Detection of Skin Cancer Based on Skin Lesion Images Using Deep Learning

Healthcare, Год журнала: 2022, Номер 10(7), С. 1183 - 1183

Опубликована: Июнь 24, 2022

An increasing number of genetic and metabolic anomalies have been determined to lead cancer, generally fatal. Cancerous cells may spread any body part, where they can be life-threatening. Skin cancer is one the most common types its frequency worldwide. The main subtypes skin are squamous basal cell carcinomas, melanoma, which clinically aggressive responsible for deaths. Therefore, screening necessary. One best methods accurately swiftly identify using deep learning (DL). In this research, method convolution neural network (CNN) was used detect two primary tumors, malignant benign, ISIC2018 dataset. This dataset comprises 3533 lesions, including malignant, nonmelanocytic, melanocytic tumors. Using ESRGAN, photos were first retouched improved. augmented, normalized, resized during preprocessing step. lesion could classified a CNN based on an aggregate results obtained after many repetitions. Then, multiple transfer models, such as Resnet50, InceptionV3, Inception Resnet, fine-tuning. addition experimenting with several models (the designed CNN, Resnet), study's innovation contribution use ESRGAN Our model showed comparable pretrained model. Simulations ISIC 2018 that suggested strategy successful. 83.2% accuracy rate achieved by in comparison Resnet50 (83.7%), InceptionV3 (85.8%), Resnet (84%) models.

Язык: Английский

---

An innovation prediction of DNA damage of melanoma skin cancer patients using deep learning

2022 13th International Conference on Computing Communication and Networking Technologies (ICCCNT), Год журнала: 2023, Номер unknown, С. 1 - 7

Опубликована: Июль 6, 2023

There are now a variety of intriguing options for the study genetic data thanks to recent developments in artificial neural networks and deep learning. In this study, we use learning-based prediction model find possible DNA damage individuals with melanoma skin cancer. We create convolutional network (CNN) forecast susceptibility cancer cells using publically available genome sequencing dataset. This preprocesses genomic data, extracts features, categorises them. Comparing results our CNN those traditional logistic regression model, that reported superior performance identifying differences between healthy cancerous samples an accuracy nearly 96%. The can be used augment standard clinical diagnosis melanoma, which only uses visual assessment histology. By intervening sooner, clinicians put forward more personalized informed plans care surveillance each patient, reducing medical costs improving quality patient diagnosis.

Язык: Английский

---

Artificial intelligence-driven enhanced skin cancer diagnosis: leveraging convolutional neural networks with discrete wavelet transformation

Egyptian Journal of Medical Human Genetics, Год журнала: 2024, Номер 25(1)

Опубликована: Апрель 9, 2024

Abstract Background Artificial intelligence (AI) has shown great promise in the field of healthcare as a means improving diagnosis skin cancer. The objective this research is to enhance precision and effectiveness cancer identification by incorporation convolutional neural networks (CNNs) discrete wavelet transformation (DWT). Making use AI-driven techniques potential completely transform process providing quicker more accurate evaluations lesions. In an effort improve dermatology give physicians reliable resources for early precise diagnosis, work explores combination CNNs with DWT. Methods timely classification lesions plays crucial role effective treatment. this, we propose novel approach using DWT employed extract relevant features from lesion images, which are then used train model. suggested assessed through examination dataset images known classes (malignant or benign). Results outcomes experiment demonstrate that model successfully attained result sensitivity 94% specificity 91% when compared artificial network (ANN) multilayer perceptron methods. Conclusions HAM 10000 explore evaluate proposed model, leading improved accuracy existing machine learning algorithms utilization. results DWT-based accurately classifying lesions, thus aiding detection diagnosis.

Язык: Английский

---

Hybrid Feature Fusion and Machine Learning Approaches for Melanoma Skin Cancer Detection

Опубликована: Янв. 18, 2022

Skin cancer is an exquisite disease globally nowadays. Because of the poor contrast and apparent resemblance between skin lesions, automatic identification complicated. The rate human death can be massively reduced if melanoma detected quickly using dermoscopy images. In this research, anisotropic diffusion filtering method used on images to remove multiplicative speckle noise fast-bounding box (FBB) applied segment region. Furthermore, paper consists two feature extractor parts. One features parts hybrid (HFE) part another convolutional neural network VGG19 based CNN part. HFE portion combines three extraction approaches into a single fused vector: Histogram-Oriented Gradient (HOG), Local Binary Pattern (LBP), Speed Up Robust Feature (SURF). also extract additional from test training datasets. This two-feature vector design classification model. classifier performs whether it or non-melanoma cancer. proposed methodology performed ordinary datasets achieved accuracy 99.85%, sensitivity 91.65%, specificity 95.70%, which makes more successful than previous machine learning algorithms.

