Fluorescence images of skin lesions and automated diagnosis using convolutional neural networks
In recent years, interest in applying deep learning (DL) to medical diagnosis has rapidly increased, driven primarily by the development of Convolutional Neural Networks and Transformers. Despite advancements in DL, the automated diagnosis of skin cancer remains a significant challenge. Emulating de...
Saved in:
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-04-01
|
| Series: | Photodiagnosis and Photodynamic Therapy |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1572100024004988 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832583077628477440 |
|---|---|
| author | Matheus B. Rocha Sebastiao Pratavieira Renan Souza Vieira Juliana Duarte Geller Amanda Lima Mutz Stein Fernanda Sales Soares de Oliveira Tania R.P. Canuto Luciana de Paula Vieira Renan Rossoni Maria C.S. Santos Patricia H.L. Frasson Renato A. Krohling |
| author_facet | Matheus B. Rocha Sebastiao Pratavieira Renan Souza Vieira Juliana Duarte Geller Amanda Lima Mutz Stein Fernanda Sales Soares de Oliveira Tania R.P. Canuto Luciana de Paula Vieira Renan Rossoni Maria C.S. Santos Patricia H.L. Frasson Renato A. Krohling |
| author_sort | Matheus B. Rocha |
| collection | DOAJ |
| description | In recent years, interest in applying deep learning (DL) to medical diagnosis has rapidly increased, driven primarily by the development of Convolutional Neural Networks and Transformers. Despite advancements in DL, the automated diagnosis of skin cancer remains a significant challenge. Emulating dermatologists, deep learning approaches using clinical images acquired from smartphones and considering patient lesion information have achieved performance levels close to those of specialists. While including clinical information, such as whether the lesion bleeds, hurts, or itches, improves diagnostic metrics, it is insufficient for correctly differentiating some major skin cancer lesions. An alternate technology for diagnosing skin cancer is fluorescence widefield imaging, where the skin lesion is illuminated with excitation light, causing it to emit fluorescence. Since there is no public dataset of fluorescence images for skin lesions, to the best of our knowledge, an effort has been made and resulted in 1,563 fluorescence images of major skin lesions taken by smartphones using the handheld LED wieldfield fluorescence device. The collected images were annotated and analyzed, creating a new dataset named FLUO-SC. Convolutional neural networks were then applied to classify skin lesions using these fluorescence images. Experimental results indicate that fluorescence images are competitive with clinical images (baseline) for classifying major skin lesions and show promising potential for discrimination. |
| format | Article |
| id | doaj-art-35e4bcf879f8472a9adae674e669ff45 |
| institution | Kabale University |
| issn | 1572-1000 |
| language | English |
| publishDate | 2025-04-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Photodiagnosis and Photodynamic Therapy |
| spelling | doaj-art-35e4bcf879f8472a9adae674e669ff452025-01-29T05:00:23ZengElsevierPhotodiagnosis and Photodynamic Therapy1572-10002025-04-0152104462Fluorescence images of skin lesions and automated diagnosis using convolutional neural networksMatheus B. Rocha0Sebastiao Pratavieira1Renan Souza Vieira2Juliana Duarte Geller3Amanda Lima Mutz Stein4Fernanda Sales Soares de Oliveira5Tania R.P. Canuto6Luciana de Paula Vieira7Renan Rossoni8Maria C.S. Santos9Patricia H.L. Frasson10Renato A. Krohling11Labcin - Nature Inspired Computing Laboratory, Federal University of Espírito Santo, Vitória, Brazil; Graduate Program in Computer Science, Federal University of Espírito Santo, Vitória, Brazil; Correspondence to: Nature-inspired Computing Lab, CT-X, Federal University of Espirito Santo, Vitória, ES, CEP 29075-910, Brazil.Sao Carlos Institute of Physics, University of São Paulo, São Carlos, BrazilDermatological Assistance Program (PAD), Federal University of Espírito Santo, Vitória, BrazilDermatological Assistance Program (PAD), Federal University of Espírito Santo, Vitória, BrazilDermatological Assistance Program (PAD), Federal University of Espírito Santo, Vitória, BrazilDermatological Assistance Program (PAD), Federal University of Espírito Santo, Vitória, BrazilDermatological Assistance Program (PAD), Federal University of Espírito Santo, Vitória, Brazil; Secretary of Health of the Espírito Santo State, Governor of Espírito Santo state, Vitória, BrazilDermatological Assistance Program (PAD), Federal University of Espírito Santo, Vitória, Brazil; Secretary of Health of the Espírito