Skripsi

DETEKSI RUANG JANTUNG JANIN MENGGUNAKAN METODE FASTER R-CNN DALAM KASUS INTER-PASIEN

---

Penilaian

---

Fetal heart examination using ultrasound imaging is a crucial step in the early detection of congenital heart defects. However, manually identifying fetal heart structures requires specialized expertise and is prone to subjectivity. This study proposes an automated system for fetal heart chamber detection using the Faster R-CNN method under an inter-patient scenario, where testing is conducted on data from patients not involved in training. Eight models were trained using combinations of ResNet50 and VGG16 backbones, two learning rates (0.0001 and 0.001), and two epoch settings (100 and 150), on an ultrasound image dataset labeled with 18 anatomical heart objects and a background class. Model performance was evaluated using metrics including Mean Average Precision (mAP), precision, recall, and F1-score. Results show that models with the ResNet50 backbone achieved the highest performance, with a maximum mAP of 99.79%, while the best VGG16 model reached 99.26%. These findings indicate that the Faster R-CNN method is effective in detecting fetal heart structures with high accuracy, even when applied to unseen patient data.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2507005400	T182707	T1827072025	Central Library (Reference)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T1827072025

Publisher

Indralaya : Prodi Sistem Komputer, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xxii, 549 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

006.320 7

Content Type

Text

Media Type

unmediated

Carrier Type

-

Edition

-

Subject(s)

Prodi Sistem Komputer
Jantung Janin

Specific Detail Info

-

Statement of Responsibility

KA

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment