DOWNLOAD THE COMPLETE PROJECT

7 Users found this project useful

Effect of AI-Assisted Skin Cancer Screening on Skin Cancer Detection among Adults in Nigeria

Abstract

Skin cancer is an important public health concern that requires timely recognition and appropriate clinical evaluation, particularly when suspicious skin lesions are present. Although skin cancer incidence and mortality patterns vary across populations, delayed recognition of suspicious lesions may contribute to late diagnosis and poorer treatment outcomes. In Nigeria, limited access to dermatologists, specialized diagnostic services, and skin-health screening programmes may create barriers to timely assessment, particularly in underserved communities. Artificial intelligence (AI)-assisted skin cancer screening is an emerging digital health approach that can support the assessment of skin lesions by analyzing clinical or dermoscopic images and identifying features that may warrant further professional evaluation. AI-assisted screening may improve access to preliminary risk assessment and facilitate referral of individuals with suspicious lesions. However, limitations related to diagnostic accuracy, image quality, algorithmic bias, skin-tone representation, privacy, and the need for professional clinical confirmation remain important considerations. Against this background, this study investigates the effect of AI-assisted skin cancer screening on skin cancer detection among adults in Nigeria. The study will be anchored on the Technology Acceptance Model, Health Belief Model, and Health Systems Framework. The Technology Acceptance Model explains how perceived usefulness, perceived ease of use, trust, and perceived reliability of AI-assisted screening may influence acceptance and utilization of the technology. The Health Belief Model explains how perceived susceptibility to skin cancer, perceived severity, perceived benefits of screening, perceived barriers, and cues to action may influence participation in skin cancer screening and follow-up. The Health Systems Framework emphasizes service delivery, healthcare workforce, health information, medical technologies, financing, accessibility, referral systems, and continuity of care as essential components of effective health screening and diagnostic services. Collectively, these theoretical perspectives provide a suitable framework for explaining how AI-assisted skin cancer screening may influence skin cancer detection among adults in Nigeria. The study will adopt a quantitative quasi-experimental or analytical cross-sectional research design. The study population will comprise adults aged 18 years and above residing in selected urban, semi-urban, and rural communities across Nigeria. A multistage sampling technique will be used to select geopolitical zones, states, local government areas, communities, screening sites, and eligible participants. AI-assisted skin cancer screening will be assessed using indicators such as availability of AI-assisted screening tools, number of participants screened, image acquisition procedures, image quality, AI-generated lesion risk classifications, screening time, referral recommendations, accessibility, and availability of trained personnel to operate or interpret the technology. Skin cancer detection will be assessed using indicators such as suspicious skin lesions identified, AI-positive screening results, clinically confirmed lesions, referrals for dermatological evaluation, histopathological confirmation where indicated, newly detected skin cancers, lesion characteristics, and appropriate follow-up after abnormal screening results. Data will be collected using structured questionnaires, skin-screening forms, AI-assisted screening records, clinical examination records, dermatological assessment reports, referral registers, and histopathology reports where available. Descriptive statistics will be used to summarize participants' characteristics, screening findings, AI classifications, referrals, and confirmed cases. Inferential statistical techniques, including chi-square tests, t-tests, correlation analysis, logistic or multiple regression analysis, and diagnostic performance measures such as sensitivity, specificity, positive predictive value, and negative predictive value where appropriate, will be used to determine the effect of AI-assisted screening on skin cancer detection. Where a quasi-experimental design is adopted, detection rates before and after implementation of AI-assisted screening may be compared with those of a comparison group using conventional screening approaches. Diagnostic tests will also be conducted to assess the reliability, validity, and robustness of the findings. The study is expected to find that AI-assisted skin cancer screening has a significant positive effect on skin cancer detection among adults in Nigeria. Adults who participate in AI-assisted screening are expected to have a greater likelihood of suspicious skin lesions being identified and appropriately referred for professional evaluation than adults without comparable access to screening. AI-assisted tools may support rapid preliminary assessment of skin lesions and help healthcare workers identify individuals who require further dermatological evaluation. The technology may be particularly useful in settings where access to dermatologists is limited, provided that it is used as a screening or decision-support tool rather than a replacement for professional diagnosis. AI-assisted screening may also reduce some geographical barriers by supporting preliminary assessment in primary healthcare and community settings. However, inadequate image quality, limited representation of diverse Nigerian skin tones in training datasets, algorithmic errors, false-positive or false-negative results, limited availability of suitable devices, inadequate digital infrastructure, privacy concerns, and weak referral systems may affect effectiveness. The study therefore expects appropriately validated, locally relevant, professionally supervised, and ethically implemented AI-assisted screening to contribute significantly to improved detection and referral of suspicious skin lesions among adults in Nigeria. The study is expected to contribute to the literature on artificial intelligence in healthcare, AI-assisted cancer screening, dermatological screening, skin cancer detection, digital health, health technology, early cancer detection, and public health in Nigeria. The findings will provide useful information to the Federal Ministry of Health and Social Welfare, National Primary Health Care Development Agency, hospitals, dermatology departments, primary healthcare centres, healthcare professionals, digital health developers, technology providers, development partners, and policymakers regarding the responsible use of AI-assisted screening technologies. The study will also provide evidence-based recommendations for validating AI screening tools across relevant Nigerian populations, improving representation of diverse skin tones in AI datasets, establishing appropriate clinical referral pathways, protecting patient privacy, training healthcare workers, monitoring screening accuracy, and integrating AI-assisted skin lesion assessment into appropriate healthcare services without replacing necessary professional clinical evaluation.

Keywords: AI-assisted screening, artificial intelligence, skin cancer detection, skin cancer screening, digital health, dermatology, early cancer detection, health technology, adults, Nigeria, public health.