Andriyan Bayu Suksmono

Background and Objective: Cephalometric radiography, which displays craniofacial structures, is a key tool in dentistry, serving as the gold standard for analyzing, evaluating, and diagnosing dental and skeletal relationships. Cephalometric analysis identifies landmarks and measures angles on radiographs. However, this process is time-consuming, inconsistent, and requires experience. Artificial intelligence (AI) has been extensively explored for cephalometric landmark identification to improve accuracy and efficiency. Despite efforts to develop automated landmark detection systems, they remain inadequate for orthodontic applications due to the lack of reliable datasets.Methods: To address this gap, we propose IndoCEPH, designed to support the development of robust AI solutions for quantitative morphometric analysis. Differences in facial morphology across various racial or ethnic groups are widely acknowledged, highlighting the importance of region-specific datasets. IndoCEPH was developed as a publicly available dataset consist of lateral cephalometric radiographs from the Indonesian population annotated with 17 anatomical landmarks by oral and maxillofacial radiology experts. Radiographs were selected using quality-based inclusion and exclusion criteria and included various malocclusion conditions. Dataset performance was evaluated using a deep learning-based landmark detection model and compared with a public dataset using localization metrics and detection performance metrics.Results: Experimental results using the IndoCEPH dataset demonstrated that the model achieved a mean radial error (MRE) of 1.00 mm and a successful detection rate (SDR) of 86.9% within a 2 mm range. The model also achieved a precision of 0.881, recall of 0.743, and mAP@50 of 0.731. Compared with the public dataset, IndoCEPH demonstrated improved overall landmark detection performance and successfully detected all 17 anatomical landmarks.Conclusions: The dataset provides valuable insights and fosters progress in landmark identification. Additionally, it can be applied in dental and skeletal age determination along with anthropometric comparisons across populations.

Social media has become a vital platform for user interaction, offering valuable insights into service quality for the video streaming industry. As competition intensifies, understanding user satisfaction through Quality of Experience (QoE) is critical for retaining users and improving services. Traditional QoE measurement methods are often costly and time-consuming, making efficient alternatives necessary. This study investigates the correlation between user sentiments on social media and QoE parameters—human, content, context, and system—while exploring sentiment analysis as a reliable tool for service improvement. A quantitative correlational design was employed, analyzing data from the official social media accounts of the top four global video streaming platforms and one leading service in Indonesia. After that, text mining techniques such as tokenization, removal of stop-words, and stemming were applied. However, they were labeled and classified into QoE parameters to prevent invalid user comments. The study employed correlation analysis to examine the connection between user sentiments and QoE parameters in video streaming services. A total of 2,556 data points were chosen for a complete analysis of the 11,280 data points obtained. Specifically, the analysis revealed a strong positive correlation of 98.16% between user sentiments and QoE parameters. The research results confirm that social media sentiment analysis is an efficient and accurate alternative to traditional QoE measurement approaches. This approach offers a data-based, practical evaluation of user satisfaction and QoE based on the analysis. The analysis concludes with practical recommendations for optimizing service delivery based on user perception in the video streaming industry.

Decoding speech imagery from brain signals potentially assists individuals with speech impairments. However, limited data and complex brain activity in recent studies have made accurate decoding challenging. We analyzed both the spectral (frequency) and spatial (location) aspects of brain activity to enhance decoding accuracy in small datasets. To our knowledge, previous studies have mainly used generated features without adequately considering spectro-spatial aspects. We trained machine learning with time-frequency representation (TFR) features using a public dataset from the Brain-Computer Interface (BCI) competition (BCI-DB) and our own recordings (PrimAudio-DB). The results showed prominent feature patterns of speech imagery in the frontal region and Gamma band, achieving accuracies of 98.6% in BCI-DB (exceeding benchmarks) and 81.7% in PrimAudio-DB. Moreover, our analysis revealed insights regarding speech differences (language and semantics). This study contributed to non-invasive speech imagery decoding and offered valuable insights for future speech rehabilitation and assistive technologies.

Presents corrections to the paper, (Corrections to “Depth Map Reconstruction for Scleral Bleb Models Using Structured Light in Light Field Imaging”).

