Muhammad Izzad Ramli
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Muhammad Izzad Ramli has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
Jian Qin
Faculty of Health Sciences, Universiti Teknologi MARA, Shah Alam, Selangor, Malaysia
3 shared publicationsJiang Li
Faculty of Health Sciences, Universiti Teknologi MARA, Shah Alam, Selangor, Malaysia
3 shared publications- Nursuriati Jamil
Universiti Teknologi MARA
3 shared publications - Ahmad Khusaini Mohd Kharip Shah
Agro-Biotechnology Institute (ABI)
2 shared publications - Ahmad Nazem Norali
GBA Corporation Sdn Bhd
2 shared publications - Ismail Mamat
Agro-Biotechnology Institute (ABI)
2 shared publications
Latest Publications
- Open Access
- 1 May 2026
- Published by Taylor & Francis in Clinical Interventions in Aging
- Vol. ume 21, 1-15
Deep Learning-Assisted Radiomics for Predicting Thoracolumbar Osteoporotic Vertebral Fractures
Rationale and Objectives: Osteoporotic vertebral compression fractures (OVFs), particularly at the thoracolumbar spinal junction spanning levels T11 to L2, represent a significant and debilitating global health challenge. The objective of the present research was to establish and verify a robust radiomics model using routine non-contrast computed tomography (CT) to predict thoracolumbar OVFs risk. Materials and Methods: In this retrospective cohort study, 80 patients with new thoracolumbar OVFs were propensity-matched 1:2 with 160 controls. A 3D U-Net automatically segmented T11-L2 cortical and cancellous bone. Following minimum redundancy maximum relevance (mRMR) and least absolute shrinkage and selection operator (LASSO) feature selection, 13 cortical and 16 cancellous radiomics features were extracted. Logistic regression models were developed using corresponding radscores. The best-performing model was compared to volumetric bone mineral density (vBMD) and evaluated via ROC curves, decision curve analysis (DCA), and calibration plots. Results: A total of 240 patients (147 females, 93 males) were enrolled, with no significant age or sex differences between groups. The combined cortical and cancellous radscore model (vertebral model) attained an AUC of 0.825 and 0.840 for OVF prediction, outperforming the vBMD model (AUC: 0.752/0.735). DCA and calibration plots verified its outstanding predictive performance. Notably, integrating vBMD with the vertebral model did not yield a statistically significant improvement (p > 0.05). The selected features highlighted crucial microstructural insights, with cancellous features reflecting trabecular heterogeneity and cortical features indicating mineralization uniformity and integrity. Conclusion: Our findings demonstrate that radiomics derived from routine non-contrast CT, leveraging deep learning for automated bone compartment segmentation, offers a superior and practical tool for Early identification of high-risk individuals for thoracolumbar OVFs. This approach provides valuable, complementary information beyond vBMD, potentially enhancing clinical decision-making and reducing the burden of osteoporosis.
- 1 May 2026
- Published by Taylor & Francis in Clinical Interventions in Aging
- Vol. ume 21, 1-15
Open AccessDeep Learning-Assisted Radiomics for Predicting Thoracolumbar Osteoporotic Vertebral Fractures
Rationale and Objectives: Osteoporotic vertebral compression fractures (OVFs), particularly at the thoracolumbar spinal junction spanning levels T11 to L2, represent a significant and debilitating global health challenge. The objective of the present research was to establish and verify a robust radiomics model using routine non-contrast computed tomography (CT) to predict thoracolumbar OVFs risk. Materials and Methods: In this retrospective cohort study, 80 patients with new thoracolumbar OVFs were propensity-matched 1:2 with 160 controls. A 3D U-Net automatically segmented T11-L2 cortical and cancellous bone. Following minimum redundancy maximum relevance (mRMR) and least absolute shrinkage and selection operator (LASSO) feature selection, 13 cortical and 16 cancellous radiomics features were extracted. Logistic regression models were developed using corresponding radscores. The best-performing model was compared to volumetric bone mineral density (vBMD) and evaluated via ROC curves, decision curve analysis (DCA), and calibration plots. Results: A total of 240 patients (147 females, 93 males) were enrolled, with no significant age or sex differences between groups. The combined cortical and cancellous radscore model (vertebral model) attained an AUC of 0.825 and 0.840 for OVF prediction, outperforming the vBMD model (AUC: 0.752/0.735). DCA and calibration plots verified its outstanding predictive performance. Notably, integrating vBMD with the vertebral model did not yield a statistically significant improvement (p > 0.05). The selected features highlighted crucial microstructural insights, with cancellous features reflecting trabecular heterogeneity and cortical features indicating mineralization uniformity and integrity. Conclusion: Our findings demonstrate that radiomics derived from routine non-contrast CT, leveraging deep learning for automated bone compartment segmentation, offers a superior and practical tool for Early identification of high-risk individuals for thoracolumbar OVFs. This approach provides valuable, complementary information beyond vBMD, potentially enhancing clinical decision-making and reducing the burden of osteoporosis.
