P

Skin Disease Detection

1. Vishwakarma Institute of Technology Pune, Maharashtra, India

[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]

Abstract. The application of deep learning methods to diagnose diseases has become a new
research topic in the medical field. In the field of medicine, skin disease is one of the most common
diseases, and its visual representation is more prominent compared with the other types of diseases.
Accordingly, the use of deep learning methods for skin disease image recognition is of great
significance and has attracted the attention of researchers. Due to lack of medical facilities
available in the remote areas, patients usually ignore early symptoms which may worsen the
situation as time progresses. Hence, there is a rising need for automatic skin disease detection
system with high accuracy. Thus, we develop a multiclass deep learning model to differentiate
between Healthy Skin Vs Skin suffering from a Disease and Classification of Skin Diseases The
advent of deep learning methods has ushered in a transformative era in medical research,
particularly in the context of disease diagnosis. Within the medical landscape, skin diseases
represent a prevalent health concern, often distinguished by their prominent visual manifestations.
Consequently, the application of deep learning techniques to facilitate accurate skin disease image
recognition has garnered substantial attention from the research community. Compounded by the
lack of accessible medical facilities in remote regions, early symptoms of skin diseases often go
unnoticed by patients, potentially exacerbating their conditions over time. Consequently, the
imperative for a high-precision automated skin disease detection system has become increasingly
apparent.
In light of these challenges, we present a comprehensive multiclass deep learning model
designed to discern between healthy skin, skin afflicted by diseases, and the classification of
specific skin diseases. Our approach integrates a robust dataset, meticulously curated and
preprocessed to ensure optimal model performance. Leveraging a carefully constructed deep
learning architecture, our model achieves notable accuracy and efficiency in distinguishing various
skin conditions. Through extensive experimentation and evaluation, we demonstrate the efficacy of
our proposed model, highlighting its ability to accurately classify and diagnose a range of common
skin diseases.
This research contributes significantly to the ongoing efforts to enhance early detection and
intervention in skin disease management, particularly in underserved regions. By presenting a
reliable and accessible automated system, we aim to mitigate the challenges associated with timely
diagnosis and treatment, ultimately improving patient outcomes and alleviating the Pburden on
healthcare systems.
Keywords— Deep Learning , Skin Disease Diagnosis ,Medical Image Recognition ,Multiclass
Classification , Healthcare Technology , Automated Diagnosis , Remote Healthcare , Disease
Detection System , Image Analysis , Medical Informatics
1 Introduction

In recent years, the application of deep learning techniques has catalyzed a revolution in the
medical field, offering a promising avenue for enhanced disease diagnosis and management.
Among the myriad health concerns, skin diseases have emerged as a significant focus, primarily
due to their prevalence and the prominent visual cues they present. Dermatological conditions,
ranging from common ailments such as eczema and psoriasis to more severe afflictions like
melanoma, significantly impact the global population, often resulting in substantial morbidity and
decreased quality of life if left undiagnosed or untreated. This has underscored the critical need for
efficient and accurate methods to diagnose and differentiate between various skin diseases.

In addition to the challenges posed by the diverse nature of skin ailments, the accessibility of
healthcare facilities remains a persistent issue, particularly in remote and underserved areas.
Patients in these regions often lack timely access to specialized dermatologists or modern
diagnostic technologies, leading to delayed or inaccurate diagnoses. As a result, early symptoms of
skin diseases are frequently overlooked, potentially exacerbating the severity of the conditions and
complicating treatment regimens. To address these limitations and bridge the gap in accessible
healthcare, there is an urgent demand for reliable and user-friendly automated systems capable of
accurately identifying and classifying various skin diseases.

Building upon these imperatives, this research aims to develop a comprehensive multiclass deep
learning model that can effectively differentiate between healthy skin, skin suffering from diseases,
and various specific types of skin diseases. By harnessing the power of deep learning algorithms,
we seek to create a robust and reliable tool that can aid in the early detection and accurate
classification of a wide spectrum of dermatological conditions. Through the integration of
advanced image recognition techniques and a meticulously curated dataset, our model endeavors to
provide an accessible and effective solution, particularly for regions with limited access to
specialized medical care.

Previous research efforts have laid the groundwork for utilizing deep learning in medical image
analysis, showcasing promising results in diverse fields, including radiology, pathology, and
ophthalmology. Notably, studies have demonstrated the efficacy of convolutional neural networks
(CNNs) in accurately diagnosing and classifying various diseases based on medical imaging data.
For instance, Esteva et al. (2017) demonstrated the potential of deep learning models in the
automated classification of skin cancer, yielding results comparable to experienced dermatologists.
Similarly, Rajpurkar et al. (2018) showcased the applicability of deep learning in detecting
pathologies from chest radiographs, highlighting the potential for accurate disease identification
and classification.

