AIAnggun Islami

Anggun Islami

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Anggun Islami has a total of 31 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Deep learning-based classification of acanthosis nigricans as a clinical sign of high insulin resistance in type 2 diabetes mellitus

    by Annisa Darmawahyuni, Siti Nurmaini, Sutarno Sutarno, Anggun Islami, Muhammad Raffly Riyansyah Putera, Muhammad Mukhlashyn Az Zahrawi

    Acanthosis nigricans (AN) is a cutaneous manifestation strongly linked to insulin resistance and often serves as an early indicator of type 2 diabetes mellitus (T2DM). Early detection of AN is therefore essential for preventive healthcare. Conventional diagnosis typically relies on visual inspection and grading, which may be subjective and inconsistent. To address this limitation, this study proposes a computer-aided diagnostic framework based on deep learning for automatic AN detection from skin fold images. A comprehensive evaluation was conducted using nine widely adopted deep learning architectures, namely VGG16, VGG19, Dense-Net121, DenseNet201, ResNet50, ResNet101, MobileNetV2, Xception, and the vision transformer (ViT). To ensure a fair comparison, all models were trained under identical experimental settings. Model performance was assessed using standard classification metrics. The comparative analysis demonstrates that ResNet101 outperformed the other architectures, achieving an accuracy of 94.12% for AN detection, with a precision and recall of 96.55% and an F1-score of 93.33%. Overall, the results highlight the robustness and effectiveness of convolutional neural network (CNN) based models, particularly ResNet101, as non-invasive and objective tools for automated AN screening. This approach has the potential to support early detection of individuals at risk of T2DM and to enhance preventive healthcare delivery.

  • Open Access

    Integrating Prompt-Guided Pipeline of YOLOv8 and Segment Anything Model for Cervical Precancerous Lesion Segmentation

    by Akhiar Wista Arum, Siti Nurmaini, Muhammad Naufal Rachmatullah, Bambang Tutuko, Firdaus Firdaus, Annisa Darmawahyuni, Ade Iriani Sapitri, Anggun Islami, Ghita Athalina, Muhammad Alpim Alfa Rolis

    Background: Segmenting precancerous cervical lesions from cervicography images remains challenging due to irregular lesion morphology, blurred boundaries, low contrast, and visual similarity between abnormal and normal cervical tissue. Automated segmentation methods are needed to support more objective and efficient cervical cancer screening. Objective: This study aimed to develop and evaluate a prompt-guided integration framework that integrates variants of YOLOv8 with the Segment Anything Model (SAM) for the detection and segmentation of precancerous cervical lesions. Methods: A total of 1,350 cervicography images were used, consisting of 300 primary clinical images (TeleOTIVA Dataset) and 1,050 public images from the International Agency for Research on Cancer (IARC). The dataset was divided into 80% training data and 20% validation data. Five variants of YOLOv8—namely YOLOv8n, YOLOv8s, YOLOv8m, YOLOv8l, and YOLOv8x—were evaluated for detecting the cervical region, columnar region, and lesion classes. The best YOLOv8 model was then used to generate bounding box prompts for SAM-based segmentation using the ViT-B, ViT-L, and ViT-H encoders. Results: YOLOv8m achieved the best overall detection performance, with an average mAP of 65.7%. In the segmentation stage, ViT-L SAM achieved the best performance. The optimal model configuration depends on the object being segmented. YOLOv8m–ViT-L SAM delivered the best performance in lesion segmentation, with an IoU of 0.591 and a Dice coefficient of 0.705. Statistical analysis shows that YOLOv8m–ViT-L SAM achieved the best average score for lesion segmentation and significantly outperformed the evaluated comparison models. Conclusion: The proposed YOLOv8m-SAM pipeline demonstrates the feasibility of detection-guided segmentation using prompts generated by YOLOv8. However, more detailed lesion segmentation remains challenging and requires further improvement through prompt refinement, domain adaptation, and validation on larger, multicenter datasets. Keywords: Cervicography, Precancerous Cervical Lesions, Segmentation, Segment Anything Model, YOLOv8

  • Open Access

    Delineating 12-lead ECG for automated ST-elevation and ST depression detection using deep learning

    by Bambang Tutuko, Annisa Darmawahyuni, Alexander Edo Tondas, Muhammad Naufal Rachmatullah, Firdaus Firdaus, Ade Iriani Sapitri, Anggun Islami, Sukemi Sukemi, Muhammad Fachrurrozi, Siti Nurmaini, et al.