- Open Access
- 1 April 2026
- Published by Taylor & Francis in International Journal of General Medicine
- Vol. ume 19, 1-11
Deep Learning-Based Assessment of the Value of Vertebral Structural Parameters in Predicting Osteoporotic Vertebral Compression Fractures on Opportunistic CT Scans
Background: Osteoporotic vertebral compression fractures are common fragility fractures in older adults and are associated with substantial disability and healthcare burden. Opportunistic CT may provide a practical no-extra-radiation pathway for fracture-risk assessment, but quantitative vertebral structural parameters, especially cortical parameters, remain insufficiently studied. Purpose: To evaluate the value of vertebral structural parameters derived from opportunistic CT in identifying osteoporotic vertebral compression fractures and to develop a nomogram for individualized risk estimation. Methods: This retrospective single-center study included 298 patients aged 45 years or older who underwent chest or abdominal CT at the Second Affiliated Hospital of Shandong First Medical University between January 2020 and May 2024. Osteoporotic vertebral compression fracture status on sagittal CT was determined by two readers (one radiology resident and one senior physician) by consensus. A high-resolution 3D Dense-U-Net was used for automated vertebral segmentation and extraction of L1 structural parameters. Group comparisons were performed with t tests, one-way analysis of variance, and chi-squared tests as appropriate. Independent predictors were identified with univariate and multivariate logistic regression, and the nomogram was evaluated with receiver operating characteristic and calibration analyses. Results: Of the 298 participants (182 men and 116 women; mean age, 62.33 ± 9.56 years), 134 had osteoporotic vertebral compression fractures and 164 did not. In multivariate analysis, L1 cortical average area (OR, 0.99; 95% CI, 0.99– 1.00; P = 0.002) and L1 cortical average thickness (OR, 0.22; 95% CI, 0.13– 0.38; P < 0.001) were independent predictors. The nomogram achieved an area under the curve of 0.867 (95% CI, 0.817– 0.918) in the training cohort and 0.804 (95% CI, 0.709– 0.899) in the validation cohort. Conclusion: Quantitative vertebral structural parameters derived from opportunistic CT, particularly cortical parameters, showed good performance for identifying osteoporotic vertebral compression fracture risk in this single-center cohort. These findings support the potential value of deep learning-assisted quantitative assessment for opportunistic screening, although external validation is still required before broader clinical implementation.
- 1 April 2026
- Published by Taylor & Francis in International Journal of General Medicine
- Vol. ume 19, 1-11
Open AccessDeep Learning-Based Assessment of the Value of Vertebral Structural Parameters in Predicting Osteoporotic Vertebral Compression Fractures on Opportunistic CT Scans
Background: Osteoporotic vertebral compression fractures are common fragility fractures in older adults and are associated with substantial disability and healthcare burden. Opportunistic CT may provide a practical no-extra-radiation pathway for fracture-risk assessment, but quantitative vertebral structural parameters, especially cortical parameters, remain insufficiently studied. Purpose: To evaluate the value of vertebral structural parameters derived from opportunistic CT in identifying osteoporotic vertebral compression fractures and to develop a nomogram for individualized risk estimation. Methods: This retrospective single-center study included 298 patients aged 45 years or older who underwent chest or abdominal CT at the Second Affiliated Hospital of Shandong First Medical University between January 2020 and May 2024. Osteoporotic vertebral compression fracture status on sagittal CT was determined by two readers (one radiology resident and one senior physician) by consensus. A high-resolution 3D Dense-U-Net was used for automated vertebral segmentation and extraction of L1 structural parameters. Group comparisons were performed with t tests, one-way analysis of variance, and chi-squared tests as appropriate. Independent predictors were identified with univariate and multivariate logistic regression, and the nomogram was evaluated with receiver operating characteristic and calibration analyses. Results: Of the 298 participants (182 men and 116 women; mean age, 62.33 ± 9.56 years), 134 had osteoporotic vertebral compression fractures and 164 did not. In multivariate analysis, L1 cortical average area (OR, 0.99; 95% CI, 0.99– 1.00; P = 0.002) and L1 cortical average thickness (OR, 0.22; 95% CI, 0.13– 0.38; P < 0.001) were independent predictors. The nomogram achieved an area under the curve of 0.867 (95% CI, 0.817– 0.918) in the training cohort and 0.804 (95% CI, 0.709– 0.899) in the validation cohort. Conclusion: Quantitative vertebral structural parameters derived from opportunistic CT, particularly cortical parameters, showed good performance for identifying osteoporotic vertebral compression fracture risk in this single-center cohort. These findings support the potential value of deep learning-assisted quantitative assessment for opportunistic screening, although external validation is still required before broader clinical implementation.
