<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE ArticleSet PUBLIC "-//NLM//DTD PubMed 2.7//EN" "https://dtd.nlm.nih.gov/ncbi/pubmed/in/PubMed.dtd">
<ArticleSet>
<Article>
<Journal>
				<PublisherName>University of Sistan and Baluchestan</PublisherName>
				<JournalTitle>Iranian Journal of Fuzzy Systems</JournalTitle>
				<Issn>1735-0654</Issn>
				<Volume>23</Volume>
				<Issue>6</Issue>
				<PubDate PubStatus="epublish">
					<Year>2026</Year>
					<Month>12</Month>
					<Day>01</Day>
				</PubDate>
			</Journal>
<ArticleTitle>FLinkNet-GAN: An integrated deep learning and fuzzy logic architecture for membership based cerebral palsy risk assessment</ArticleTitle>
<VernacularTitle></VernacularTitle>
			<FirstPage>139</FirstPage>
			<LastPage>156</LastPage>
			<ELocationID EIdType="pii">10170</ELocationID>
			
<ELocationID EIdType="doi">10.22111/ijfs.2026.54858.9720</ELocationID>
			
			<Language>EN</Language>
<AuthorList>
<Author>
					<FirstName>M.</FirstName>
					<LastName>Sweetline Sonia</LastName>
<Affiliation>Mahendra Engineering College</Affiliation>

</Author>
<Author>
					<FirstName>S.</FirstName>
					<LastName>Sumathi</LastName>
<Affiliation>Mahendra Engineering College</Affiliation>

</Author>
</AuthorList>
				<PublicationType>Journal Article</PublicationType>
			<History>
				<PubDate PubStatus="received">
					<Year>2026</Year>
					<Month>02</Month>
					<Day>20</Day>
				</PubDate>
			</History>
		<Abstract>Early detection of cerebral palsy (CP) is challenging due to subtle and heterogeneous neuromotor abnormalities in  infants. This paper proposes FLinkNet-GAN, a hybrid deep learning and fuzzy logic framework for membership-based CP  risk assessment from RGB-D infant videos. Initially, a Bright Contrast Dynamic Histogram Equalization (BCDHE) filter  enhances motion visibility. A Generative Adversarial Network (GAN) performs multi-pose estimation to extract kinematic  features, including Movement Deviation (MD) and Joint Trajectory Discontinuity (JD). These features are processed  through a LinkNet backbone to preserve fine spatial details and subsequently mapped into a Mamdani-type Fuzzy  Inference System using Gaussian membership functions. Unlike conventional hard classifiers, the proposed approach  generates a continuous Risk Score (RS) via centroid defuzzification, providing interpretable risk levels: Normal, Borderline, High, and Very High. Experimental evaluation on the RVI-38 dataset achieves 94.85% accuracy and 94.5% pose precision,  demonstrating robust and uncertainty-aware CP risk stratification. </Abstract>
		<ObjectList>
			<Object Type="keyword">
			<Param Name="value">Cerebral palsy detection</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">Fuzzy Inference System</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">generative adversarial network</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">pose estimation</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">Gaussian Membership function</Param>
			</Object>
			<Object Type="keyword">
			<Param Name="value">risk score modeling</Param>
			</Object>
		</ObjectList>
<ArchiveCopySource DocType="pdf">https://ijfs.usb.ac.ir/article_10170_4843b09fb9b826fc48f92cdcf974a33f.pdf</ArchiveCopySource>
</Article>
</ArticleSet>
