Publication Search

The rapid proliferation of Internet of Things (IoT) devices across diverse industries has significantly increased the vulnerability of IoT edge networks to sophisticated cyber threats. Traditional intrusion detection systems (IDS), such as signature-based and anomaly-based approaches, are often insufficient in addressing the dynamic and evolving nature of these threats. This study proposes a hybrid intrusion detection system (IDS) framework that combines supervised machine learning (ML) techniques with deep reinforcement learning (DRL) to enhance detection performance in real-time, resource-constrained IoT environments. The proposed framework utilizes supervised learning for initial traffic classification and DRL for adaptive decision-making, enabling the system to continuously learn and optimize its detection policies based on new attack patterns. The hybrid approach significantly improves detection accuracy and reduces false positives when compared to conventional signature-based and single-model ML systems. In addition to improved detection capabilities, the framework's computational efficiency allows it to operate effectively within the constraints of IoT devices, ensuring that it is suitable for large-scale deployments. Benchmark evaluations using publicly available datasets, such as NSL-KDD, IoT-23, and BoT-IoT, show that the hybrid IDS framework outperforms traditional methods, providing a more robust and adaptive solution to cybersecurity challenges in IoT edge networks. The findings of this study suggest that combining machine learning with deep reinforcement learning offers a promising approach to secure IoT environments and address the limitations of existing IDS techniques. Future work will explore enhancing real-time adaptability, scalability, and the detection of zero-day attacks in evolving IoT ecosystems.

The integration of Internet of Things (IoT) with cloud computing has revolutionized healthcare systems, offering scalable and real-time patient monitoring. However, optimizing response times and energy consumption remains crucial for efficient healthcare delivery. This research evaluates various algorithmic approaches for workload migration and resource management within IoT cloud-based healthcare systems. The performance of the implemented algorithm in this research, Hybrid Dynamic Programming and Long Short-Term Memory (Hybrid DP+LSTM), was analyzed against other six key algorithms, namely Gradient Optimization with Back Propagation to Input (GOBI), Deep Reinforcement Learning (DRL), improved GOBI (GOBI2), Predictive Offloading for Network Devices (POND), Mixed Integer Linear Programming (MILP), and Genetic Algorithm (GA) based on their average response time and energy consumption. Hybrid DP+LSTM achieves the lowest response time (82.91ms) with an energy consumption of 2,835,048 joules per container. The outcome of the analysis showed that Hybrid DP+LSTM have significant response times improvement, with percentage increases of 89.3%, 79.0%, 83.8%, 97.0%, 99.8%, and 99.94% against GOBI, GOBI2, DRL, POND, MILP, and GA, respectively. In terms of energy consumption, Hybrid DP+LSTM outperforms other approaches, with GOBI2 (3,664,337 joules) consuming 29.3% more energy, DRL (2,973,238 joules) consuming 4.9% more, GOBI (4,463,010 joules) consuming 57.4% more, POND (3,310,966 joules) consuming 16.8% more, MILP (3,005,498 joules) consuming 6.0% more, and the GA (3,959,935 joules) consuming 39.7% more. The result of ablation of the Hybrid DP+LSTM model achieves a 47.05% improvement over DP-only (156.57ms) and a 70.64% improvement over LSTM-only (282.41ms) in response time. On the energy efficiency side, Hybrid DP+LSTM shows 22.80% improvement over LSTM-only (3,671,51 joules), but 7.34% underperformance compared to DP-only (2,640,93). These research findings indicate that the Hybrid DP+LSTM technique provides the best trade-off between response time and energy efficiency. Future research should further explore hybrid approaches to optimize these metrics in IoT cloud-based healthcare systems.

Background: There are still shortcomings in the implementation of a truly safe and optimal thoracic examination protocol for toddlers in certain hospital settings. Furthermore, data related to direct radiation dose measurements and evaluation of the effectiveness of thoracic examination techniques for toddlers specifically in the local context in Indonesia are very limited. Objective: To examine the thoracic examination procedure that can be performed with a high level of safety without compromising the quality of diagnostic results and to evaluate the radiation exposure dose and thoracic examination techniques in toddlers at Hospital. Methodology: This study used a mixed methods approach with a convergent parallel design. Quantitative data were obtained from radiation dose measurements and examination parameters, while qualitative data were collected through observation, interviews, and group discussions, then analyzed thematically to understand the factors that influence radiation dose in infant thoracic examinations. Results: Research on thoracic radiology examinations in toddlers at Heart and Vascular Hospital was conducted systematically and in accordance with established procedures. Some limitations emerged from limited radiation dose records and inconsistent use of protective shields. Efforts to reduce radiation exposure include optimizing examinations, proper collimation, selecting exposure parameters, and educating families and staff. Continuous training and strict implementation of standard operating procedures (SOP) are essential to raise awareness of the ALARA principle. Internal policies and routine oversight are also needed to improve radiation dose monitoring, with the hope of improving the quality of radiology services and optimizing protection for toddler patients.

