cover
Contact Name
Purwono
Contact Email
purwono@ptti.web.id
Phone
+6282113940427
Journal Mail Official
jahir@ptti.web.id
Editorial Address
Jl. Empu Sedah No. 12, Pringwulung, Condongcatur, Kec. Depok, Kabupaten Sleman, Daerah Istimewa Yogyakarta 55281, Indonesia
Location
Kab. sleman,
Daerah istimewa yogyakarta
INDONESIA
Journal of Advanced Health Informatics Research
ISSN : -     EISSN : 29856124     DOI : https://doi.org/10.59247/jahir.v1i1
Journal of Advanced Health Informatics Research (JAHIR) is a scientific journal that focuses on the application of computer science to the health field. JAHIR is a peer-reviewed open-access journal that is published three times a year (April, August and December). The scientific journal is published by Peneliti Teknologi Teknik Indonesia (PTTI). The JAHIR aims to provide a national and international forum for academics, researchers, and professionals to share their ideas on all topics related to Informatics in Healthcare Research
Articles 36 Documents
A Bibliometric Analysis of Knowledge Distillation in Medical Image Segmentation Muntiari, Novita Ranti; Rania Majdoubi; Rajiansyah
Journal of Advanced Health Informatics Research Vol. 2 No. 3 (2024)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/jahir.v2i3.297

Abstract

This study conducts a bibliometric analysis and systematic review to examine research trends in the application of knowledge distillation for medical image segmentation. A total of 806 studies from 343 distinct sources, published between 2019 and 2023, were analyzed using Publish or Perish and VOSviewer, with data retrieved from Scopus and Google Scholar. The findings indicate a rising trend in publications indexed in Scopus, whereas a decline was observed in Google Scholar. Citation analysis revealed that the United States and China emerged as the leading contributors in terms of both publication volume and citation impact. Previous research predominantly focused on optimizing knowledge distillation techniques and their implementation in resource-constrained devices. Keyword analysis demonstrated that medical image segmentation appeared most frequently with 144 occurrences, followed by medical imaging with 110 occurrences. This study highlights emerging research opportunities, particularly in leveraging knowledge distillation for U-Net architectures with large-scale datasets and integrating transformer models to enhance medical image segmentation performance
Virus Host Prediction with Metagenomic Features using Support Vector Machine Algorithm and Grid Search Cross Validation Optimization Purwono, Purwono; Annastasya Nabila Elsa Wulandari; Novieta Hardeani Sari
Journal of Advanced Health Informatics Research Vol. 2 No. 3 (2024)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/jahir.v2i3.298

Abstract

Viruses and bacteria continue to evolve alongside humans. Viruses are spreading too fast and causing a huge loss of life in the world. Viruses play an important role as dangerous pathogens that continue to spread various infectious diseases. Metegenomics is the application of large sequencing technology to genetic material obtained directly from one or more environmental samples, resulting in at least 50Mb random samples and multiple long sequences. It is important to identify the origin of the virus to prevent the spread of outbreaks. Understanding the biology of these viruses and how they affect their ecosystems depends on knowing which host they infect. We can use metagenomic features derived from the viral genome to determine the type of virus host. The activity of predicting virus hosts has traditionally taken a lot of time and effort in the process. Technology can be one of the solutions that can be used to predict virus host types. One of the technologies that can be used is machine learning. We chose one of the machine learning algorithms, SVM, to predict viral hosts with metagenomics features, namely genome size, GC% and number of CDS from viral genomes derived from 7326 viral genomes. The SVM model was further optimised with GS and K-CV methods. This optimisation resulted in an increase in the accuracy value of the model when predicting virus hosts from 80% to 84%.
Potential Use of U-Net and Fuzzy Logic in Diabetic Foot Ulcer Segmentation: A Comprehensive Review Rachman Hidayat; Annastasya Nabila Elsa Wulandari; Purwono, Purwono; Khoirun Nisa
Journal of Advanced Health Informatics Research Vol. 2 No. 3 (2024)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/jahir.v2i3.299