Objectives: This review article aims to discuss the development of lateral cephalometric radiography use in science until now. Review: The search for studies on the identification of lateral cephalometric anatomical landmarks based on artificial intelligence was conducted by involving four databases: PubMed, IEEE Xplore, Google Scholar, and Scopus. The article selection was conducted using the keywords "Cephalometric Radiograph," "Automatic Cephalometric," "Cephalometric Landmarking," and "Cephalometric Digital" from January 2000 to March 2022. A total of 11 articles were obtained for this study. Cephalometric radiography is a radiographic technique that shows a picture of the skull and is widely used in dentistry to analyze and assess the relationship between teeth, jaws, and facial bones. Cephalometric analysis can be done by identifying anatomical landmark points and measuring angles on lateral cephalometric radiographs. The development of cephalometric radiography in biomedical imaging, especially in terms of the processing of cephalometric radiograph images from the process of forming X-rays to their potential use in the process of determining automatic anatomical landmark points. Conclusion: The results of the literature review of the development of dental radiology, especially digital cephalometric radiography, continue to increase, and its development is supported by computing technology, especially Artificial Intelligence. Keywords: Lateral cephalometric; biomedical imaging; artificial intelligence

Light field image processing offers promising opportunities for the development of three-dimensional object imaging technology in healthcare. However, implementing it in real-world scenarios often encounters numerous external challenges that can degrade the quality of captured data. One example is imaging blebs on the eye sclera after trabeculectomy. The surface of the blebs has a glossy appearance and, in some cases, contains minimal blood vessels, resulting in a predominantly white appearance. This condition leads to inaccuracies in depth map reconstruction from light field images. A depth map reconstruction method is introduced for bleb surface models on the eye sclera model, utilizing information from the center of the sub-aperture image as a boundary condition within a Markov Random Field. This approach reduces outliers at the edges of the depth map where intensity variations occur. Additionally, the use of structured light on textureless object surfaces serves as an auxiliary enhancement, adding contour detail to the clear and white bleb surface model. Four Heidelberg Collaboratory for Image Processing datasets were used to evaluate the performance of this approach, with the number of sub-aperture images reduced to match the amount of data generated by the light field camera employed in this study. Based on the tests conducted, the proposed method improves the accuracy of depth map reconstruction for the input dataset. Furthermore, the addition of structured light on textureless bleb model images significantly reduces the depth map reconstruction errors while improving the similarity index of the reconstructed depth maps compared to the ground truth.

We implemented a quantum version of the Kolmogorov-Arnold Network (KAN) on a quantum circuit. It optimizes the weight of each synaptic connection to enhance the overall performance of the neurons. This approach offers a significant boost in efficiency over traditional neural networks, as it can be optimized more accurately and requires only a few neurons to solve complex problems. The KAN is a multi-layer network, similar to a traditional neural network, with each layer represented as a matrix containing parameters and feedback. This structure is linear and mirrors the form of Variational Quantum Algorithms. Thus, we propose the Variational Quantum Kolmogorov-Arnold Network (VQKAN) as a quantum version of the Kolmogorov-Arnold Network. We demonstrated the optimization of VQKAN for fitting the given function in various ways, such as by using ansatzes and comparisons.

Microscopic detection of acid-fast bacilli (AFB) from mycobacterium tuberculosis (MTB) in Ziehl-Neelsen (ZN)-stained sputum samples is a crucial step in the detection of TB (tuberculosis) disease. Pathologists encounter many challenges that may result in incorrect diagnoses, such as the heterogeneous shape and irregular appearance of MTB, low-quality ZN staining, and errors in in scanning each of the field of view (FoV) using a conventional microscope. Additionally, multiple manual observations may cause fatigue that leads to human error. Several studies have created microscopic imaging databases of sputum samples, aiding researchers in creating computer-aided diagnosis (CAD) for tuberculosis, which is a promising method that offers timely, reliable, and repeatable assistance. Nevertheless, the implementation of CAD systems for TB diagnosis remains an area of ongoing research and development owing to the lack of microscopic image datasets of sputum samples, which represent whole-slide imaging (WSI) that follows the WHO (World Health Organization) regulations. To address this issue, this study developed a novel digitized microscopic image from sputum smear samples of Indonesian patients in the WSI that conform to the WHO regulation. These images are collected as a Microscopic Imaging Database of Tuberculosis Indonesia (MIDTI). This study also proposed a method based on the YOLOv7 (You Only Look Once seventh version) algorithm to develop a CAD for tuberculosis diagnosis by classifying ZN-stained sputum smear samples into International Union Against Tuberculosis and Lung Disease (IUATLD) grades, which has never been revealed in any previous studies.