Explainable AI-Based Skin Cancer Detection Using CNN, Particle Swarm Optimization and Machine Learning

Journal of Imaging, Год журнала: 2024, Номер 10(12), С. 332 - 332

Опубликована: Дек. 22, 2024

Skin cancer is among the most prevalent cancers globally, emphasizing need for early detection and accurate diagnosis to improve outcomes. Traditional diagnostic methods, based on visual examination, are subjective, time-intensive, require specialized expertise. Current artificial intelligence (AI) approaches skin face challenges such as computational inefficiency, lack of interpretability, reliance standalone CNN architectures. To address these limitations, this study proposes a comprehensive pipeline combining transfer learning, feature selection, machine-learning algorithms accuracy. Multiple pretrained models were evaluated, with Xception emerging optimal choice its balance efficiency performance. An ablation further validated effectiveness freezing task-specific layers within architecture. Feature dimensionality was optimized using Particle Swarm Optimization, reducing dimensions from 1024 508, significantly enhancing efficiency. Machine-learning classifiers, including Subspace KNN Medium Gaussian SVM, improved classification Evaluated ISIC 2018 HAM10000 datasets, proposed achieved impressive accuracies 98.5% 86.1%, respectively. Moreover, Explainable-AI (XAI) techniques, Grad-CAM, LIME, Occlusion Sensitivity, enhanced interpretability. This approach provides robust, efficient, interpretable solution automated in clinical applications.

Язык: Английский

---

An ensemble-based deep learning model for detection of mutation causing cutaneous melanoma

Scientific Reports, Год журнала: 2023, Номер 13(1)

Опубликована: Дек. 14, 2023

Abstract When the mutation affects melanocytes of body, a condition called melanoma results which is one deadliest skin cancers. Early detection cutaneous vital for raising chances survival. Melanoma can be due to inherited defective genes or environmental factors such as excessive sun exposure. The accuracy state-of-the-art computer-aided diagnosis systems unsatisfactory. Moreover, major drawback medical imaging shortage labeled data. Generalized classifiers are required diagnose avoid overfitting dataset. To address these issues, blending ensemble-based deep learning (BEDLM-CMS) model proposed detect by integrating long short-term memory (LSTM), Bi-directional LSTM (BLSTM) and gated recurrent unit (GRU) architectures. dataset used in study contains 2608 human samples 6778 mutations total along with 75 types genes. most prominent that function biomarkers early prognosis utilized. Multiple extraction techniques this extract most-prominent features. Afterwards, we applied different DL models optimized through grid search technique melanoma. validity confirmed using several techniques, including tenfold cross validation (10-FCVT), independent set (IST), self-consistency (SCT). For multiple metrics include accuracy, specificity, sensitivity, Matthews’s correlation coefficient. BEDLM gives highest 97% test whereas it 94% 93% respectively. Accuracy test, (96%, 94%, 92%), GRU (93%, 91%), BLSTM (99%, 98%, 93%), findings demonstrate BEDLM-CMS effectively treatment efficacy evaluation

Язык: Английский

---

RvXmBlendNet: A Multi-architecture Hybrid Model for Improved Skin Cancer Detection

