Short reports or Letters · Open access
Cnn-Based Method For Benign And Malignant Skin Lesion Classification
Abstract
Skin cancer is one of the most common forms of cancer, with its diagnosis typically involving manual screening and visual examination by dermatologists. However, these traditional methods are often complex, time-consuming, and prone to errors. With the growing incidence and significant mortality rates associated with skin cancer, early detection is crucial for effective treatment. In this study, we present an optimized deep convolutional neural network (DCNN) model specifically designed to classify skin lesions as either benign or malignant, thereby improving diagnostic accuracy. Our approach utilizes a dataset of 3,297 dermoscopic images and incorporates rigorous data preprocessing along with the softmax activation function. We evaluated multiple optimizers—Adam, RMSProp, and SGD—each configured with a learning rate of 0.0001. Our findings indicate that the Adam optimizer outperforms others, achieving an accuracy of 84%, a loss rate of 32%, a recall of 85%, a precision of 85%, an F1-score of 85%, and a ROC-AUC of 83% on the ISIC dataset. This CNN-based approach offers a promising tool for enhancing early skin cancer detection and diagnosis.
Keywords
Cite this paper
Choose a citation style, then copy the formatted reference for your manuscript or academic record.
Naim, M. M. H. (2024). Cnn-Based Method For Benign And Malignant Skin Lesion Classification. International Journal of Advanced Scientific Discoveries, 4(1), online first. https://doi.org/10.66917/ijasd.a000019
Naim, Md. Mehedi Hasan. “Cnn-Based Method For Benign And Malignant Skin Lesion Classification.” International Journal of Advanced Scientific Discoveries, vol. 4, no. 1, 2024, pp. online first. doi:10.66917/ijasd.a000019
Naim, Md. Mehedi Hasan. “Cnn-Based Method For Benign And Malignant Skin Lesion Classification.” International Journal of Advanced Scientific Discoveries 4(1) (2024): online first. https://doi.org/10.66917/ijasd.a000019
Naim, M. M. H. (2024) ‘Cnn-Based Method For Benign And Malignant Skin Lesion Classification’, International Journal of Advanced Scientific Discoveries, 4(1), pp. online first. doi:10.66917/ijasd.a000019
BibTeX
@article{msisijasd-a000019,
title = {Cnn-Based Method For Benign And Malignant Skin Lesion Classification},
author = {Md. Mehedi Hasan Naim},
journal = {International Journal of Advanced Scientific Discoveries},
year = {2024},
volume = {4},
number = {1},
doi = {10.66917/ijasd.a000019}
}Related articles
Browse this journal
IJASD
One Year Descriptive Analysis Of Reported Animal Bite Cases At Techiman Municipality Of Ghana
Background: Rabies is a fatal but preventable zoonotic viral disease of the central nervous system. In developed countries, rabies is mostly found in wild animals like bats, but the developing countries like Ghana, dogs still carry rabies, and most rabies deaths in people around the world are caused by dog bites. There are limited studies on rabies in the Techiman Municipality, Bono East Region of Ghana. The current study focuses on the incidence of bites and resultant rabies infectivity, management and control in Techiman South Municipality. The objective of this study is to detect the gaps in rabies control and prevention practices while providing information for mass education. Methods: The study design was cross-sectional with Excel Format for the sample collection from the two departments under Veterinary Service Directorate in Techiman that is Disease Investigation Farm (DIF)/Regional Veterinary Laboratory (RVL) and the Municipal Veterinary Clinic (MVC) all in Techiman. In all, 163 reported cases were collected. Data on dog bites and rabies cases in Bono East Region from January to December 2022 were also analysed. Results: One hundred and Sixty-three (163) cases of animal bites were reported from January to December 2022. The majority of the animals involved were dogs which were 158 representing 96.93% and 73.61% of dogs were not vaccinated. Complete data were available for 157 completed cases. The age range of reported cases was 11months-83 years with few unknown ages. Nearly twenty-five percent (24.54%) were children under 10 years, the majority being males representing 47.85% but slightly above females representing 47.23%. All cases of animal bites were reported first to health facilities before they referred to veterinary facilities. There was a rapid increase of reported cases in the last quarter of 2022 83/163. Recommendation: Ghana Health Service and VSD operated independent systems of zoonotic disease reporting, collected very different sets of information and did not have a common platform for data sharing and reconciliation. This needs to be addressed to ensure better disease information collection, management and dissemination. Animals suspected of having rabies on dying have their brain tissue examined at KCCR as DIF/RVL not having the standard equipment for rabies diagnosis. This calls for upgrading the DIF/RVL and also further improvement in the skills and/or equipment for rabies detection at DIF/RVL. This was the reason why most of the samples were not able to be summited to KCCR due to the high cost of transportation to Kumasi.
