arrow
Return

Real-Time Multi-Task Deep Learning Model for Polyp Detection, Characterization, and Size Estimation

delete2025-01-01
delete0
delete
OA
AI
P
Phanukorn Sunthornwetchapong
K
Kasichon Hombubpha
K
Kasenee Tiankanon
S
Satimai Aniwan
P
Pasit Jakkrawankul
N
Natawut Nupairoj
P
Peerapon Vateekul *
R
Rungsun Rerknimitr *
DOI:10.1109/ACCESS.2025.3527720delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
While performing a colonoscopy, there are many tasks to be done: finding polyps, classifying them, and deciding the next procedure for the polyps, whether to incise them or not. Such tasks are challenging for fellow doctors. All these three tasks can have an intrapersonal error, which varies among endoscopists. A proven method for enhancing performance is computer-aided detection and a diagnosis system for endoscopists, which tends to be a real-time system. In this work, we present a modified convolutional neural network (CNN) based deep learning (DL) model to perform these tasks in real-time, utilizing existing object detection models: YOLOv5 and YOLOv8. For the various tasks, the models are trained using datasets with incomplete labels, leading to a comparison of different training strategies. Our model, YOLOv8, achieved an F1-score of 95.96% for the polyp detection task, 85.24% F1-score for the polyp classification task, and 78.41% macro F1-score for the polyp size estimation task. Such results, when compared with fellow doctors' findings proved superior in both accuracy and macro F1-score, maintaining a real-time inference speed.
Keywords:
Real-time systems
YOLO
Estimation
Accuracy
Endoscopes
Colon
Training
Medical diagnostic imaging
Mathematical models
Convolutional neural networks
Colonic polyp
deep learning
real-time image classification
real-time object detection
real-time size estimation

Journal

IEEE Access cover
IEEE Access
IF:
3.6
Papers:
9.8W
Citations:
29.4W

Organization

C
Chulalongkorn University
Scholars:
1.8W
Papers: 1.4W
Citations: 1.5W
Cited Papers

Cited Papers

Polyp Segmentation With the FCB-SwinV2 Transformer
err2024-01-01
err5
errOAAI
errFitzgerald, Kerr; Bernal, Jorge; Histace, Aymeric; Matuszewski, Bogdan J.
errShare
errSave
errShare
errSave
Endoscopic overestimation of colorectal polyp size
err2016-01-01
err77
PREAI
errAnderson, Bradley W.; Smyrk, Thomas C.; Anderson, Kari S.; Mahoney, Douglas W.; Devens, Mary E.; Sweetser, Seth R.; Kisiel, John B.; Ahlquist, David A.
errShare
errSave
Endoscopic prediction of deep submucosal invasive carcinoma: validation of the Narrow-Band Imaging International Colorectal Endoscopic (NICE) classification
err2013-10-01
err346
PREAI
errHayashi, Nana; Tanaka, Shinji; Hewett, David G.; Kaltenbach, Tonya R.; Sano, Yasushi; Ponchon, Thierry; Saunders, Brian P.; Rex, Douglas K.; Soetikno, Roy M.
errShare
errSave
Medical image analysis using deep learning algorithms
err2023-11-07
err74
errOAAI
errLi, Mengfang; Jiang, Yuanyuan; Zhang, Yanzhou; Zhu, Haisheng
errShare
errSave
Artificial intelligence-based measurement outperforms current methods for colorectal polyp size measurement
err2022-05-19
err25
PREAI
errKwak, Min Seob; Cha, Jae Myung; Jeon, Jung Won; Yoon, Jin Young; Park, Jong Wook
errShare
errSave
SiPh Self-coherent Transmitter Circuit with on-chip CSPR Control Capability based on a Tunable Power Splitter
err2019-01-01
err0
PREAI
errZhenping Xing; David Patel; Eslam El-fiky; Meng Xiang; Rui Li; Md Ghulam Saber; Luhua Xu; Michael Hui; David V. Plant
errShare
errSave
researcher View more