An Improved YOLO-SCS Model for AQP4 Antibody Immunofluorescence Image Detection
DOI:
https://doi.org/10.66535/myw05225Keywords:
AQP4 antibody; immunofluorescence image; target detection; YOLOv11; SPD-Conv; Shape-NWDAbstract
Neuromyelitis optica spectrum disorder (NMOSD) diagnosis currently relies on the cell-based assay (CBA) as the gold standard for detecting aquaporin-4 (AQP4) antibodies. However, manual interpretation of AQP4 antibody immunofluorescence images is subjective, time-consuming, and prone to variability. These images often exhibit challenges such as small, densely distributed positive signals, weak fluorescence intensity, and severe background interference from non-specific binding or artifacts.To address these issues, we propose an improved YOLO-based model, termed YOLO-SCS, built upon YOLOv11, for accurate detection of positive signals in AQP4 immunofluorescence images. First, a high-quality dataset of AQP4 immunofluorescence images was constructed with precise annotations. A Gaussian Peak-optimized Channel Adjustment (GPCA) algorithm was designed for preprocessing to enhance weak fluorescence contrast while effectively suppressing background noise. On the YOLOv11 architecture, we incorporated the SPD-Conv module in the downsampling stages to preserve fine-grained spatial features critical for small targets. Additionally, a shape-aware Normalized Wasserstein Distance (NWD) loss function was adopted to improve bounding box regression accuracy for densely overlapping and irregularly shaped positive regions. Experimental evaluations on a clinically representative dataset (positive-to-negative ratio of 1:9) demonstrated superior performance: the proposed YOLO-SCS achieved a positive detection rate of 98.4%, negative specificity of 100%, mAP@0.5 of 0.6971, and mAP@0.5:0.95 of 0.4256. These metrics significantly outperform the baseline YOLOv11 and other state-of-the-art improved models. The model reliably distinguishes specific positive fluorescence signals from background impurities, providing robust automated interpretation support for AQP4 immunofluorescence in NMOSD clinical diagnosis and potentially reducing diagnostic turnaround time and inter-observer variability.
References
1.ABDULRAHMAN Z, HENDRIKS N, KRUSE A J, et al. Immune-based biomarker accurately predicts response to imiquimod immunotherapy in cervical high-grade squamous intraepithelial lesions [J]. JOURNAL FOR IMMUNOTHERAPY OF CANCER, 2022, 10(11).
2.ALESSI J V, LINDSAY J R, GIOBBIE-HURDER A, et al. Immunoprofiling at an Institutional Scale Reveals That High Numbers of Intratumoral CD8+ and PD-1+ Cells Predict Superior Patient Survival Across Major Cancer Types Independent of Major Risk Factors [J]. JCO PRECISION ONCOLOGY, 2025, 9.
3.ANDRIC F, AL-FAIROUZI A, WETTERGREN Y, et al. Immune Microenvironment in Sporadic Early-Onset versus Average-Onset Colorectal Cancer [J]. CANCERS, 2023, 15(5).
4.AUNG T N, SHA S B, WILMOTT J S, et al. Objective assessment of tumor infiltrating lymphocytes as a prognostic marker in melanoma using machine learning algorithms [J]. EBIOMEDICINE, 2022, 82.
5.AZIMI M. From Spatial Patterns to Prognosis: Decoding Single-Cell Architecture in Cancer with Hyperplex Immunofluorescence Imaging [J]. JOURNAL OF CANCER, 2025, 16(12): 3654-3663.
6.BAI R L, JIA Y H, WANG B, et al. In vivo spatiotemporal mapping of proliferation activity in gliomas via water-exchange dynamic contrast-enhanced MRI [J]. THERANOSTICS, 2025, 15(10): 4693-707.
7.WANG Q, HUANG Y X, ZHANG Y, et al. Digital pathology-based prognostic model for hepatocellular carcinoma: Integrating pathomics signatures with clinical parameters for recurrence prediction and biological interpretation [J]. COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE, 2026, 275.
8.WATSON S S, DUC B, KANG Z Q, et al. Microenvironmental reorganization in brain tumors following radiotherapy and recurrence revealed by hyperplexed immunofluorescence imaging [J]. NATURE COMMUNICATIONS, 2024, 15(1).
9.WESSELINK E, THOMAS C E, TAKASHIMA Y, et al. Associations between Calcium Intake and T-cell Infiltration in Colorectal Tumors [J]. CANCER PREVENTION RESEARCH, 2025, 18(9): 561-571.
