Method for realizing automated detection of melanoma skin cancer, involves calculating total dermoscopy score based on extracted features, and classifying images using convolutional neural network for melanoma detection
	DUTTA B; DESHMUKH T; SINGRU M C; GAIDHANI A R; MOGARE S B
	2024-09-03
专利权人	UNIV SANDIP SIJOUL (UYSA-Non-standard) ; SANDIP ENG & MANAGEMENT INST (SAND-Non-standard) ; UNIV SANDIP NASHIK (UYSA-Non-standard) ; SANDIP TECHNOLOGY & RES CENT INST (SAND-Non-standard)
申请日期	2024-09-03
专利号	IN202421066488-A
成果简介	NOVELTY - The method involves collecting a dermoscopy image database. Images are preprocessed to enhance quality and remove noise. The images are segmented using thresholding techniques. Statistical features such as gray level co-occurrence matrix (GLCM), asymmetry, border, volor, diameter (ABCD) are extracted. Relevant features are selected using principal component analysis (PCA). A total dermoscopy score is calculated based on the extracted features. The images are classified using a convolutional neural network (CNN) for melanoma detection. USE - Method for realizing automated detection of melanoma skin cancer. ADVANTAGE - The method enables realizing user interface interacts with both the image database and the CNN classifier, facilitating input of images and display of classification results to ensure efficient data flow and processing for automated melanoma detection. The method enables integrating computer vision, image processing, and CNNs to enhance the accuracy and efficiency of melanoma diagnosis, reducing time-consuming and prone to errors in discerning early-stage lesions. The method enables analyzing dermoscopy images with high precision, enabling rapid and reliable identification of suspicious lesions indicative of melanoma to extract relevant information from the images, employing advanced classification techniques to distinguish between benign and malignant lesions, facilitating prompt medical intervention when necessary. The method enables utilizing a user-friendly tool for healthcare professionals to expedite diagnosis of melanoma and improve patient outcomes through timely treatment and intervention. The method enables realizing early detection of melanoma skin cancer using machine learning techniques to enhance patient outcomes by facilitating early diagnosis, reducing risk of metastasis, and increasing survival rates for individuals diagnosed with melanom. DETAILED DESCRIPTION - INDEPENDENT CLAIMS are included for: (1) a system for realizing automated detection of melanoma skin cancer; (2) a omputer-readable storage medium for storing a set of instructions for realizing automated detection of melanoma skin cancer. DESCRIPTION OF DRAWING(S) - The drawing shows a block diagram of a system for realizing automated detection of melanoma skin cancer. 101Preprocessing module 102Feature selection module 103User interface 105CNN classifier 106Segmentation algorithm
IPC 分类号	A24B-003/18 ; A61B-017/86 ; A61B-017/88 ; A61B-090/00 ; A61F-002/46
国家	印度
专业领域	农业科学
语种	英语
成果类型	专利
文献类型	科技成果
条目标识符	http://119.78.100.226:8889/handle/3KE4DYBR/15271
专题	中国科学院新疆生态与地理研究所
作者单位	1.UNIV SANDIP SIJOUL (UYSA-Non-standard) 2.SANDIP ENG & MANAGEMENT INST (SAND-Non-standard) 3.UNIV SANDIP NASHIK (UYSA-Non-standard) 4.SANDIP TECHNOLOGY & RES CENT INST (SAND-Non-standard)
推荐引用方式 GB/T 7714	DUTTA B,DESHMUKH T,SINGRU M C,et al. Method for realizing automated detection of melanoma skin cancer, involves calculating total dermoscopy score based on extracted features, and classifying images using convolutional neural network for melanoma detection. IN202421066488-A[P]. 2024.

条目包含的文件
条目无相关文件。

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。