Abstract

Diabetic foot ulcer (DFU) image segmentation is still an interesting concern of researchers. Various new deep learning-based methods have been proposed to handle this image segmentation problem. Some research problems that are still faced by many researchers are dataset problems that are considered limited and need further clinical trials. The challenges of data problems include heterogeneity and image quality variations in the shape of skin lesions and subjectivity when annotating. The evaluation results from previous studies also show a considerable difference where there are still low accuracy results, but also too high accuracy is still found so that it is considered to have the potential for overfitting. As a result of the review of various related studies, there is an interesting potential of applying fuzzy logic to the U-Net architecture. This architecture has become very popular because it is widely used in medical image segmentation. The application of fuzzy logic can be applied to the U-Net architecture such as encoder, decoder, skip connection to adjust various U-Net parameters.
Classification of Skin Disease Images Using K-Nearest Neighbour (KNN) Ari Peryanto; Susanto, Dwi; Jihad, Bagus Hayatul
Journal of Advanced Health Informatics Research Vol. 2 No. 3 (2024)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/jahir.v2i3.300

Abstract

The skin is the outermost part of the human body that is often exposed to the environment, so it is easy to experience disease disorders. Some of the skin diseases that are often contracted in humans are ulcers, herpes, and warts. Untreated skin diseases will be very annoying because of the sensation of itching so it can cause irritation and inflammation. The ability to classify skin diseases using technology is one solution. This study uses the K-Nearest Neighbour (KNN) method to detect images of skin diseases. KNN is one of the machine learning methods with a calculation method based on the proximity of k. KNN was chosen because it is fast and has high-accuracy results. The results of the research that has been carried out have obtained results of accuracy of 63%, precision of 63%, recall of 63%, and F1 Score of 63%. From the results of the study, it can be concluded that disease detection using KNN has been successfully applied and can be used in classification.
Expert System For Diagnosis Of Gerd Disease Forward Chaining Methods Dhinur Aini, Fadhilah; Peryanto, Ari
Journal of Advanced Health Informatics Research Vol. 3 No. 1 (2025)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/jahir.v3i1.332

Abstract

This study presents the development of an expert system for diagnosing gastric diseases using the forward chaining method. The system is designed to assist patients in identifying possible conditions such as Gastroesophageal Reflux Disease (GERD), dyspepsia, and peptic ulcer based on reported symptoms through a web-based interface. The diagnosis process relies on a rule-based knowledge system that maps symptoms to disease categories and provides preliminary results along with simple treatment recommendations. The implementation demonstrates that the system can facilitate early screening and improve patient awareness. Nonetheless, it remains limited to common gastric diseases and depends on subjective symptom reporting. Accordingly, the system is intended as a supporting tool for early detection and patient guidance, rather than a substitute for clinical examination
Hybrid Ensemble Learning for Classifying Prescription vs. Over-the-Counter Medicines on Large-Scale Categorical and Textual Data Reina Melani; Dina Febrina
Journal of Advanced Health Informatics Research Vol. 3 No. 1 (2025)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/jahir.v3i1.341

Abstract

The classification of drugs into Prescription (Rx) and Over-the-Counter (OTC) categories is an important aspect of pharmaceutical governance because it has a direct impact on patient safety, drug access, and regulatory compliance. However, large-scale pharmaceutical data often consists of heterogeneous categorical variables and short texts, such as product names or indications, which poses challenges in the form of duplication, inconsistencies, and potential class imbalances. This condition demands a modeling approach that is not only accurate, but also lightweight and explainable. This study proposes a hybrid ensemble model that combines three algorithms, namely CART, Random Forest, and LightGBM, through a weighted soft-voting mechanism. This approach combines decision tree transparency with the reliability of modern boosting techniques. The main contribution of this study is to show that a low-complexity domain-based pipeline can produce accurate, efficient, and easily auditable Rx and OTC classifications for both clinical and regulatory needs. The pre-processing pipeline includes TF-IDF for short text, One-Hot Encoding for categorical features, as well as simple dosage variables. All features were combined into a solid matrix, then trained using weighted ensembles [1,1,8]. Evaluations include Accuracy, Precision, Recall, F1-score, ROC-AUC, Brier score, confusion matrix, and ROC curve. Test results on a dataset of 50,000 balanced samples showed consistent in-sample performance: Accuracy = 0.742; Accuracy = 0.742; Recall = 0.742; F1 = 0.742; ROC-AUC = 0.819; then Brier score = 0.214. The model is able to stably distinguish classes with a balance between False Positive and False Negative errors. In conclusion, this lightweight ensemble is able to present competitive prediction performance as well as interpretation, so that it has the potential to be applied to pharmacovigilance and drug classification. Further studies suggest adding cross-validation, probability calibration, as well as robustness tests to data outside the distribution to strengthen the reliability of the model

Page 4 of 4 | Total Record : 36