    ST-elevation or ST-depression are markers of an abnormal heart condition detected through an electrocardiogram (ECG) where the tracing in the ST-segment is unusually elevated above the TP-segment (baseline). Identifying the localization of the ST-segment on an ECG is difficult because even a minor change in the ST-segment can be obscured by filtering processes. The 12-lead ECG signal is a non-invasive tool in the early detection of ST-elevation based on ST- and TP-segment, with quick and accurate interpretation. This study proposes a standard 12-lead ECG delineation model using deep learning (DL). The ECG signal has been segmented to Pstart–Pend, Pend–QRSstart, QRSstart–Rpeak, Rpeak–QRSend, QRSend-Tstart, Tstart–Tend, and Tend–Pstart. The study interpreted ST-elevation or -depression using an ECG delineation approach guided by medical rules. The findings revealed that the DL model achieved an average accuracy of 99.18%, sensitivity of 92.55%,specificity of 99.55%, precision of 92.61%, and F1-score of 92.52% in limb leads. Similarly, in chest leads, the DL model attained an accuracy of 99.16%, sensitivity of 93.10%, specificity of 99.53%, precision of 93.32%, and F1-score of 93.11%. This study also validated the DL-predicted results by a cardiologist from Mohammad Hoesin Hospital, Indonesia.

  • Open Access

    Enhancing ultrasound image quality using deep structure of residual network

    by Ade Iriani Sapitri, Siti Nurmaini, Muhammad Naufal Rachmatullah, Annisa Darmawahyuni, Firdaus Firdaus, Anggun Islami, Bambang Tutuko, Akhiar Wista Arum

    Ultrasonography, a medical imaging technique, is often affected by various types of noise and low brightness, which can result in low image quality. These drawbacks can significantly impede accurate interpretation and hinder effective medical diagnoses. Therefore, improving image quality is an essential aspect of the field of ultrasound systems. This study aims to enhance the quality of ultrasound images using deep learning (DL). The experiment is conducted using a custom dataset consisting of 2,175 infant heart ultrasound images collected from Indonesian hospitals, and the model is subsequently generalized using other datasets. We propose enhanced deep residual network combined convolutional neural networks (EDR-CNNs) to improve the image quality. After the enhancement process, our model achieved peak signal-to-noise ratio (PSNR) and structural similarity index metrics (SSIM) scores of 38.35 and 0.92 respectively, outperforming other methods. The benchmarking with other ultrasound medical images indicates that our proposed model produces good performance, as evidenced by higher PSNR, lower SSIM, a decrease in mean square error (MSE), and a lower contrast improvement index (CII). In conclusion, this study encapsulates the forthcoming trends in advancing low-illumination image enhancement, along with exploring the prevailing challenges and potential directions for further research.


  • Open Access

    iCOR: End-to-End Electrocardiography Morphology Classification Combining Multi-Layer Filter and BiLSTM

    by Siti Nurmaini, Wisnu Jatmiko, Satria Mandala, Bambang Tutuko, Erwin Erwin, Alexander Edo Tondas, Annisa Darmawahyuni, Firdaus Firdaus, Muhammad Naufal Rachmatullah, Ade Iriani Sapitri, et al.

    Accurate delineation of ECG signals is critical for effective cardiovascular diagnosis and treatment. However, previous studies indicate that models developed for specific datasets and environments perform poorly when used with varying ECG signal morphology characteristics. This paper presents a novel approach to ECG signal delineation using a multi-layer filter (MLF) combined with a bidirectional long short-term memory (BiLSTM) model, namely iCOR. The proposed iCOR architecture enhances noise removal and feature extraction, resulting in improved classification of the P-QRS-T-wave morphology with a simpler model. Our method is evaluated on a combination of two standard ECG databases, the Lobachevsky University Electrocardiography Database (LUDB) and QT Database (QTDB). It can be observed that the classification performance for unseen sets of LUDB datasets yields above 90.4% and 98% accuracy, for record-based and beat-based approaches, respectively. Beat-based approaches outperformed the record-based approach in overall performance metric results. Similar results were shown in an unseen set of the QTDB, in which beat-based approaches performed with accuracy above 97%. These results highlight the robustness and efficacy of the iCOR model across diverse ECG signal datasets. The proposed approach offers a significant advancement in ECG signal analysis, paving the way for more reliable and precise cardiac health monitoring.


  • Open Access

    Multi-scale input reconstruction network and one-stage instance segmentation for enhancing heart defect prediction rate

    by Sutarno Sutarno, Siti Nurmaini, Ade Iriani Sapitri, Muhammad Naufal Rachmatullah, Bambang Tutuko, Annisa Darmawahyuni, Firdaus Firdaus, Anggun Islami, Samsuryadi Samsuryadi

    Artifacts and unpredictable fetal movements can hinder clear fetal heart imaging during ultrasound scans, complicating anatomical identification. This study presents a new medical imaging approach that combines one-stage instance segmentation with ultrasound (US) video enhancement for precise fetal heart defect detection. This innovation allows real-time identification and timely medical intervention. The study acquired 100 fetal heart US videos from an Indonesian Hospital featuring cardiac septal defects, generating 1,000 frames for training, validation, and testing. Utilizing a combination of the multi-scale input reconstruction network (MIRNet) for image enhancement and YOLOv8l-seg for real-time instance segmentation, the method achieved outstanding validation results, boasting a 99.50% mAP for bounding box prediction and 98.40% for mask prediction. It delivered a remarkable real-time processing speed of 68.4 frames per second. In application to new patients, the method yielded a 65.93% mAP for bounding box prediction and 57.66% for mask prediction. This proposed approach offers a promising solution to early fetal heart defect detection using ultrasound, holding substantial potential for enhancing healthcare outcomes.