10.XIE T, XUE H C, HUANG A L, et al. Integrating multi-omics data with artificial intelligence to decipher the role of tumor-infiltrating lymphocytes in tumor immunotherapy [J]. PATHOLOGY RESEARCH AND PRACTICE, 2025, 271.
11.YANG C K, LEE C Y, WANG H S, et al. Glomerular disease classification and lesion identification by machine learning [J]. BIOMEDICAL JOURNAL, 2022, 45(4): 675-685.
12.BAUCKNEHT M, MARINI C, COSSU V, et al. Gene's expression underpinning the divergent predictive value of [18F]F-fluorodeoxyglucose and prostate-specific membrane antigen positron emission tomography in primary prostate cancer: a bioinformatic and experimental study [J]. JOURNAL OF TRANSLATIONAL MEDICINE, 2023, 21(1).
13.BERTIN D, BONGRAND P, BARDIN N. Comparison of the Capacity of Several Machine Learning Tools to Assist Immunofluorescence-Based Detection of Anti-Neutrophil Cytoplasmic Antibodies [J]. INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES, 2024, 25(6).
14.BOROWSKY J, HARUKI K, LAU M C, et al. Association of Fusobacterium nucleatum with Specific T-cell Subsets in the Colorectal Carcinoma Microenvironment [J]. CLINICAL CANCER RESEARCH, 2021, 27(10): 2816-2826.
15.CAI S Y, HU T, VENKATESAN M, et al. Multiplexed protein profiling reveals spatial subcellular signaling networks [J]. ISCIENCE, 2022, 25(9).
16.CHEN Y, CHEN Y Y, FU S S, et al. VGG16-based intelligent image analysis in the pathological diagnosis of IgA nephropathy [J]. JOURNAL OF RADIATION RESEARCH AND APPLIED SCIENCES, 2023, 16(3).
17.COSSELLU D, VIVADO E, BATTI L, et al. Volumetric atlas of the rat inner ear from microCT and iDISCO plus cleared temporal bones [J]. PEERJ, 2025, 13.
18.MENCATTINI A, TOCCI T, NUCCETELLI M, et al. Automatic classification of HEp-2 specimens by explainable deep learning and Jensen-Shannon reliability index [J]. ARTIFICIAL INTELLIGENCE IN MEDICINE, 2025, 160.
19.MUNTINGA C L P, VAN STEENWIJK P J D, KOOREMAN L F S, et al. Pre-existing infiltration with T cells and CD14+myeloid cells is associated with treatment response to imiquimod in primary and recurrent vulvar high-grade squamous intraepithelial lesions [J]. JOURNAL FOR IMMUNOTHERAPY OF CANCER, 2025, 13(10).
20.NESBIT M, MAMO J C, MAJIMBI M, et al. Automated Quantitative Analysis of ex vivo Blood-Brain Barrier Permeability Using Intellesis Machine-Learning [J]. FRONTIERS IN NEUROSCIENCE, 2021, 15.
21.QIN Y, HUTTLIN E L, WINSNES C F, et al. A multi-scale map of cell structure fusing protein images and interactions [J]. NATURE, 2021, 600(7889): 536-541.
22.RODRIGUEZ S G D, HERRERA P H, FLORES C A, et al. A Machine Learning Workflow of Multiplexed Immunofluorescence Images to Interrogate Activator and Tolerogenic Profiles of Conventional Type 1 Dendritic Cells Infiltrating Melanomas of Disease-Free and Metastatic Patients [J]. JOURNAL OF ONCOLOGY, 2022, 2.
23.ROJAS F, HERNANDEZ S, LAZCANO R, et al. Multiplex Immunofluorescence and the Digital Image Analysis Workflow for Evaluation of the Tumor Immune Environment in Translational Research [J]. FRONTIERS IN ONCOLOGY, 2022, 12.
24.SCHWARZE S, SCHAADT N S, SOBOTTA V M G, et al. Task design for crowdsourced glioma cell annotation in microscopy images [J]. SCIENTIFIC REPORTS, 2024, 14(1).
25.SHISHIDO S N, COURCOUBETIS G, KUHN P, et al. Automated Single-Cell Analysis in the Liquid Biopsy of Breast Cancer [J]. CANCERS, 2025, 17(17).
26.SUN A X, HAYAT H, LIU S H, et al. 3D in vivo Magnetic Particle Imaging of Human Stem Cell-Derived Islet Organoid Transplantation Using a Machine Learning Algorithm [J]. FRONTIERS IN CELL AND DEVELOPMENTAL BIOLOGY, 2021, 9.