- Open Access
- 9 May 2025
- Published by Frontiers Media SA in Frontiers in Medicine
- Vol. 12, 1530462
Study on the correlation between abnormal bone metabolism and cognitive impairment in type 2 diabetes mellitus
Introduction: Type 2 diabetes mellitus (T2DM) is often accompanied by bone metabolic disorders and cognitive impairment, forming an interactive network through metabolic derangements, oxidative stress, and inflammatory responses. Hyperglycemia and insulin resistance disrupt bone remodeling leading to osteoporosis while simultaneously impairing cognition via blood-brain barrier damage and neuroinflammation. Osteogenic factors like osteocalcin may bidirectionally regulate glucose metabolism and brain function, suggesting that “bone-brain axis” dysregulation could be a potential mechanism underlying cognitive impairment in T2DM. This study aims to characterize cognitive function patterns in T2DM patients with bone metabolic abnormalities and their clinical correlations, providing a basis for multisystemic interventions.Methods: The general clinical data, osteocalcin (OC), glycosylated hemoglobin (HbA1c), bone mineral density (BMD), and the Montreal Cognitive Assessment (MoCA) scores of 50 patients with T2DM were collected. According to whether cognitive impairment occurred or not, one-way ANOVA was performed to analyze the correlation between cognitive and clinical indicators, BMD and OC. Multiple linear regression analysis was performed with cognition and bone density as dependent variables and other factors as independent variables.Results: T2DM subjects were grouped according to bone mass. The osteoporosis group had the lowest MoCA score and bone density, followed by the osteopenia group. There were 16 cases (16/17 94.12%) of cognitive impairment in the osteoporosis group, 13 cases (13/17 76.47%) of cognitive impairment in the osteopenia group, and 3 cases (3/16 18.75%) of cognitive impairment in the normal bone mass group. Compared with the normal cognitive group, the MoCA score, OC measurement and BMD of the patients in the cognitive impairment group were lower (P < 0.05). BMD (r = 0.686, P = 0.000), OC (r = 0.756, P = 0.000) are positively correlated with MoCA score. OC (r = 0.690, P = 0.000) and Age (r = −0.032, P = 0.045) are positively correlated with BMD. Multivariate linear regression analysis found that with cognition as the dependent variable, the decrease in BMD (P = 0.028) and OC (P = 0.000) aggravated the occurrence of cognitive impairment; with BMD as the dependent variable, the decline in cognition (P = 0.028) and OC (P = 0.029) aggravated the decrease in BMD.Conclusion: T2DM, osteoporosis, and cognitive impairment form pathological connections through metabolic disorders, chronic inflammation, and bidirectional regulatory networks of the “bone-brain axis,” with osteocalcin serving as a key mediator that maintains bone remodeling balance while also exerting cross-domain regulation over central insulin signaling and synaptic plasticity. Understanding these interactive mechanisms provides a basis for developing combined screening models integrating bone density and cognitive assessments, and promotes multidisciplinary collaborative interventions across endocrinology, orthopedics, and neurology to improve overall outcomes for T2DM patients.
