Zangana, Hewa Majeed and Omar, Marwan and Li, Shuai and Al-Karaki, Jamal N. and Vitianingsih, Anik Vega (2025) Small Object Detection in Medical Imaging Using Enhanced CNN Architectures for Early Disease Screening. Buletin Ilmiah Sarjana Teknik Elektro, 7 (3). pp. 595-607.

Text 14015-Article Text-64523-1-10-20251004.pdf Download (749kB)

Abstract

Early detection of subtle pathological features in medical images is critical for improving patient outcomes but remains challenging due to low contrast, small lesion size, and limited annotated data. The research contribution is a hybrid attention-enhanced CNN specifically tailored for small object detection across mammography, CT, and retinal fundus images. Our method integrates a ResNet-50 backbone with a modified Feature Pyramid Network, dilated convolutions for contextual scale expansion, and combined channel–spatial attention modules to preserve and amplify fine-grained features. We evaluate the model on public benchmarks (DDSM, LUNA16, IDRiD) using standardized preprocessing, extensive augmentation, and cross-validated training. Results show consistent gains in detection and localization: ECNN achieves an F1-score of 88.2% (95% CI: 87.4–89.0), mAP@0.5 of 86.8%, IoU of 78.6%, and a low false positives per image (FPPI = 0.12) versus baseline detectors. Ablation studies confirm the individual contributions of dilated convolutions, attention modules, and multi-scale fusion.

Item Type:	Artikel Umum
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisi / Prodi:	Faculty of Industrial Technology (Fakultas Teknologi Industri) > S1-Electrical Engineering (S1-Teknik Elektro)
Depositing User:	M.Eng. Alfian Ma'arif
Date Deposited:	03 Nov 2025 04:08
Last Modified:	03 Nov 2025 04:08
URI:	http://eprints.uad.ac.id/id/eprint/88422
Dosen Pembimbing:	UNSPECIFIED | [error in script]

Actions (login required)

View Item