Santo State, Governor of Espírito Santo state, Vitória, BrazilDermatological Assistance Program (PAD), Federal University of Espírito Santo, Vitória, BrazilPathological Anatomy Unit of the University Hospital Cassiano Antônio Moraes (HUCAM), Federal University of Espírito Santo, Vitória, BrazilDepartment of Specialized Medicine, Federal University of Espírito Santo, Vitória, Brazil; Dermatological Assistance Program (PAD), Federal University of Espírito Santo, Vitória, BrazilLabcin - Nature Inspired Computing Laboratory, Federal University of Espírito Santo, Vitória, Brazil; Graduate Program in Computer Science, Federal University of Espírito Santo, Vitória, BrazilIn recent years, interest in applying deep learning (DL) to medical diagnosis has rapidly increased, driven primarily by the development of Convolutional Neural Networks and Transformers. Despite advancements in DL, the automated diagnosis of skin cancer remains a significant challenge. Emulating dermatologists, deep learning approaches using clinical images acquired from smartphones and considering patient lesion information have achieved performance levels close to those of specialists. While including clinical information, such as whether the lesion bleeds, hurts, or itches, improves diagnostic metrics, it is insufficient for correctly differentiating some major skin cancer lesions. An alternate technology for diagnosing skin cancer is fluorescence widefield imaging, where the skin lesion is illuminated with excitation light, causing it to emit fluorescence. Since there is no public dataset of fluorescence images for skin lesions, to the best of our knowledge, an effort has been made and resulted in 1,563 fluorescence images of major skin lesions taken by smartphones using the handheld LED wieldfield fluorescence device. The collected images were annotated and analyzed, creating a new dataset named FLUO-SC. Convolutional neural networks were then applied to classify skin lesions using these fluorescence images. Experimental results indicate that fluorescence images are competitive with clinical images (baseline) for classifying major skin lesions and show promising potential for discrimination.http://www.sciencedirect.com/science/article/pii/S1572100024004988Skin cancer detectionFluorescence imagingDeep learningSmartphone-based imagingClinical images |
| spellingShingle | Matheus B. Rocha Sebastiao Pratavieira Renan Souza Vieira Juliana Duarte Geller Amanda Lima Mutz Stein Fernanda Sales Soares de Oliveira Tania R.P. Canuto Luciana de Paula Vieira Renan Rossoni Maria C.S. Santos Patricia H.L. Frasson Renato A. Krohling Fluorescence images of skin lesions and automated diagnosis using convolutional neural networks Photodiagnosis and Photodynamic Therapy Skin cancer detection Fluorescence imaging Deep learning Smartphone-based imaging Clinical images |
| title | Fluorescence images of skin lesions and automated diagnosis using convolutional neural networks |
| title_full | Fluorescence images of skin lesions and automated diagnosis using convolutional neural networks |
| title_fullStr | Fluorescence images of skin lesions and automated diagnosis using convolutional neural networks |
| title_full_unstemmed | Fluorescence images of skin lesions and automated diagnosis using convolutional neural networks |
| title_short | Fluorescence images of skin lesions and automated diagnosis using convolutional neural networks |
| title_sort | fluorescence images of skin lesions and automated diagnosis using convolutional neural networks |
| topic | Skin cancer detection Fluorescence imaging Deep learning Smartphone-based imaging Clinical images |
| url | http://www.sciencedirect.com/science/article/pii/S1572100024004988 |
| work_keys_str_mv | AT matheusbrocha fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT sebastiaopratavieira fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT renansouzavieira fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT julianaduartegeller fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT amandalimamutzstein fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT fernandasalessoaresdeoliveira fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT taniarpcanuto fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT lucianadepaulavieira fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT renanrossoni fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT mariacssantos fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT patriciahlfrasson fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks AT renatoakrohling fluorescenceimagesofskinlesionsandautomateddiagnosisusingconvolutionalneuralnetworks |