Язык: Английский

---

BreastMultiNet: A multi-scale feature fusion method using deep neural network to detect breast cancer

Array, Год журнала: 2022, Номер 16, С. 100256 - 100256

Опубликована: Ноя. 7, 2022

Breast cancer is predominantly seen in women and the leading cause of death females worldwide. Diagnosis breast using biopsy tissue images expensive, time-intensive, fraught with conflicts among doctors. Pathologists can now diagnose more consistently promptly because advances Computer-Aided (CAD) system. As a result, there has been surge demand for CAD-based machine learning techniques. This study describes “BreastMultiNet” framework that focuses on transfer concept identifying distinct types by utilizing two publicly available datasets. The suggested architecture allows rapid comprehensive diagnosis. scheme extracts features from microscope help well-known conventional deep models such as HOG, LBP, SURP, DenseNet201, VGG19. Comparatively, provide good accuracy than models. collected properties are subsequently dispatched into summing layer, resulting fused vector. proposed achieves 99% 95% classification both BreakHis ICIAR dataset respectively, outperforming all other state art In terms accuracy, may be employed modeling approach hospitals medical care contexts.

Язык: Английский

---

Biorthogonal wavelet based entropy feature extraction for identification of maize leaf diseases

Journal of Agriculture and Food Research, Год журнала: 2023, Номер 14, С. 100756 - 100756

Опубликована: Авг. 31, 2023

Crop disease is considered as a major constraint to both food quality and production. Even in this era of precision agriculture, the lacking compulsory infrastructure has made rapid identification crop diseases quite hard numerous regions around world. In paper, we introduced new method based on biorthogonal wavelet transform (BWT) identify prime maize leaf diseases. We performed decomposition pixel wise morphological operation segment lesion from input image. For feature extraction, by applying 2-D at multiple levels proposed novel extract colour channel entropy features investigating discriminatory potential three different filters (bior3.3, bior3.5, bior3.7). The effectiveness our extracted were evaluated employing five classifiers obtaining 95.78% overall accuracy with 10-fold cross validation. All materials related study can be found at: https://github.com/BadhanMazumder/BiorthogonalWavelet_MaizeDiseaseDetection.git.

Язык: Английский

---

An efficient multiscale enhancement network with attention mechanism for aluminium defect detection

Nondestructive Testing And Evaluation, Год журнала: 2024, Номер unknown, С. 1 - 18

Опубликована: Окт. 14, 2024

Real-time defect detection is required to efficiently control the quality of aluminium. However, aluminium defects have characteristics small-size, low contrast, and multiscale variations, which pose great challenges detection. This article aims improve accuracy propose an effective enhancement network with attention mechanism for (AMMENet). First, capture key features mitigate interference from background, a pluggable parallel residual module (PRAM) proposed feature extraction network. To compensate loss deep features, multilevel semantic (C2f-MFF) fuse maps. Finally, model was applied Tianchi Aluminium Surface Defect Dataset (TC-ASDD) ablation experiments comparisons. The experimental results show that mean average precision ([email protected]) AMMENet 73.6% real-time speed 66.2 frames per second (FPS). Compared YOLOv8 baseline network, improves [email protected] by 2.8% only slight in speed. Moreover, superior state-of-the-art methods terms accuracy.

Язык: Английский

---

A novel artificial intelligence-based predictive analytics technique to detect skin cancer

PeerJ Computer Science, Год журнала: 2023, Номер 9, С. e1387 - e1387

Опубликована: Май 24, 2023

One of the leading causes death among people around world is skin cancer. It critical to identify and classify cancer early assist patients in taking right course action. Additionally, melanoma, one main illnesses, curable when detected treated at an stage. More than 75% fatalities worldwide are related A novel Artificial Golden Eagle-based Random Forest (AGEbRF) created this study predict cells Dermoscopic images used instance as dataset for system's training. dermoscopic image information processed using established AGEbRF function segment cancer-affected area. approach simulated a Python program, current research's parameters assessed against those earlier studies. The results demonstrate that, compared other models, new research model produces better accuracy predicting by segmentation.

Язык: Английский

---

Enhancing histopathological medical image classification for Early cancer diagnosis using deep learning and explainable AI – LIME & SHAP

Biomedical Signal Processing and Control, Год журнала: 2024, Номер 100, С. 107014 - 107014

Опубликована: Окт. 12, 2024

Язык: Английский

---

Multi-stage feature extraction-based classification of skin cancer detection

Soft Computing, Год журнала: 2023, Номер 28(S2), С. 633 - 633

Опубликована: Июнь 29, 2023

Язык: Английский

---