- 9 May 2025
- Published by Frontiers Media SA in Frontiers in Medicine
- Vol. 12, 1530462
Open AccessStudy on the correlation between abnormal bone metabolism and cognitive impairment in type 2 diabetes mellitus
Introduction: Type 2 diabetes mellitus (T2DM) is often accompanied by bone metabolic disorders and cognitive impairment, forming an interactive network through metabolic derangements, oxidative stress, and inflammatory responses. Hyperglycemia and insulin resistance disrupt bone remodeling leading to osteoporosis while simultaneously impairing cognition via blood-brain barrier damage and neuroinflammation. Osteogenic factors like osteocalcin may bidirectionally regulate glucose metabolism and brain function, suggesting that “bone-brain axis” dysregulation could be a potential mechanism underlying cognitive impairment in T2DM. This study aims to characterize cognitive function patterns in T2DM patients with bone metabolic abnormalities and their clinical correlations, providing a basis for multisystemic interventions.Methods: The general clinical data, osteocalcin (OC), glycosylated hemoglobin (HbA1c), bone mineral density (BMD), and the Montreal Cognitive Assessment (MoCA) scores of 50 patients with T2DM were collected. According to whether cognitive impairment occurred or not, one-way ANOVA was performed to analyze the correlation between cognitive and clinical indicators, BMD and OC. Multiple linear regression analysis was performed with cognition and bone density as dependent variables and other factors as independent variables.Results: T2DM subjects were grouped according to bone mass. The osteoporosis group had the lowest MoCA score and bone density, followed by the osteopenia group. There were 16 cases (16/17 94.12%) of cognitive impairment in the osteoporosis group, 13 cases (13/17 76.47%) of cognitive impairment in the osteopenia group, and 3 cases (3/16 18.75%) of cognitive impairment in the normal bone mass group. Compared with the normal cognitive group, the MoCA score, OC measurement and BMD of the patients in the cognitive impairment group were lower (P < 0.05). BMD (r = 0.686, P = 0.000), OC (r = 0.756, P = 0.000) are positively correlated with MoCA score. OC (r = 0.690, P = 0.000) and Age (r = −0.032, P = 0.045) are positively correlated with BMD. Multivariate linear regression analysis found that with cognition as the dependent variable, the decrease in BMD (P = 0.028) and OC (P = 0.000) aggravated the occurrence of cognitive impairment; with BMD as the dependent variable, the decline in cognition (P = 0.028) and OC (P = 0.029) aggravated the decrease in BMD.Conclusion: T2DM, osteoporosis, and cognitive impairment form pathological connections through metabolic disorders, chronic inflammation, and bidirectional regulatory networks of the “bone-brain axis,” with osteocalcin serving as a key mediator that maintains bone remodeling balance while also exerting cross-domain regulation over central insulin signaling and synaptic plasticity. Understanding these interactive mechanisms provides a basis for developing combined screening models integrating bone density and cognitive assessments, and promotes multidisciplinary collaborative interventions across endocrinology, orthopedics, and neurology to improve overall outcomes for T2DM patients.
- Open Access
- 1 May 2025
- Published by Taylor & Francis in Clinical Interventions in Aging
- Vol. ume 20, 649-658
Resting-State Functional MRI Reveals Altered Seed-Based Connectivity in Diabetic Osteoporosis Patients
Background: Diabetic osteoporosis (DOP) can cause abnormal brain neural activity, but its mechanism is still unclear. This study aims to further explore the abnormal functional connectivity between different brain regions based on the team’s previous research. Methods: Resting-state functional magnetic resonance imaging (rs-fMRI) data were obtained from 14 participants diagnosed with type 2 diabetes mellitus (T2DM) and osteoporosis. For comparison, data from 13 T2DM patients without osteoporosis were analyzed. The seed regions for functional connectivity (FC) analysis were chosen according to brain areas previously reported to exhibit abnormal regional homogeneity (ReHo). Results: DOP patients exhibited significantly decreased BMD, T-scores, MoCA scores, and osteocalcin (OC) levels compared to controls (p< 0.05). FC analysis revealed: 1) Reduced connectivity between the left middle temporal gyrus (increased ReHo) and middle occipital gyrus; 2) Enhanced connectivity between the right angular gyrus (increased ReHo) and left Rolandic operculum; 3) Weakened the left precuneus (increased ReHo) and right superior/left middle frontal gyri. These alterations correlated with deficits in visual processing, working memory, and executive function. Conclusion: Distinct FC reorganization in DOP patients reflects synergistic effects of metabolic and skeletal pathologies on neural networks, potentially mediating cognitive decline through visual pathway disruption and prefrontal-default mode network decoupling. The findings highlight neuroimaging biomarkers for metabolic bone disease-related cognitive disorders.