Despite these advancements, the unique challenges posed by skin disease diagnosis warrant a
specialized approach tailored to the complexities of dermatological conditions. In contrast to
internal diseases, skin diseases are often characterized by their external manifestations,
necessitating a nuanced understanding of visual cues and subtle differentiations in skin textures,
colors, and patterns. This calls for a customized deep learning framework that can effectively
discern these intricate visual differentiations, enabling precise and reliable classification of various
skin ailments.
 P
In this context, our research endeavors to contribute to the growing body of knowledge surrounding
the application of deep learning in dermatology. By developing a specialized multiclass deep
learning model, we seek to advance the frontier of automated skin disease diagnosis, particularly in
regions with limited access to specialized healthcare services. Through the integration of advanced
image recognition techniques and a comprehensive dataset encompassing a diverse range of skin
conditions, our proposed model aspires to offer a user-friendly and accessible solution, facilitating
timely and accurate diagnosis for a multitude of patients.

2 Literature Survey:
Skin diseases are the 4th common cause of skin burden worldwide. Robust and
Automated system have been developed to lessen this burden and to help the patients
to conduct the early assessment of

3:Referance:
References:

[1] Naziahameed et al. (2020): Multi-Class Skin Diseases Classification Using Deep
Convolutional Neural Network and Support Vector Machine, 2018 12th International
Conference on Software, Knowledge, Information Management & Applications (SKIMA)

[2] Parvathaneni et al. (2021): Classification of Skin Disease Using Deep Learning
Neural Networks with MobileNet V2 and LSTM, https://www.mdpi.com/1424-
8220/21/8/2852

[3] Li, H., Chen, Z., Wu, Y., & Zhang, X. (2022). A Skin Disease Classification Model
Based on DenseNet and ConvNeXt Fusion. https://www.mdpi.com/2079-9292/12/2/438

[4] Nasiriavanaki, M., Sengupta, S., & Yu, G. (2020). Artificial intelligence-powered skin
disease diagnosis: opportunities and challenges. Journal of Investigative Dermatology,
140(11), 1883-1886.

[5] Codella, N., Gutman, D., Celebi, M. E., Mendonça, B., Marques, O., Kittler, H., ... &
Blum, A. (2018). Skin lesion analysis towards melanoma detection: A challenge dataset
and baseline method. IEEE transactions on medical imaging, 38(7), 1662-1674.

[6] Zhou, Y., Cao, M., Liu, C., Zheng, Y., & Liu, X. (2019). DermaNet: A convolutional
neural network for large-scale skin lesion classification. arXiv preprint arXiv:1905.06739.

[7] Tschandl, P., Rosendahl, C., & Kittler, H. (2020). The HAM10000 dataset: A large
collection of multi-source dermatoscopic images of common pigmented skin lesions.
Scientific Data, 7(1), 1-9

[8] Shamsul Arifin, M., Golam Kibria, M., Firoze, A., Ashraful Amini, M., & Hong Yan.
(2012). Dermatological disease diagnosis using color skin images. 2012 International
Conference on Machine Learning and Cybernetics. doi:10.1109/icmlc.2012.6359626).

[9] Jana, E., Subban, R., & Saraswathi, S. (2017). Research on Skin Cancer Cell
Detection Using Image Processing. 2017 IEEE International Conference on
Computational Intelligence and Computing Research (ICCIC).
doi:10.1109/iccic.2017.8524554.

[10] Mhaske, H. R., & Phalke, D. A. (2013). Melanoma skin cancer detection and
classification based on supervised and unsupervised learning. 2013 International P
Conference on Circuits, Controls and Communications (CCUBE).
doi:10.1109/ccube.2013.6718539.

[11] Alfed, N., Khelifi, F., Bouridane, A., & Seker, H. (2015). Pigment network-based
skin cancer detection. 2015 37th Annual International Conference of the IEEE
Engineering in Medicine and Biology Society (EMBC). doi:10.1109/embc.2015.7320056.
[12] Lau, H. T., & Al-Jumaily, A. (2009). Automatically Early Detection of Skin Cancer:
Study Based on Nueral Netwok Classification. 2009 International Conference of Soft
Computing and Pattern Recognition. doi:10.1109/socpar.2009.80.