Human-Centric Intelligent Systems, Год журнала: 2024, Номер unknown

Опубликована: Сен. 9, 2024

Abstract Skin cancer, one of the most dangerous cancers, poses a significant global threat. While early detection can substantially improve survival rates, traditional dermatologists often face challenges in accurate diagnosis, leading to delays treatment and avoidable fatalities. Deep learning models like CNN transfer have enhanced diagnosis from dermoscopic images, providing precise timely detection. However, despite progress made with hybrid models, many existing approaches still challenges, such as limited generalization across diverse datasets, vulnerability overfitting, difficulty capturing complex patterns. As result, there is growing need for more robust effective that integrate multiple architectures advanced mechanisms address these challenges. Therefore, this study aims introduce novel multi-architecture deep model called "RvXmBlendNet," which combines strengths four individual models: ResNet50 (R), VGG19 (v), Xception (X), MobileNet (m), followed by "BlendNet" signify their fusion into unified architecture. The integration achieved through synergistic combination architectures, incorporating self-attention using attention layers adaptive content blocks. This used HAM10000 dataset refine image preprocessing enhance accuracy. Techniques OpenCV-based hair removal, min–max scaling, histogram equalization were employed quality feature extraction. A comparative between proposed "RvXmBlendNet" (CNN, ResNet50, VGG19, Xception, MobileNet) demonstrated highest accuracy 98.26%, surpassing other models. These results suggest system facilitate earlier interventions, patient outcomes, potentially lower healthcare costs reducing invasive diagnostic procedures.

Язык: Английский

---

AI in dermatology: a comprehensive review into skin cancer detection

PeerJ Computer Science, Год журнала: 2024, Номер 10, С. e2530 - e2530

Опубликована: Дек. 5, 2024

Artificial Intelligence (AI) is significantly transforming dermatology, particularly in early skin cancer detection and diagnosis. This technological advancement addresses a crucial public health issue by enhancing diagnostic accuracy, efficiency, accessibility. AI integration medical imaging procedures offers promising solutions to the limitations of traditional methods, which often rely on subjective clinical evaluations histopathological analyses. study systematically reviews current applications classification, providing comprehensive overview their advantages, challenges, methodologies, functionalities. In this study, we conducted analysis artificial intelligence classification cancer. We evaluated publications from three prominent journal databases: Scopus, IEEE, MDPI. thorough selection process using PRISMA guidelines, collecting 1,156 scientific articles. Our methodology included evaluating titles abstracts thoroughly examining full text determine relevance quality. Consequently, total 95 final study. analyzed categorized articles based four key dimensions: difficulties, AI-based models exhibit remarkable performance leveraging advanced deep learning algorithms, image processing techniques, feature extraction methods. The advantages include improved faster turnaround times, increased accessibility dermatological expertise, benefiting underserved areas. However, several challenges remain, such as concerns over data privacy, complexities integrating systems into existing workflows, need for large, high-quality datasets. methods detection, including CNNs, SVMs, ensemble aim improve lesion accuracy increase detection. enhance healthcare enabling remote consultations, continuous patient monitoring, supporting decision-making, leading more efficient care better outcomes. review highlights transformative potential While technologies have accessibility, remain. Future research should focus ensuring developing robust that can generalize across diverse populations, creating Integrating tools workflows critical maximizing utility effectiveness. Continuous innovation interdisciplinary collaboration will be essential fully realizing benefits

Язык: Английский

---

Custom CNN architectures for skin disease classification: binary and multi-class performance

Multimedia Tools and Applications, Год журнала: 2024, Номер unknown

Опубликована: Дек. 16, 2024

Язык: Английский

---

Can Vision Transformers Be the Next State-of-the-Art Model for Oncology Medical Image Analysis?

AI in Precision Oncology, Год журнала: 2024, Номер 1(6), С. 286 - 305

Опубликована: Дек. 1, 2024

Язык: Английский

---

Development of Convolutional Neural Network-Based AI-Dermatoscope for Non-Invasive Skin Assessments

Опубликована: Дек. 7, 2023

Early detection of skin conditions is crucial, and some can become more difficult to treat if left untreated. The gold standard Dermatoscope a non-invasive technique used for the examination evaluation lesions, which equipped with magnifying lens light source. However, precise inspection existing dermatoscopes has limitation due unavailability image-analyzing methods. Herein, this study reports successful development Convolutional Neural Networks (CNN) based, Artificial intelligence (AI)-Dermatoscope integrating optics smart illumination system enhance accurate acne skin. was trained on large dataset accurately identify classify conditions. Finally, utilizes CNN knowledge predict new images provide diagnostic information doctors other healthcare professionals. Thus, will improve accuracy speed diagnosis, consequently, health-related quality life patients.

Язык: Английский

---