Method for performing cervical area segmentation and classification using logistic regression compared over linear regression, involves training algorithm on labeled dataset containing annotated regions of interest to classify pixels as cervical or non-cervical and achieve accurate segmentation
	HABIBA U H; SAHIB K A S K A; HEMALATHA; SRINIVASULU K; PRASAD K N; KUMAR J M R V
	2023-10-16
专利权人	BHARATH HIGHER EDUCATION & RES INST (BHAR-Non-standard)
申请日期	2023-10-16
专利号	IN202341069957-A
成果简介	NOVELTY - The method involves performing logistic regression using a well-established machine learning (ML) algorithm for binary classification tasks, is employed to segment the cervical area from digital cervicography images. An algorithm is trained on a labeled dataset containing annotated regions of interest to classify pixels as cervical or non-cervical and achieve accurate segmentation. A linear regression is adapted for binary classification using a threshold value. The performance of linear regression is compared with logistic regression in terms of accuracy. USE - Method for performing cervical area segmentation and classification using logistic regression compared over linear regression. Can also be used in medical imaging to detect cervical cancer. ADVANTAGE - The method performs automated segmentation and classification on cervical areas, avoids manual intervention, increases efficiency, improves accuracy and patient care, leverages logistic regression, identifies abnormal cervical regions, achieves early detection of abnormalities, enables early detection of abnormal cervical regions, facilitates timely treatment and improving patient outcomes, reduces inter-observer variability, manual segmentation and workload of medical professionals, ensures consistent and reliable results, is seamlessly integrated into clinical workflows and imaging systems and evaluated using various accuracy metrics, e.g. precision, recall and F 1-score, on a large dataset of annotated cervical images and allows widespread adoption and usability in diverse healthcare environments. DESCRIPTION OF DRAWING(S) - The drawing shows a flow chart illustrating the method for performing cervical area segmentation and classification.
IPC 分类号	A61B-005/00 ; A61B-006/00 ; G06N-020/00 ; G06T-007/00 ; G16H-050/20
国家	印度
专业领域	信息技术
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/19869
专题	中国科学院新疆生态与地理研究所
作者单位	BHARATH HIGHER EDUCATION & RES INST (BHAR-Non-standard)
推荐引用方式 GB/T 7714	HABIBA U H,SAHIB K A S K A,HEMALATHA,et al. Method for performing cervical area segmentation and classification using logistic regression compared over linear regression, involves training algorithm on labeled dataset containing annotated regions of interest to classify pixels as cervical or non-cervical and achieve accurate segmentation. IN202341069957-A[P]. 2023.

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。