| Title |
Konvoliuciniai neuroniniai tinklai ankstyvam melanomos aptikimui naudojant dermatoskopinius vaizdus: literatūros apžvalga / |
| Translation of Title |
Convolutional neural networks for early melanoma detection using dermoscopic images: a literature review. |
| Authors |
Karmazinaitė, Martyna ; Šalaševičius, Rapolas |
| DOI |
10.53453/ms.2024.3.9 |
| Full Text |
|
| Is Part of |
Medicinos mokslai = Journal of medical sciences.. Kėdainiai : VšĮ Lietuvos sveikatos mokslinių tyrimų centras. 2024, vol. 12, iss. 2, p. 78-84.. ISSN 2345-0592 |
| Keywords [eng] |
melanoma ; convolutional neural networks ; patterns ; borders ; augmentation ; data preprocessing ; machine learning algorithms ; dermoscopy |
| Abstract [eng] |
Background: artificial intelligence (AI) is reshaping dermatology, offering faster and more accurate diagnoses, particularly for conditions like melanoma. Melanoma, a deadly form of skin cancer, requires early detection for effective treatment. Dermoscopy, aided by Convolutional Neural Networks (CNN), has emerged as a promising approach for early melanoma detection. As melanoma incidence rises, AI-driven solutions offer hope for more efficient diagnoses and improved patient outcomes. Aim: To analize the potential of convolutional neural networks (CNN) for early detection of melanoma using dermatoscopy images. We aim to review how CNN can effectively analyze and classify dermatoscopic images to accurately identify suspicious skin lesions. Methodology: articles were searched in the PubMed and Google Scholar databases. The search used keywords and their combinations, such as melanoma, convolutional neural networks, dermoscope. Articles whose title or keywords matched the purpose of this literature review were selected for analysis. Results: CNN have shown superior diagnostic accuracy compared to traditional methods. Collaborative approaches between CNN and dermatologists showing promise in improving diagnostic outcomes. Conclusions: the integration of artificial intelligence into dermatology is transforming the detection and diagnosis of skin diseases like melanoma. Convolutional neural networks have shown exceptional performance in analyzing dermoscopic images and distinguishing benign from malignant lesions. With melanoma incidence on the rise, leveraging artificial intelligence becomes increasingly essential for swift and accurate diagnoses globally, ultimately enhancing outcomes for patients battling the disease. |
| Published |
Kėdainiai : VšĮ Lietuvos sveikatos mokslinių tyrimų centras |
| Type |
Journal article |
| Language |
Lithuanian |
| Publication date |
2024 |
| CC license |
|