- 1 May 2025
- Published by Taylor & Francis in Clinical Interventions in Aging
- Vol. ume 20, 649-658
Open AccessResting-State Functional MRI Reveals Altered Seed-Based Connectivity in Diabetic Osteoporosis Patients
Background: Diabetic osteoporosis (DOP) can cause abnormal brain neural activity, but its mechanism is still unclear. This study aims to further explore the abnormal functional connectivity between different brain regions based on the team’s previous research. Methods: Resting-state functional magnetic resonance imaging (rs-fMRI) data were obtained from 14 participants diagnosed with type 2 diabetes mellitus (T2DM) and osteoporosis. For comparison, data from 13 T2DM patients without osteoporosis were analyzed. The seed regions for functional connectivity (FC) analysis were chosen according to brain areas previously reported to exhibit abnormal regional homogeneity (ReHo). Results: DOP patients exhibited significantly decreased BMD, T-scores, MoCA scores, and osteocalcin (OC) levels compared to controls (p< 0.05). FC analysis revealed: 1) Reduced connectivity between the left middle temporal gyrus (increased ReHo) and middle occipital gyrus; 2) Enhanced connectivity between the right angular gyrus (increased ReHo) and left Rolandic operculum; 3) Weakened the left precuneus (increased ReHo) and right superior/left middle frontal gyri. These alterations correlated with deficits in visual processing, working memory, and executive function. Conclusion: Distinct FC reorganization in DOP patients reflects synergistic effects of metabolic and skeletal pathologies on neural networks, potentially mediating cognitive decline through visual pathway disruption and prefrontal-default mode network decoupling. The findings highlight neuroimaging biomarkers for metabolic bone disease-related cognitive disorders.
- Open Access
- 1 April 2025
- Published by Institute of Advanced Engineering and Science in International Journal of Electrical and Computer Engineering (IJECE)
- Vol. 15 (2) , 1949-1960
Revisiting self-supervised contrastive learning for imbalanced classification
Class imbalance remains a formidable challenge in machine learning, particularly affecting fields that depend on accurate classification across skewed datasets, such as medical imaging and software defect prediction. Traditional approaches often fail to adequately address the underrepresentation of minority classes, leading to models that exhibit high performance on majority classes but have poor performance on critical minority classes. Self-supervised contrastive learning has become an extremely encouraging method for this issue, enabling the utilization of unlabeled data to generate robust and generalizable models. This paper reviews the advancements in self-supervised contrastive learning for imbalanced classification, focusing on methodologies that enhance model performance through innovative contrastive loss functions and data augmentation strategies. By pulling similar instances closer and pushing dissimilar ones apart, these techniques help mitigate the biases inherent in imbalanced datasets. We critically analyze the effectiveness of these methods in diverse scenarios and propose future research directions aimed at refining these approaches for broader application in real-world settings. This review serves as a guide for researchers exploring the potential of contrastive learning to address class imbalances, highlighting recent successes and identifying crucial gaps that need addressing.
- 1 April 2025
- Published by Institute of Advanced Engineering and Science in International Journal of Electrical and Computer Engineering (IJECE)
- Vol. 15 (2) , 1949-1960
Open AccessRevisiting self-supervised contrastive learning for imbalanced classification
Class imbalance remains a formidable challenge in machine learning, particularly affecting fields that depend on accurate classification across skewed datasets, such as medical imaging and software defect prediction. Traditional approaches often fail to adequately address the underrepresentation of minority classes, leading to models that exhibit high performance on majority classes but have poor performance on critical minority classes. Self-supervised contrastive learning has become an extremely encouraging method for this issue, enabling the utilization of unlabeled data to generate robust and generalizable models. This paper reviews the advancements in self-supervised contrastive learning for imbalanced classification, focusing on methodologies that enhance model performance through innovative contrastive loss functions and data augmentation strategies. By pulling similar instances closer and pushing dissimilar ones apart, these techniques help mitigate the biases inherent in imbalanced datasets. We critically analyze the effectiveness of these methods in diverse scenarios and propose future research directions aimed at refining these approaches for broader application in real-world settings. This review serves as a guide for researchers exploring the potential of contrastive learning to address class imbalances, highlighting recent successes and identifying crucial gaps that need addressing.
