THE ANALYSIS STUDY OF ARTIFICIAL INTELIGENCE FOR SKIN CANCER: A COMPREHENSIVE SYSTEMATIC REVIEW

Background: Skin cancer diagnosis relies heavily on the interpretation of visual patterns, making it a complex task that requires extensive training in dermatology and dermatoscopy. However, AI algorithms have been shown to accurately diagnose skin cancers, even outperforming experienced dermatologists in image classification tasks in constrained settings. The aim: The aim of this study to show about artificial intelligence for skin cancer. Methods: By the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) 2020, this study was able to show that it met all of the requirements. This search approach, publications that came out between 2014 and 2024 were taken into account. Several different online reference sources, like Pubmed, SagePub, and Science Direct were used to do this. It was decided not to take into account review pieces, works that had already been published, or works that were only half done. Result: Eight publications were found to be directly related to our ongoing systematic examination after a rigorous three-level screening approach. Subsequently, a comprehensive analysis of the complete text was conducted, and additional scrutiny was given to these articles. Conclusion: The use of AI has high potential to facilitate the way skin cancer is diagnosed. Two main branches of AI are used to detect and classify skin cancer, namely shallow and deep techniques.


INTRODUCTION
Artificial intelligence (AI) stands at the forefront of technological innovation and has permeated into almost every industry and field.In dermatology, significant progress has been made toward the application of AI in skin cancer screening and diagnosis.Notably, a milestone that marked the era of modern artificial intelligence in dermatology was the demonstration of skin cancer classification abilities by deep learning convolutional neural networks (CNNs), which was on par with the performance of board-certified dermatologists.This CNN was trained on a dataset that was two orders of magnitude greater than those previously utilized.The dermatologist-level classification ability has since been experimentally validated by other papers.Recent progress in the field of AI enables models to not only analyze image data but also integrate clinical information, including patient demographics and past medical history.[3] Skin cancer is the abnormal growth of skin cells.The cancerous growth may affect both the layers-dermis and epidermis, but this review is concerned primarily with epidermal skin cancer; the two types of skin cancers that can arise from the epidermis are carcinomas and melanomas, depending on their cell type keratinocytes or melanocytes, respectively.It is a challenge to estimate the incidence of skin cancer due to various reasons, such as the multiple sub-types of skin cancer.This poses as a problem while collating data, as non-melanoma is often not tracked by registries or are left incomplete because most cases are treated via surgery.As of 2020, the World Cancer Research Fund International reported a total of 300,000 cases of melanoma in skin, and a total of 1,198,073 cases of non-melanoma skin cancer.The reasons for the occurrence of skin cancer cannot be singled out, but they include and are not limited to exposure to ultraviolet rays, family history, or a poor immune system.[6] Cancer is one of the major healthcare burdens across the world.Global statistics suggest almost 10.0 million deaths (9.9 million excluding non-melanoma skin cancer) due to cancer in the year 2020.The most commonly diagnosed cancers include breast cancer in females, lung cancer, and prostate cancers.Lung, liver, and stomach cancers are the major contributors of cancer related deaths.Skin cancer, including both malignant melanoma and non-melanoma skin cancer (NMSC), are common cancers in Caucasians and their incidence is on the rise.According to the US Skin Cancer Foundation, skin cancer affects more people in the United States each year than all other cancers combined. 7lanoma is the skin cancer with the worst prognosis.If diagnosed early, it can be treated successfully with surgical procedures.However, once there is metastasis, rates of survival are reduced significantly.Diagnosis of melanoma depends on the clinical examination and classic findings on the lesion biopsy.Examples of NMSC include basal cell carcinoma (NMSC) and squamous cell carcinoma.The success of skin cancer depends on early diagnosis and appropriate treatment.Visual inspection may not be sufficient to differentiate benign lesions from malignant tumors.The gold standard procedure is histopathology examination of the skin biopsy.The invasive nature of the procedure, associated pain, and the need for repeated samples in suspected lesions with varied presentations are some of the limitations for skin biopsy.[9]

METHODS PROTOCOL
By following the rules provided by Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) 2020, the author of this study made certain that it was up to par with the requirements.This is done to ensure that the conclusions drawn from the inquiry are accurate.

CRITERIA FOR ELIGIBILITY
For the purpose of this literature review, we compare and contrast artificial intelligence for skin cancer.It is possible to accomplish this by researching of artificial intelligence for skin cancer.As the primary purpose of this piece of writing, demonstrating the relevance of the difficulties that have been identified will take place throughout its entirety.
In order for researchers to take part in the study, it was necessary for them to fulfil the following requirements: 1) The paper needs to be written in English, and it needs to determine about artificial intelligence for skin cancer.In order for the manuscript to be considered for publication, it needs to meet both of these requirements.2) The studied papers include several that were published after 2014, but before the time period that this systematic review deems to be relevant.Examples of studies that are not permitted include editorials, submissions that do not have a DOI, review articles that have already been published, and entries that are essentially identical to journal papers that have already been published.