IJASD
Ai Explainability In Insurance Applications With Telematics Data
This paper describes a step-by-step process of creating an insurance-based application that predicts insurance premium using Artificial Intelligence (AI) method of machine learning (ML). Artificial Intelligence technologies will have a high impact on all aspects of the insurance industry, from distribution to underwriting and pricing to claims. Nowadays, AI is being used in insurance claims processing, underwriting, fraud detection, anomaly detection and customer service. In this research paper, we are going to predict a driver's premium charges based on the driver's driving behavior. Here, we used telematics, machine learning, and deep learning technologies for predicting the insurance premium cost in the auto insurance space. In this research journey, we propose a hybrid framework architecture with multiple combinations of Supervised Learning (LR, DS, RFR and SVM) Algorithms and interpretability of AI model.
IJASD
Irradiation Deactivation Of Anti-Nutritional Factors In Moringa Oleifera Seeds For Meat Chicken
Phytochemical composition and deactivation of anti-nutritional factors in Moringa oleifera seeds (MOS) to replace soybeans in diet for chicks were investigated. 5-experimental diets were formulated and fed to 180 -Arbor Acre broiler chicks at hatch over a feeding trial lasting 56-days. Each dietary treatment contained 3-replicates of 36-chicks. The diets were made of a control diet1 (free of MOS), a negative control diet 2 containing 5% MOS and 3 other diets 3, 4 and 5 containing 5% gamma-irradiated Moringa oleifera Seed Meal (GIMOSM) at graded doses of 10, 30 and 50kGy respectively. The experiment followed a one -way classification design. Chemical analysis of the raw seed gave valuable amounts of nutrients as well as anti-nutrients which presence could elicit health hazards when consumed in diet in certain amount. Results on performance traits, nutrient digestibility (crude protein, fat and fibre) gave poor results in the negative/raw MOS control diet (p < 0.05) compared to the control and GIMOSM diets. GIMOSM diets improved (p < 0.05) haematological (PCV, RBC, HB) and biochemical (urea, AST, ALT, ALP) parameters measured in broiler nutrition. Findings of this study showed that the detrimental effect induce by the anti-nutrients present in the raw MOS on the measured biological parameters were reversed to good by g-irradiation treatment. It is suggested that MOS meal be treated at 10kGy level to safely detoxify its anti-nutritional and/or toxic factors for optimum use in poultry nutrition.
IJASD
Drug Utilization Evaluation Among Hypertensive Patients In A Tertiary Care Setting Of Southern Rajasthan
Hypertension, a prevalent cardiovascular condition, is a significant public health concern globally. This burden is particularly pronounced in India, affecting a substantial proportion of the population. Effective management of hypertension relies heavily on the appropriate utilization of antihypertensive medications. This prospective observational cohort study, conducted in a tertiary hospital in Southern Rajasthan, recruited 220 hypertensive patients on antihypertensive drug therapy and followed them at intervals of 0, 24 hours, 7 days, and 30 days. The study found that angiotensin receptor blockers (ARBs) were the most frequently used class of drugs in monotherapy, accounting for 15%. In dual therapies, ARB combined with calcium channel blockers (CCBs) was used by 20% of patients, while 11% of patients were on triple therapy involving ARB, CCB, and diuretics.