- Open Access
- 28 November 2024
- Published by MDPI AG in Applied Sciences
- Vol. 14 (23) , 11093
CL-SR: Boosting Imbalanced Image Classification with Contrastive Learning and Synthetic Minority Oversampling Technique Based on Rough Set Theory Integration
Image recognition models often struggle with class imbalance, which can impede their performance. To overcome this issue, researchers have extensively used resampling methods, traditionally focused on tabular datasets. In contrast to the original method, which generates data at the data level, this paper introduces a novel strategy that combines contrastive learning with the Synthetic Minority Oversampling Technique based on Rough Set Theory (SMOTE-RSB) specifically tailored for imbalanced image datasets. Our method leverages contrastive learning to refine representation learning and balance features, thus effectively mitigating the challenges of imbalanced image classification. We begin by extracting features using a pre-trained contrastive learning encoder. Subsequently, SMOTE-RSB is applied to these features to augment underrepresented classes and reduce irrelevant features. We evaluated our approach on several modified benchmark datasets, including CIFAR-10, SVHN, and ImageNet-LT, achieving notable improvements: an F1 score of 72.43% and a Gmean of 82.53% on the CIFAR-10 long-tailed dataset, F1 scores up to 79.57% and Gmean of 88.20% on various SVHN datasets, and a Top-1 accuracy of 68.67% on ImageNet-LT. Both qualitative and quantitative results confirm the effectiveness of our method in managing imbalances in image datasets. Additional ablation studies exploring various contrastive learning models and oversampling techniques highlight the flexibility and efficiency of our approach across different settings, underscoring its significant potential for enhancing imbalanced image classification.
- 28 November 2024
- Published by MDPI AG in Applied Sciences
- Vol. 14 (23) , 11093
Open AccessCL-SR: Boosting Imbalanced Image Classification with Contrastive Learning and Synthetic Minority Oversampling Technique Based on Rough Set Theory Integration
Image recognition models often struggle with class imbalance, which can impede their performance. To overcome this issue, researchers have extensively used resampling methods, traditionally focused on tabular datasets. In contrast to the original method, which generates data at the data level, this paper introduces a novel strategy that combines contrastive learning with the Synthetic Minority Oversampling Technique based on Rough Set Theory (SMOTE-RSB) specifically tailored for imbalanced image datasets. Our method leverages contrastive learning to refine representation learning and balance features, thus effectively mitigating the challenges of imbalanced image classification. We begin by extracting features using a pre-trained contrastive learning encoder. Subsequently, SMOTE-RSB is applied to these features to augment underrepresented classes and reduce irrelevant features. We evaluated our approach on several modified benchmark datasets, including CIFAR-10, SVHN, and ImageNet-LT, achieving notable improvements: an F1 score of 72.43% and a Gmean of 82.53% on the CIFAR-10 long-tailed dataset, F1 scores up to 79.57% and Gmean of 88.20% on various SVHN datasets, and a Top-1 accuracy of 68.67% on ImageNet-LT. Both qualitative and quantitative results confirm the effectiveness of our method in managing imbalances in image datasets. Additional ablation studies exploring various contrastive learning models and oversampling techniques highlight the flexibility and efficiency of our approach across different settings, underscoring its significant potential for enhancing imbalanced image classification.
- Open Access
- 1 August 2023
- Published by Institute of Advanced Engineering and Science in Bulletin of Electrical Engineering and Informatics
- Vol. 12 (4) , 2269-2281
SiulMalaya: an annotated bird audio dataset of Malaysia lowland forest birds for passive acoustic monitoring
The laborious point count method of conducting bird surveys is still a common practice in Malaysia. An alternative method known as passive acoustic monitoring (PAM) is deployed in many countries by placing sound recorders at surveying sites to collect bird sounds. Studies revealed that the number of bird densities counted by human observers was agreeable with those obtained using PAM. However, one of the most essential gaps in conducting PAM is the lack of expert-verified bird-call databases. Therefore, the aim of this study is to construct the first annotated Malaysia lowland forest bird sounds called SiulMalaya to be used as ground-truth datasets for PAM-related activities. The raw bird sounds dataset was downloaded from Macaulay Library using the eBird platform. Data pre-processing was done to produce annotated audio tracks that can be used as training datasets for bird classification. SiulMalaya dataset was further validated by two bird experts from the Department of Wildlife and National Parks, Malaysia. A bird identification experiment was carried out to assess and validate SiulMalaya dataset using a convolutional neural network (CNN) learning model. Even though the accuracy of bird identification is slightly above 50%, the annotated dataset is shown to be viable for PAM-related operations.