SEARCH STRATEGY
We used " Artificial intelligence for skin cancer."as keywords.The search for studies to be included in the systematic review was carried out using the PubMed, SagePub, and Science Direct databases by inputting the words:

DATA RETRIEVAL
After reading the abstract and the title of each study, the writers performed an examination to determine whether or not the study satisfied the inclusion criteria.The writers then decided which previous research they wanted to utilise as sources for their article and selected those studies.After looking at a number of different research, which all seemed to point to the same trend, this conclusion was drawn.All submissions need to be written in English and cannot have been seen anywhere else.

Figure 1. Article search flowchart
Only those papers that were able to satisfy all of the inclusion criteria were taken into consideration for the systematic review.This reduces the number of results to only those that are pertinent to the search.We do not take into consideration the conclusions of any study that does not satisfy our requirements.After this, the findings of the research will be analysed in great detail.The following pieces of information were uncovered as a result of the inquiry that was carried out for the purpose of this study: names, authors, publication dates, location, study activities, and parameters.

QUALITY ASSESSMENT AND DATA SYNTHESIS
Each author did their own study on the research that was included in the publication's title and abstract before making a decision about which publications to explore further.The next step will be to evaluate all of the articles that are suitable for inclusion in the review because they match the criteria set forth for that purpose in the review.After that, we'll determine which articles to include in the review depending on the findings that we've uncovered.This criteria is utilised in the process of selecting papers for further assessment.in order to simplify the process as much as feasible when selecting papers to evaluate.Which earlier investigations were carried out, and what elements of those studies made it appropriate to include them in the review, are being discussed here.

RESULT
Using reputable resources like Science Direct, PubMed, and SagePub, our research team first gathered 1702 publications.A thorough three-level screening strategy was used to identify only eight papers as directly relevant to our ongoing systematic evaluation.Next, a thorough study of the entire text and further examination of these articles were selected.Table 1 compiles the literature that was analyzed for this analysis in order to make it easier to view.

DISCUSSION
Artificial intelligence (AI) is transforming health care.Deep learning (DL) has become the dominant AI technology for high-dimensional complex data, such as images.In brief, DL leverages artificial neural networks, which learn complex mappings between inputs (e.g., images) and outputs (e.g., diagnoses) without explicit human engineering.Inspired by the brain, artificial neurons arranged in deep layers adapt the strength of their connections to one another as the model selflearns features from the input, such as visual patterns, that are most relevant for predicting the output.In experimental settings across multiple specialties, DL performs equivalently to health-care professionals for detecting disease from medical imaging. 18,19odern machine learning techniques rely on vast datasets to identify patterns useful for classification.Diagnostic imaging represents one of the most promising arenas for AI research.Skin cancer detection in particular serves as an appealing application for AI, given that diagnoses often hinge on the subjective visual interpretation of clinical and dermoscopic images.AI-assisted diagnosis promises several advantages.For instance, AI could improve access to specialist-level expertise.The scarcity of dermatologists is a serious problem in many regions, often leading to protracted waiting times for specialist appointments.In addition, there is growing optimism that AI-based systems might offer greater consistency and higher accuracy than human experts.First demonstrated the efficacy of convolutional neural networks (CNNs) for the task of image-based classification in dermatology.CNNs are specialized types of neural networks that are optimally suited for image analysis and are predominantly trained using supervised learning techniques. 20,21in cancer is the most common form of cancer worldwide.Over the past decade, there has been a concerning 27% increase in the annual diagnosis of invasive melanoma cases.Alarmingly, more than 5,400 people die from non-melanoma skin cancer every month.In the United States alone, the annual financial burden of treating skin cancer is estimated at a staggering US$8.1 billion, with approximately US$4.8 billion allocated to non-melanoma skin cancer and US$3.3 billion to melanoma.Among skin cancer types, basal cell carcinoma ranks as the most common, followed by squamous cell carcinoma and melanoma, which stands out as the most aggressive and lethal type of skin cancer.Merkel cell carcinoma also stands out among aggressive tumors.These tumors can arise anywhere on the body but are frequently observed in regions more exposed to the sun, including the face, neck, arms, and hands.[24] The deployment of this AI skin app at a broad scale shows the real-world costs of more false positives (benign lesion claims) and fewer true positives (malignant lesion claims) compared to the management of non-app users.More false positives and fewer true positives compared with conventional care can take an emotional and financial toll on patients and the healthcare system. 25e overall cost-effectiveness of the screening may be comparable to that of a dermatologist.A recent study in the US found that the cost of detecting an additional skin premalignancy or malignancy through total body exams was $2346.Depending on the assumptions of these calculations, the skin app performed at a comparable cost per new positive identification.In context, increased total costs per app user at a comparable cost-benefit ratio suggests that the app users are enjoying more of the "benefits"-i.e., they had more skin lesions diagnosed than non-app users, likely due to increased access.This supports using AI skin apps insofar as access is the limiting determinant of diagnosis. 25

CONCLUSION
The use of AI has high potential to facilitate the way skin cancer is diagnosed.Two main branches of AI are used to detect and classify skin cancer, namely shallow and deep techniques.However, the reliability of such AI tools is questionable since different data set sizes, image types, and number of diagnostic classes are being used and evaluated with different evaluation metrics.