thyroglobulin antibodies measured from cystic fluid from a skin lesion was more than 2000 IU/ml. Skin metastasis was diagnosed at 20 months after primary surgery for thyroid cancer and brain metastasis at 12 months after diagnosis of skin metastasis. Although skin metastasis is an ominous prognostic indicator in patients with thyroid carcinoma, a radical treatment for skin lesions and early diagnosis of distant metastasis could provide a chance to the patients to improve their survival.

• 대한두개안면성형외과학회지
• /
• 제13권1호
• /
• pp.76-79
• /
• 2012

Purpose: Sebaceous epithelioma (sebaceoma) is a benign tumor with sebaceous differentiation. It presents primarily as a yellowish papule or nodule on the face and scalp. It must be differentiated from basal cell carcinoma and other appendageal tumors. We report a giant sebaceous epithelioma on the scalp and describe the immunohistochemical character of the cells in sebaceous epithelioma to epithelial membrane antigen (EMA). Methods: A 55-year-old-man who presented with 5-cm-diameter 2-cm-height, round shape exophytic ulcerated tumor on his head presented for treatment. The patient had noticed the lesion 40 years prior as a small yellowish plaque and 18 months ago, the plaque started to grow progressively larger. We excised the lesion with 1 cm resection margin, considering the possibility of malignancy because this lesion grossly resembled basal cell carcinoma (BCC). The defect was repaired with the use of a splitthickness skin graft. Results: When we excised the lesion, the margin was clear. Histology showed nodules that consisted of an admixture of basaloid cells and mature adipocytes lacking an organized lobular architecture. Strong expression of EMA on mature adipose cells confirmed the differential diagnosis from BCC with sebaceous differentiation because of the absence of a nuclear palisade pattern and cleft-like spaces on the hematoxylin and eosin (H&E) section. Conclusion: We treated the giant sebaceous epithelioma on the scalp with surgical excision and a split-thickness skin graft. It is important to know that the diagnosis of sebaceous epithelioma should be made based on the histologic pattern of the H&E section. Immunohistochemistry with EMA can help to confirm the differential diagnosis between sebaceous epithelioma and BCC.

• 대한수의학회지
• /
• 제55권2호
• /
• pp.141-143
• /
• 2015

Dermatophytosis was found on the right front leg of a 4-month-old female African lion cub (Panthera leo) kept at a zoo with locally marginal alopecia. For diagnosis, culturing on sabouraud dextrose agar was performed and skin scrapings from the lesion were analyzed. The ones from the culture and skin scrapings were identified as Trichophyton mentagrophytes. A zookeeper that had been in contacted with the lion for artificial rearing developed skin lesions with well-defined erythematous plaques on the right arm about 1 month after the lesion in the lion was observed. The ringworm was probably transmitted from the lion through continuous contact.

• Asian Pacific Journal of Cancer Prevention
• /
• 제16권8호
• /
• pp.3403-3406
• /
• 2015

Background: Skin biopsy is the method to assist clinicians to make definite dermatological diagnosis which further helps in holistic management. Skin cancers are relatively rare clinical diagnosis in developing countries like Nepal, but the prevalence is on rise. Objectives: To investigate the profile of skin biopsies and frequencies and pattern of skin cancers in a tertiary care centre of Western Nepal. Materials and Methods: The materials consisted of 434 biopsies (1.37%) out of 31,450 OPD visits performed in the Department of Dermatology, Manipal Teaching Hospital, Pokhara, Nepal, during the period of Dec 2011-Nov 2014. Data were collected and analyzed using SPSS-16 with reference to incidence, age, sex, race and clinical and histopathological features. Results: The commonest disorders observed in biopsies were papulosquamous lesions, skin tuberculosis of different types, benign skin tumors, leprosy, collagen and fungal diseases. Viral diseases were rarely seen, probably due to straight forward clinical diagnosis. Dermatological malignancies accounted for 55/434 (12.67%) of biopsies. Skin disorders in general were commoner in females 280/434 (64%), including malignancies 32/55(58.2%). Mean age of patients with skin cancer was 54.5 years. Facilities for proper laboratory investigation of dermatological disorders will improve the quality of life. Conclusions: The most prevalent lesion in skin biopsies was papulosquamous disorders followed by skin tuberculosis of different types. Dermatological malignancy constituted 55/434 (12.67%) cases. The prevalence of skin malignancy is on rise in Nepalese society probably due to increase in life expectancy and better diagnostic services.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• 제27권6호
• /
• pp.570-574
• /
• 2005

Keratoacanthoma is a benign, self-limited epithelial lesion that closely resembles Squamous cell carcinoma(SCC). Keratoacanthoma occur primarily exposed skin in male patients over 45 years of ages. although etiology is unknown, sunlight, genetic, and human papillomavirus factor have been considered. in clinical feature, rapid enlargement occurs over 4$\sim$8 weeks, resulting ultimately in a hemispheric, firm, elevated, asymptomatic nodule that contains a central plug of keratin. When fully developed, the keratoacanthoma contains a core of keratin surrounded by a concentric collar of raised skin. Over the next 4$\sim$8 weeks, static lesion persists. Then undergoes spontaneous regression over the next 6$\sim$8weeks period by expulsion of the keratin core with resorption of the mass. In histologic feature, Keratoacanthoma consists of hyperplastic squamous epithelium growing into the underlying connective tissue. The surface is covered by a thickened layer of parakeratin with central plugging. Epithelium cell shows dysplastic features and the margins the normal adjacent epithelium is elevated. The differential diagnosis includes SCC. Keratoacanthoma present as a exophytic lesion with horny keratin occupying a depression on the top of the lesion, persists static period and undergoes rapid growth compared with SCC. Keratoacanthoma is usually treated by surgical excision or curettage of the base, spontaneous regression does not occur in every case. A 60 years old male who present facial lesion visit our hospital and surgical excision was done. Biopsy result was keratoacanthoma. We report case with review of literatures.