Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Data Enhancement to Improve CNN Accuracy for Early Detection of Colorectal Cancer

Rektorat UDINUS 2025 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2025
Goals
Record
c51fc1ec-8b37-4398-a4c7-4a0f328b5a3c
Published
2026-10-30

Universitas Dian Nuswantoro conducted research on data enhancement to support the early detection of colorectal cancer through the classification of colon endoscopic images. The activity involved Athiyah, Ummi; Rustad, Supriadi; Soeleman, Moch Arief; Muljono, Muljono; Akrom, Muhamad; and Konoralma, Rahel Amanda as participants. The research implemented a Convolutional Neural Network model using the VGG-16 architecture. The model classified colon endoscopic images into three categories: normal, polyp, and cancer.

The activity applied several data-processing techniques. K-means clustering was used to manage outliers, while data augmentation was applied to increase dataset diversity. Pearson correlation analysis was conducted to confirm the relationship between the augmented dataset and the initial dataset. The processing stages produced an enhanced dataset that included outlier handling, augmentation, validation, and class balancing.

The results showed that CNN models trained with the enhanced dataset achieved a significant increase in accuracy and generalizability compared with models trained with the initial dataset. Performance evaluation reported an accuracy of 86 percent, together with notable improvements in the F1-score and recall for the cancer class. The findings indicated that the model classified images better after dataset enhancement. In relation to the stated goal, this UDINUS activity addresses health-focused research by applying computational methods to improve image classification for early colorectal cancer detection.

Reference https://www.scopus.com/pages/publications/105022197948?origin=resultslist