Search

We aimed to develop an accurate and efficient skin cancer classification system using deep-learning technology with a relatively small dataset of clinical images. We proposed a novel skin cancer classification method, SkinFLNet, which utilizes model fusion and lifelong learning technologies. The SkinFLNet's deep convolutional neural networks were trained using a dataset of 1215 clinical images of skin tumors diagnosed at Taichung and Taipei Veterans General Hospital between 2015 and 2020. The dataset comprised five categories: benign nevus, seborrheic keratosis, basal cell carcinoma, squamous cell carcinoma, and malignant melanoma. The SkinFLNet's performance was evaluated using 463 clinical images between January and December 2021. SkinFLNet achieved an overall classification accuracy of 85%, precision of 85%, recall of 82%, F-score of 82%, sensitivity of 82%, and specificity of 93%, outperforming other deep convolutional neural network models. We also compared SkinFLNet's performance with that of three board-certified dermatologists, and the average overall performance of SkinFLNet was comparable to, or even better than, the dermatologists. Our study presents an efficient skin cancer classification system utilizing model fusion and lifelong learning technologies that can be trained on a relatively small dataset. This system can potentially improve skin cancer screening accuracy in clinical practice. [ABSTRACT FROM AUTHOR]

We aimed to develop an accurate and efficient skin cancer classification system using deep-learning technology with a relatively small dataset of clinical images. We proposed a novel skin cancer classification method, SkinFLNet, which utilizes model fusion and lifelong learning technologies. The SkinFLNet's deep convolutional neural networks were trained using a dataset of 1215 clinical images of skin tumors diagnosed at Taichung and Taipei Veterans General Hospital between 2015 and 2020. The dataset comprised five categories: benign nevus, seborrheic keratosis, basal cell carcinoma, squamous cell carcinoma, and malignant melanoma. The SkinFLNet's performance was evaluated using 463 clinical images between January and December 2021. SkinFLNet achieved an overall classification accuracy of 85%, precision of 85%, recall of 82%, F-score of 82%, sensitivity of 82%, and specificity of 93%, outperforming other deep convolutional neural network models. We also compared SkinFLNet's performance with that of three board-certified dermatologists, and the average overall performance of SkinFLNet was comparable to, or even better than, the dermatologists. Our study presents an efficient skin cancer classification system utilizing model fusion and lifelong learning technologies that can be trained on a relatively small dataset. This system can potentially improve skin cancer screening accuracy in clinical practice. [ABSTRACT FROM AUTHOR]

The coexistence of inflammatory bowel disease (IBD) and bullous pemphigoid (BP) has been reported. No large-scale study to date has explored the relationship between these diseases. This population-based case-control study examined the association between IBD and BP by using a nationwide database. A total of 5,263 BP patients and 21,052 age- and gender-, hospital visit number-matched controls were identified in the National Health Insurance Research Database of Taiwan (1997–2013). Demographic characteristics and comorbidities including IBD were compared. Logistic regression was conducted to examine the predicting factors for BP. The mean age at diagnosis was 74.88 years and 54.3% of subjects were male. BP patients tended to have more cardiovascular risk factors, autoimmune and neurologic comorbidities, and hematologic cancers than matched controls. There were 20 cases of IBD (0.38%), mostly ulcerative colitis (N = 17, 0.32%) among BP patients, compared to 33 IBD cases (0.16%) among controls (p < 0.001). Ulcerative colitis was found to be significantly associated with BP [adjusted odds ratio (OR) 3.60, 95% confidence interval (CI) 1.91–6.77, p < 0.001] on multivariate analysis. Treatment for IBD was not associated with BP development. Information about diet, lifestyle, alcohol consumption, and smoking habit was not available. We concluded that UC is independently associated with BP. [ABSTRACT FROM AUTHOR]

Background: An association between lichen simplex chronicus (LSC) and sexual dysfunction was explored. However, no data are available from investigations into the relationship between erectile dysfunction (ED) and LSC. Objectives: This retrospective population-based cohort study aimed to clarify the risk of ED associated with LSC. Methods: By using the Taiwan National Health Insurance Research dataset, we identified 5611 male patients who had been newly diagnosed with LSC from 2000 to 2004. The date of diagnosis was identified as the index date. LSC patients with incomplete demographic information or with a history of ED before the index date were excluded. In total, 22444 age-matched patients without LSC were randomly selected as the non-LSC group based on a 1:4 ratio. Subsequence occurrence of ED was measured until 2011. The association between LSC and the risk of developing ED was estimated using Cox proportional hazard regression model. Results: After adjusting for age and comorbidities, patients with LSC had a 1.74-fold greater risk of developing ED compared with those without LSC (95% confidence interval=1.44–2.10). LSC patients with comorbidities including diabetes, hyperlipidemia, hypertension, cardiovascular disease, peripheral arterial disease, chronic obstructive pulmonary disease, chronic kidney disease, depression, and anxiety were at a higher risk of ED compared with the non-LSC patients without comorbidities. Conclusions: LSC confers a greater risk in the development of ED. Physicians should be aware of the potential of ED occurrence in LSC patients. [ABSTRACT FROM AUTHOR]