27.TUOMISTO A, SIRNIö P, ELOMAA H, et al. Integrating Tumor Intraepithelial CD8+and Stromal FOXP3+T-Cell Densities as an Enhanced Immune Prognostic Index in Colorectal Cancer [J]. LABORATORY INVESTIGATION, 2025, 105(11).
28.GAVRIEL C G, DIMITRIOU N, BRIEU N, et al. Assessment of Immunological Features in Muscle-Invasive Bladder Cancer Prognosis Using Ensemble Learning [J]. CANCERS, 2021, 13(7).
29.GORMAN C, PUNZO D, OCTAVIANO I, et al. Interoperable slide microscopy viewer and annotation tool for imaging data science and computational pathology [J]. NATURE COMMUNICATIONS, 2023, 14(1).
30.HIRWAY S U, HASSAN N T, SOFRONIOU M, et al. Immunofluorescence Image Feature Analysis and Phenotype Scoring Pipeline for Distinguishing Epithelial-Mesenchymal Transition [J]. MICROSCOPY AND MICROANALYSIS, 2021, 27(4): 849-859.
31.JI X N, LI Q Y, LIU Z P, et al. Identification of Active Components for Sports Supplements: Machine Learning-Driven Classification and Cell-Based Validation [J]. ACS OMEGA, 2024, 9(10): 11347-11355.
32.JIA Y, YUAN L, WEN W J, et al. Circulating proteins and metabolites panel for noninvasive preoperative diagnosis of epithelial ovarian cancer [J]. BMC MEDICINE, 2025, 23(1).
33.JIN Y Y, BRENNECKE J, SODMANN A, et al. Antibody selection and automated quantification of TRPV1 immunofluorescence on human skin [J]. SCIENTIFIC REPORTS, 2024, 14(1).
34.KATARAS T J, JANG T J, KOURY J, et al. ACCT is a fast and accessible automatic cell counting tool using machine learning for 2D image segmentation [J]. SCIENTIFIC REPORTS, 2023, 13(1).
35.KEIKHOSRAVI A, ALMANSOUR F, BOHRER C H, et al. High-throughput image processing software for the study of nuclear architecture and gene expression [J]. SCIENTIFIC REPORTS, 2024, 14(1).
36.JIMéNEZ-SáNCHEZ D, ARIZ M, DE ANDREA C E, et al. Synplex: In Silico Modeling of the Tumor Microenvironment From Multiplex Images [J]. IEEE TRANSACTIONS ON MEDICAL IMAGING, 2023, 42(10): 3048-3058.
37.KIM G, JEON J H, PARK K, et al. High throughput screening of mesenchymal stem cell lines using deep learning [J]. SCIENTIFIC REPORTS, 2022, 12(1).
38.LANDEROS C, OH J, WEISSLEDER R, et al. Deep learning pipeline for automated cell profiling from cyclic imaging [J]. SCIENTIFIC REPORTS, 2024, 14(1).
39.MASCI A M, WHITE S, NEELY B, et al. Ontology-guided segmentation and object identification for developmental mouse lung immunofluorescent images [J]. BMC BIOINFORMATICS, 2021, 22(1).
40.MATSUDA K, UENO A, TSUZAKI J, et al. Vessels encapsulating tumor clusters contribute to the intratumor heterogeneity of HCC on Gd-EOB-DTPA-enhanced MRI [J]. HEPATOLOGY COMMUNICATIONS, 2025, 9(1).
41.MIAO K, LIN L E, QIAN C X, et al. Label-free Super-resolution Imaging Enabled by Vibrational Imaging of Swelled Tissue and Analysis [J]. JOVE-JOURNAL OF VISUALIZED EXPERIMENTS, 2022, (183).
42.NGUYEN H T, PIETRASZEK N, SHELTON S E, et al. Utilizing convolutional neural networks for discriminating cancer and stromal cells in three-dimensional cell culture images with nuclei counterstain [J]. JOURNAL OF BIOMEDICAL OPTICS, 2024, 29.
43.NOGUCHI Y, MURAKAMI M, MURATA M, et al. Microscopic image-based classification of adipocyte differentiation by machine learning [J]. HISTOCHEMISTRY AND CELL BIOLOGY, 2023, 159(4): 313-327.
Downloads
Published
Versions
- 2026-04-03 (2)
- 2026-01-01 (1)