The integration of autonomous systems in traffic management has become increasingly important as urban populations and vehicle numbers continue to rise, leading to significant congestion. Traditional traffic signal control systems, which rely on fixed timing, are no longer sufficient to handle the dynamic and complex nature of urban traffic. To address these challenges, the proposed explainable Deep Reinforcement Learning (DRL) framework aims to optimize traffic signal control by dynamically adjusting traffic signals based on real-time data. This approach enhances traffic flow efficiency, reduces congestion, and improves overall system performance. The framework leverages Vehicle-to-Everything (V2X) communication, which enables real-time data exchange between vehicles, infrastructure, and other road users, extending the perception range of autonomous vehicles and providing valuable insights for traffic signal optimization. Additionally, the integration of smart infrastructure, such as smart intersections, plays a crucial role in enabling adaptive traffic management and facilitating better coordination across multiple intersections. One of the key advantages of the proposed system is its transparency, achieved through the implementation of explainable AI (XAI) techniques. These mechanisms provide clear insights into the decision-making processes, ensuring that traffic management authorities and system users can understand the rationale behind the system’s decisions. Although challenges such as data accuracy, scalability, and cybersecurity risks remain, the proposed DRL framework shows great promise in revolutionizing traffic management systems. Future research directions include enhancing data collection methods, improving the scalability of the system for larger cities, and further developing explainability features to improve trust and adoption in real-world applications.

Cybersecurity is continuously challenged by increasingly sophisticated and dynamic cyber-attacks, necessitating advanced adaptive defense mechanisms. Deep Reinforcement Learning (DRL) has emerged as a promising approach, offering significant advantages over traditional intrusion detection methods through real-time adaptability and self-learning capabilities. This paper presents an advanced adaptive cybersecurity framework utilizing five prominent DRL algorithms: Deep Q-Network (DQN), Proximal Policy Optimization (PPO), Twin Delayed DDPG (TD3), Soft Actor-Critic (SAC), and Asynchronous Advantage Actor-Critic (A3C). The effectiveness of these algorithms is evaluated within complex, realistic simulation environments using live-streaming data, emphasizing key metrics such as accuracy (AUC-ROC), response latency, and network throughput. Experimental results demonstrate that the SAC algorithm consistently achieves superior detection accuracy (95% AUC-ROC) and minimal disruption to network performance compared to other approaches. Additionally, A3C provides the fastest response times suitable for real-time defense scenarios. This comprehensive comparative analysis addresses critical research gaps by integrating both traditional and novel DRL techniques and validates their potential to substantially improve cybersecurity defense strategies in realistic operational settings.

Background: CT scans are used for various types of examinations such as examinations of the head, thorax, abdomen and so on. The most frequently performed CT scan is a CT scan of the head. However, in the medical field, especially radiology, CT scans have the highest dose compared to other radiation equipment. Efforts to reduce radiation doses and radiation effects are by estimating the correct dose given to CT Scan patients using an optimization index called DRL (Diagnostic Level Reference), namely by evaluating the values ​​of CTDIvol and DLP on CT scans of the head, which have been determined in accordance with BAPETEN/I-DRL 2021 standards. Method: This research is a descriptive quantitative study with an observational approach which aims to analyze CTDIvol and DLP values ​​based on retrospective data on non-contrast brain CT scans at the Radiology Installation at Hospital X, Central Jakarta. Results: Calculation of the 2nd quartile (50th percentile) value of CTDIvol and DLP in Adult Non-Contrast Brain CT Scan examinations for the January-December 2022 Period at the Radiology Installation at X Hospital, Central Jakarta is CTDIvol 55.51 mGy and DLP 867.00 mGy* cm. Conclusion: The 50 percentile value of CTDIvol and DLP in the Adult Non-Contrast Brain CT Scan examination for the January-December 2022 period at the Radiology Installation at X Hospital, Central Jakarta is in accordance with the standards set by BAPETEN/I-DRL 2021. Where is the value for CT Non-Contrast Head Scan, namely a CTDIvol value of 60 mGy and a DLP value of 1275 mGy*cm.

Background. Non-contrast head CT scans utilizing X-rays are considered the gold standard in emergency units for patients with clinical head injuries. Rotation time is a parameter that influences the radiation dose received by patients. This study aims to evaluate the impact of rotation time on patient radiation doses, emphasizing the ALARA principle. The research findings can assist in optimizing CT scan settings to reduce radiation doses without compromising image quality.. Methods: This quantitative research employs an experimental approach to investigate the influence of rotation time variations on CTDI (CT dose index) and DLP (dose leght product) in non-contrast head CT scans. Results: Calculation of CTDI and DLP values to assess the impact of rotation time variations, using two variations, 1 s and 1.5 s, yielded the following results: CTDI and DLP for 1 s were 20.30 mGy and 239.54 mGycm, respectively, while CTDI and DLP for 1.5 s were 20.43 mGy and 249.26 mGycm. Conclusion: Rotation time variations affect CTDI and DLP values in non-contrast head CT scans, although both values tend to remain stabel. A rotation time of 1 s is considered optimal for CTDI and DLP in non-contrast head CT scans at the Radiology Department of Sunset Vet Kuta Animal Hospital.