- 1 August 2023
- Published by Institute of Advanced Engineering and Science in Bulletin of Electrical Engineering and Informatics
- Vol. 12 (4) , 2269-2281
Open AccessSiulMalaya: an annotated bird audio dataset of Malaysia lowland forest birds for passive acoustic monitoring
The laborious point count method of conducting bird surveys is still a common practice in Malaysia. An alternative method known as passive acoustic monitoring (PAM) is deployed in many countries by placing sound recorders at surveying sites to collect bird sounds. Studies revealed that the number of bird densities counted by human observers was agreeable with those obtained using PAM. However, one of the most essential gaps in conducting PAM is the lack of expert-verified bird-call databases. Therefore, the aim of this study is to construct the first annotated Malaysia lowland forest bird sounds called SiulMalaya to be used as ground-truth datasets for PAM-related activities. The raw bird sounds dataset was downloaded from Macaulay Library using the eBird platform. Data pre-processing was done to produce annotated audio tracks that can be used as training datasets for bird classification. SiulMalaya dataset was further validated by two bird experts from the Department of Wildlife and National Parks, Malaysia. A bird identification experiment was carried out to assess and validate SiulMalaya dataset using a convolutional neural network (CNN) learning model. Even though the accuracy of bird identification is slightly above 50%, the annotated dataset is shown to be viable for PAM-related operations.
- Open Access
- 1 August 2023
- Published by Institute of Advanced Engineering and Science in Bulletin of Electrical Engineering and Informatics
- Vol. 12 (4) , 2269-2281
SiulMalaya: an annotated bird audio dataset of Malaysia lowland forest birds for passive acoustic monitoring
The laborious point count method of conducting bird surveys is still a common practice in Malaysia. An alternative method known as passive acoustic monitoring (PAM) is deployed in many countries by placing sound recorders at surveying sites to collect bird sounds. Studies revealed that the number of bird densities counted by human observers was agreeable with those obtained using PAM. However, one of the most essential gaps in conducting PAM is the lack of expert-verified bird-call databases. Therefore, the aim of this study is to construct the first annotated Malaysia lowland forest bird sounds called SiulMalaya to be used as ground-truth datasets for PAM-related activities. The raw bird sounds dataset was downloaded from Macaulay Library using the eBird platform. Data pre-processing was done to produce annotated audio tracks that can be used as training datasets for bird classification. SiulMalaya dataset was further validated by two bird experts from the Department of Wildlife and National Parks, Malaysia. A bird identification experiment was carried out to assess and validate SiulMalaya dataset using a convolutional neural network (CNN) learning model. Even though the accuracy of bird identification is slightly above 50%, the annotated dataset is shown to be viable for PAM-related operations.
- 1 August 2023
- Published by Institute of Advanced Engineering and Science in Bulletin of Electrical Engineering and Informatics
- Vol. 12 (4) , 2269-2281
Open AccessSiulMalaya: an annotated bird audio dataset of Malaysia lowland forest birds for passive acoustic monitoring
The laborious point count method of conducting bird surveys is still a common practice in Malaysia. An alternative method known as passive acoustic monitoring (PAM) is deployed in many countries by placing sound recorders at surveying sites to collect bird sounds. Studies revealed that the number of bird densities counted by human observers was agreeable with those obtained using PAM. However, one of the most essential gaps in conducting PAM is the lack of expert-verified bird-call databases. Therefore, the aim of this study is to construct the first annotated Malaysia lowland forest bird sounds called SiulMalaya to be used as ground-truth datasets for PAM-related activities. The raw bird sounds dataset was downloaded from Macaulay Library using the eBird platform. Data pre-processing was done to produce annotated audio tracks that can be used as training datasets for bird classification. SiulMalaya dataset was further validated by two bird experts from the Department of Wildlife and National Parks, Malaysia. A bird identification experiment was carried out to assess and validate SiulMalaya dataset using a convolutional neural network (CNN) learning model. Even though the accuracy of bird identification is slightly above 50%, the annotated dataset is shown to be viable for PAM-related operations.