  • Open Access

    Empowering AI-Diagnosis: Deep Learning Abilities for Accurate Atrial Fibrillation Classification

    by Bambang Tutuko, Annisa Darmawahyuni, Siti Nurmaini, Muhammad Naufal Rachmatullah, Firdaus Firdaus, Sutarno Sutarno, Rossi Passarella, Ade Iriani Sapitri, Anggun Islami, Muhammad Wahyu Ramansyah, et al.

    Artificial intelligence (AI) is a powerful technology that can enhance clinical decision-making and the efficiency of global health systems. An AI-enabled electrocardiogram (ECG) is an essential tool for diagnosing heart abnormalities such as arrhythmias. The most prevalent arrhythmia globally is atrial fibrillation (AF), which is an irregular heart rhythm that originates in the atria and can lead to other heart-related complications. A trusted AI classification of AF is explored in this study. Deep learning (DL) has been used to analyze large amounts of publicly available ECG datasets in order to classify normal sinus rhythm (NSR), AF, and other types of arrhythmias. A convolutional neural network (CNN) has been proposed to extract ECG features and classify ECG signals. Based on a 10-fold cross-validation strategy, we conducted experiments involving three scenarios for AF classification: (i) a balanced set, an imbalanced set, and an extremely imbalanced set; (ii) a comparison of ECG denoising algorithms; and (iii) the classification of AF, NSR, and other arrhythmia types (15 classes). As a result, we have achieved 100% accuracy, sensitivity, specificity, precision, and F1-score for the AF, NSR, and non-AF classifications, both for balanced and imbalanced sets. In addition, for the classification of AF, NSR, and other types of arrhythmia (15 classes), the performance results achieved an accuracy of 99.77%, sensitivity of 96.48%, specificity of 99.87%, precision of 97.03%, and F1-score of 96.68%. The results can empower AI diagnosis and assist clinicians in classifying AF on routine screening ECGs.


  • Open Access

    Accurate Fetal QRS-Complex Classification from Abdominal Electrocardiogram Using Deep Learning

    by Annisa Darmawahyuni, Bambang Tutuko, Siti Nurmaini, Muhammad Naufal Rachmatullah, Muhammad Ardiansyah, Firdaus Firdaus, Ade Iriani Sapitri, Anggun Islami

    Fetal heart monitoring during pregnancy plays a critical role in diagnosing congenital heart disease (CHD). A noninvasive fetal electrocardiogram (fECG) provides additional clinical information for fetal heart monitoring. To date, the analysis of noninvasive fECG is challenging due to the cancellation of maternal QRS-complexes, despite significant advances in electrocardiography. Fetal QRS-complex is highly considered to measure fetal heart rate to detect some fetal abnormalities such as arrhythmia. In this study, we proposed a deep learning (DL) framework that stacked a convolutional layer and bidirectional long short-term memory for fetal QRS-complexes classification. The fECG signals are first preprocessed using discrete wavelet transform (DWT) to remove the noise or inferences. The following step beats and QRS-complex segmentation. The last step is fetal QRS-complex classification based on DL. In the experiment of Physionet/Computing in Cardiology Challenge 2013, this study achieved 100% accuracy, sensitivity, specificity, precision, and F1-score. A stacked DL model demonstrates an effective tool for fetal QRS-complex classification and contributes to clinical applications for long-term maternal and fetal monitoring.


  • Open Access

    XU-NetI: Simple U-Shaped Encoder-Decoder Network for Accurate Imputation of Multivariate Missing Data

    by Firdaus Firdaus, Siti Nurmaini, Bambang Tutuko, Muhammad Naufal Rachmatullah, Anggun Islami, Annisa Darmawahyuni, Ade Iriani Sapitri, Widya Rohadatul Ais’sy, Muhammad Irfan Karim

    Intensive care unit (ICU) patients often have multiple vital signs monitored continuously. However, missing data is common in ICU settings, negatively impacting clinical decision-making and patient outcomes. In this study, we propose a multivariate data imputation method based on simple U-Shaped encoder-decoder network imputation (XU-NetI) method to learn the underlying patterns in the data and generate imputations for missing values of vital signs data with ICU patients. To evaluate the performance of our imputation methods, we employed a publicly available database such the medical information mart for intensive care III (MIMIC III) v1.4. Our proposed model has been developed to analyze 219.281 vital sign worth of data, focusing on eight essential vital sign features: body temperature, heart rate, respiration rate, systolic blood pressure, diastolic blood pressure, mean blood pressure, oxygen saturation, and glucose. The evaluation results demonstrates the effectiveness of the imputation techniques in improving the accuracy of predictive models. We compared our approach to other state-of-the-art imputation methods including Autoencoder and Convolutional Neural Networks. As a result found, our technique with XU-NetI architecture outperformed them, in terms of root mean square error (RSME) by approximately 0.01, mean absolute error (MAE) by approximately 0.009, and R square (R2) by approximately 0.99. Our method has the potential to enhance clinical decision-making and improve patient outcomes.