Background: Abdominal MSCT examination is a type of radiodiagnostic examination that uses the MSCT device. Considering that the abdomen is close to organs that have radiosensitive properties such as the gonads and ovaries, it is necessary to monitor the radiation dose received so that it does not exceed the predetermined I-DRL value. The prevalence of contrast Abdominal MSCT examinations in the Radiodiagnostic Installation at Dr Hasan Sadikin Bandung Hospital was recorded during the last 3 months, there were 327 Contrast Abdominal MSCT examinations out of a total of 880 examinations with a percentage of contrast Abdominal MSCT examinations of 0.37%. This proves that contrast Abdominal MSCT examinations are often carried out but evaluation has never been carried out regarding the radiation dose received by the patient. Method: The type of research used in this research is descriptive quantitative with an observational approach by collecting data from patients with contrast Abdominal MSCT examinations during the period January-March 2023 with a sample of 128 patients. The local DRL value is calculated using the 2nd quart formula in the SPSS statistical application, then this local DRL value is compared with the I-DRL value determined by BAPETEN. Results: Calculation of the 50th percentile value of CTDIvol and DLP on 128 samples. The results show a CTDIvol value of 15.91 mGy and DLP of 740 mGy*cm. The DRL (50 percentile) values ​​based on gender are 13.71 mGy and 743 mGy*cm for men, and 17.74 mGy and 792 mGy*cm for women. Based on abdominal thickness, patients with a thickness of 10-19 cm have values ​​of 12.19 mGy and 587 mGy*cm, while patients with a thickness of 20-30 cm have values ​​of 22.57 mGy and 965 mGy*cm. For the most clinical cases (Ca Recti) the values ​​were 11.27 mGy and 586 mGy*cm. Conclusion: The 2nd quartile value (50 percentile) of CTDIvol and DLP received by patients during the Contrast Abdominal MSCT examination at the Radiodiagnostic Installation of RSUP Dr. Hasan Sadikin Bandung is in accordance with the standard values ​​set by BAPETEN/I-DRL 2021. However, special attention is needed for patients with an abdominal thickness of 20-30 cm, where the CTDIvol value exceeds the standards set by BAPETEN/I-DRL 2021.

Percutaneous Coronary Intervention (PCI) dikenal dengan angioplasty, merupakan prosedur non bedah yang dilakukan untuk mengobati arteri koroner stenotik (penyempitan) pada penderita jantung koroner. Pemeriksaan PCI dilakukan dengan menggunakan fluoroskopi dengan durasi fluorotime yang relatif lama dan resiko radiasi yang diterima semakin tinggi sehingga perlu adanya optimisasi dosis. Diagnostic Reference Level (DRL) merupakan salah satu alat yang digunakan untuk optimasi dosis radiasi. Tujuan  dari Diagnostic Reference Level (DRL) adalah untuk mengoptimalkan proteksi dan keselamatan radiasi pasien, dan mencegah paparan radiasi yang tidak perlu. Nilai DRL Nasional ditentukan pada nilai kuartil 3 (75 persentil) dari data sebaran dosis yang didapat dari fasilitas kesehatan. Penelitian ini menggunakan data retrospektif, Data dosis yang digunakan untuk membandingkan DRL dalam Penelitian ini menggunakan data sekunder dari aplikasi Sistem Informasi Data Dosis Pasien (SiINTAN) tahun 2021 dengan total 187 pasien pada pemeriksaan Coronary Angiography. Dengan tujuan untuk mengevaluasi dosis radiasi DAP (Dose Area Product) dan Air Kerma pada pasien pemerisaan PCI. Dari perhitungan statistik 75 persentil (kuartil ke 3) pemeriksaan PCI pada dosis DAP yaitu 146.48 Gy.cm2 dan pada Air Kerma yaitu 2385 mGy, dan pada pasien yang sama tetapi terlebih dahulu melakukan pemeriksaan coronary angiography di RSUP Prof.Dr. I.G.N.G Ngoerah di tahun 2021 memiliki nilai persentil 75 pada DAP yaitu 45.83 Gy.cm2 dan pada dosis Air Kerma nilai yaitu 609 mGy. Perbandingan dengan negara lain nilai DRL Jepang 2020 bernilai 59 Gy.cm2 dan 700 mGy,  Finland 2016 30 Gy.cm2.