Article Per Year (5 Year)
p-Index From 2021 - 2026
0.562
P-Index
This Author published in this journals
All Journal Jurnal Masyarakat Informatika Journal of Applied Data Sciences
Husni Fadhilah
School of Electrical Engineering and Informatics, Institut Teknologi Bandung
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

 1 
Transformer-Based Encoder-Decoder Model for Medical Image Captioning with Concept Embedding Husni Fadhilah; Nugraha Priya Utama
Jurnal Masyarakat Informatika Vol 17, No 1 (2026): May 2026
Publisher : Department of Informatics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/jmasif.17.1.75082

Abstract

This research presents a Transformer-based encoder-decoder model for medical image captioning that incorporates semantic medical knowledge through Concept Unique Identifiers (CUIs) from the Unified Medical Language System (UMLS). The proposed architecture employs a Swin Transformer as the visual encoder and GPT-2 as the language decoder, with CUI integration applied during both caption preprocessing and decoding. Experiments were conducted on the ROCOv2 dataset under two scenarios: baseline (raw captions) and enhanced (CUI-enriched captions). Quantitative evaluation using BLEU, ROUGE, CIDEr, and BERT-based metrics demonstrates that the CUI-integrated model outperforms several baselines, including CNN-LSTM, ViT-BioMedLM, and DeepSeek-VL, achieving a BLEU-1 score of 0.371, ROUGE-L of 0.305, CIDEr of 0.275, and PubMedBERTScore-F1 of 0.893. These results represent a 20.1% improvement in BLEU-1 and a 39.9% increase in ROUGE-L compared to the best-performing model before caption preprocessing (ViT-GPT2 with BLEU-1 = 0.309, ROUGE-L = 0.218). Qualitative assessment by expert radiologists further confirms enhanced diagnostic accuracy, descriptive completeness, and clinical relevance. This study introduces a novel integration of medical semantic knowledge into captioning models, offering a scalable solution for clinical decision support in resource-limited settings such as Indonesia.
Lightweight Brain Tumor Classification with Histogram Oriented Gradients (HOG) Features and Class-Weighted Support Vector Machine (SVM) Budi Warsito; Husni Fadhilah; Puspita Kartikasari; Arief Rachman Hakim
Journal of Applied Data Sciences Vol 7, No 1: January 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i1.1018

Abstract

Early detection of brain tumors via MRI is crucial for improving patient outcomes. This study investigates a lightweight machine learning approach for multiclass brain tumor classification (glioma, meningioma, pituitary tumor, or no tumor) using Histogram of Oriented Gradients (HOG) for feature extraction and a Support Vector Machine (SVM) classifier. This study utilizes the public Brain Tumor Classification MRI Kaggle dataset, consisting of 2870 training and 394 testing MRI images across four classes. After converting the MRIs to grayscale and resizing them to 16×16 pixels, this study extracts HOG features and applies Principal Component Analysis (PCA) to retain 98% of the variance. An SVM is then trained with a GridSearchCV-optimized kernel and hyperparameters, and a custom class-weighted variant is compared. The best model, a polynomial-kernel SVM with custom class weights, achieved 91.8% test accuracy (95% CI (confidence interval): 90.9-92.7) with an F1-score of 0.919 ± 0.01, outperforming the best unweighted SVM (accuracy 86.0% ± 0.02, F1≈0.847). These results demonstrate that HOG+SVM, with proper weighting for class imbalance, can effectively classify brain tumors on small datasets at low computational cost. The novelty of this work lies in demonstrating that an optimized, class-weighted SVM leveraging compact HOG-PCA features can deliver over 91.8% accuracy with strong generalization on small-scale MRI data, providing a viable and interpretable alternative to complex Convolutional Neural Network (CNN) models. Future work can explore CNN and hybrid feature fusion to improve accuracy and generalization further.
Lightweight Brain Tumor Classification with Histogram Oriented Gradients (HOG) Features and Class-Weighted Support Vector Machine (SVM) Budi Warsito; Husni Fadhilah; Puspita Kartikasari; Arief Rachman Hakim
Journal of Applied Data Sciences Vol 7, No 1: January 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i1.1018

Abstract

Early detection of brain tumors via MRI is crucial for improving patient outcomes. This study investigates a lightweight machine learning approach for multiclass brain tumor classification (glioma, meningioma, pituitary tumor, or no tumor) using Histogram of Oriented Gradients (HOG) for feature extraction and a Support Vector Machine (SVM) classifier. This study utilizes the public Brain Tumor Classification MRI Kaggle dataset, consisting of 2870 training and 394 testing MRI images across four classes. After converting the MRIs to grayscale and resizing them to 16×16 pixels, this study extracts HOG features and applies Principal Component Analysis (PCA) to retain 98% of the variance. An SVM is then trained with a GridSearchCV-optimized kernel and hyperparameters, and a custom class-weighted variant is compared. The best model, a polynomial-kernel SVM with custom class weights, achieved 91.8% test accuracy (95% CI (confidence interval): 90.9-92.7) with an F1-score of 0.919 ± 0.01, outperforming the best unweighted SVM (accuracy 86.0% ± 0.02, F1≈0.847). These results demonstrate that HOG+SVM, with proper weighting for class imbalance, can effectively classify brain tumors on small datasets at low computational cost. The novelty of this work lies in demonstrating that an optimized, class-weighted SVM leveraging compact HOG-PCA features can deliver over 91.8% accuracy with strong generalization on small-scale MRI data, providing a viable and interpretable alternative to complex Convolutional Neural Network (CNN) models. Future work can explore CNN and hybrid feature fusion to improve accuracy and generalization further.
Co-Authors Arief Rachman Hakim Budi Warsito Nugraha Priya Utama Puspita Kartikasari