Claim Missing Document
Check
Articles

Found 3 Documents
Search

Deep Learning-Based Image Enhancement of Low-Field Brain MRI: A Multicenter Validation of Diagnostic Image Quality and Lesion Detection Rachmat Hidayat; Linda Purnama
Sriwijaya Journal of Radiology and Imaging Research Vol. 3 No. 2 (2025): Sriwijaya Journal of Radiology and Imaging Research
Publisher : Phlox Institute: Indonesian Medical Research Organization

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59345/sjrir.v3i2.288

Abstract

Introduction: Low-field MRI (LF-MRI) widens neuroimaging access in resource-limited settings but suffers low signal-to-noise ratio (SNR), reduced resolution and artefacts. We developed and validated a deep-learning framework for image normalisation and noise reduction to elevate 0.35T brain MRI toward high-field quality, and tested whether it improves clinically significant lesion detection. Methods: In a multicentre retrospective diagnostic-accuracy study (STARD 2015), 450 adults underwent non-contrast 0.35T brain MRI (T1W, T2W, FLAIR) across three private tertiary centres in Palembang, Indonesia. Images were enhanced with a CycleGAN incorporating Vision-Transformer blocks. Three blinded neuroradiologists scored a 5-point Likert scale and recorded lesion presence; paired 1.5T MRI was the reference standard. Sensitivity, specificity, AUC and likelihood ratios were computed with 95% CIs; tests compared by McNemar and DeLong; agreement by Fleiss kappa. Results: AI enhancement improved all quality metrics (e.g., T1W PSNR 22.15 to 28.45 dB; SSIM 0.71 to 0.89; all p<0.001). For lesion detection, AI-enhanced LF-MRI achieved sensitivity 93.9% (95% CI 89.4–96.6), specificity 91.1% (87.1–94.0) and AUC 0.94 (0.91–0.97) versus 78.3%, 81.1% and 0.81 for original images (DeLong p<0.001; McNemar p<0.001). LR+ rose to 10.56 and LR− fell to 0.067. Inter-reader agreement was almost perfect (Fleiss kappa 0.78–0.85). Conclusion: A CycleGAN-with-transformer framework substantially improved objective quality and diagnostic performance of 0.35T brain MRI toward high-field standards with almost-perfect reader agreement. Pending prospective and external validation, AI enhancement is a low-cost route to more equitable neuroimaging.
Non-Invasive Neuromodulation for Drug-Resistant Epilepsy in Children: A Randomized Controlled Trial of Transcranial Magnetic Stimulation (TMS) versus Vagus Nerve Stimulation (VNS) in Jakarta, Indonesia Febria Suryani; Rinna Azrida; Linda Purnama; Vania Delma; Desiree Montesinos
Scientific Journal of Pediatrics Vol. 3 No. 1 (2025): Scientific Journal of Pediatrics
Publisher : Phlox Institute: Indonesian Medical Research Organization

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59345/sjped.v2i2.173

Abstract

Introduction: Drug-resistant epilepsy (DRE) significantly impacts the quality of life in children. While vagus nerve stimulation (VNS) is an established treatment, repetitive transcranial magnetic stimulation (rTMS) offers a non-invasive alternative. This study aimed to compare the efficacy and safety of rTMS versus VNS in a pediatric DRE population in Jakarta, Indonesia. Methods: This was a single-center, randomized, controlled, open-label trial conducted at Private Hospital, Jakarta. Children aged 5-18 years with DRE, defined as failure to achieve seizure freedom despite adequate trials of two appropriate antiepileptic drugs (AEDs), were randomly assigned (1:1) to receive either rTMS or VNS. The primary outcome was the percentage reduction in seizure frequency at 6 months post-intervention compared to baseline. Secondary outcomes included responder rate (≥50% seizure reduction), quality of life (QoL) using the PedsQL, cognitive function (using standardized neuropsychological tests), and adverse events. Results: A total of 60 children were randomized (30 rTMS, 30 VNS). At 6 months, the mean percentage reduction in seizure frequency was significantly greater in the rTMS group (48.5%, SD 15.2%) compared to the VNS group (35.2%, SD 12.8%) (p = 0.001). Responder rates were 63.3% for rTMS and 46.7% for VNS (p = 0.17). PedsQL scores showed a significant improvement in the rTMS group compared to baseline in the psychosocial health summary score (p = 0.005), but not the VNS group (p=0.1). No significant differences were observed in cognitive function between the groups. Adverse events were generally mild and transient in both groups, though VNS was associated with more voice alteration and coughing. Conclusion: rTMS demonstrated superior efficacy in reducing seizure frequency compared to VNS in this Indonesian pediatric DRE population. While VNS is an established method, rTMS may present a non-invasive and potentially more effective therapeutic alternative. Further, larger, multicenter studies are warranted to confirm these findings and explore long-term outcomes.
Spatiotemporal Trends and Attributable Risk Factors of Nasopharyngeal Carcinoma in Southeast Asia (1990–2023): An Ecological Analysis of the Global Burden of Disease Study with Projections to 2050 Linda Purnama; Priscilla Kapoor; Mustafa Mahmud; Alessandra Navos
Sriwijaya Journal of Otorhinolaryngology Vol. 3 No. 2 (2025): Sriwijaya Journal of Otorhinolaryngology
Publisher : Phlox Institute: Indonesian Medical Research Organization

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59345/sjorl.v3i2.306

Abstract

Background: Nasopharyngeal carcinoma (NPC) is an Epstein–Barr virus-associated epithelial malignancy with a strikingly uneven geography concentrated in Southeast Asia and southern China. Its region-specific burden, temporal dynamics and attributable risk factors have not been synthesised in a dedicated spatiotemporal ecological analysis. Objective: To characterise NPC burden, trends and attributable risks across Southeast Asia. Methods: We conducted an ecological analysis of Global Burden of Disease (GBD) estimates. Age-standardised incidence (ASIR), mortality (ASMR) and disability-adjusted life-year (DALY) rates for the GBD Southeast Asia region (1990–2021, GBD 2021) and constituent ASEAN countries (to 2023, GBD 2023) were compiled; temporal trends were summarised by the estimated annual percentage change (EAPC); attributable fractions for tobacco, alcohol and occupational carcinogens were obtained from GBD comparative risk assessment; and age-standardised rates were projected toward 2050. Results: In 2021 the Southeast Asia region recorded the highest NPC mortality of any world region (ASMR 1.60 per 100,000). Age-standardised rates declined (EAPC: incidence −0.39, mortality −0.76, DALYs −0.84% per year), yet absolute incident cases more than doubled (6,209 to 13,661; +120%). National ASIR ranged from 1.3 (Cambodia) to 14.2 (Singapore); Indonesia carried the greatest absolute burden (5,163 cases; 4,306 deaths). Incidence rose in Indonesia (EAPC +0.8), Laos (+0.7) and Viet Nam (+0.6). Alcohol-attributable burden increased most steeply (Southeast Asia EAPC +2.27), against declining tobacco- and occupational-attributable rates. Conclusion: Southeast Asia bears the world's heaviest NPC mortality; falling age-standardised rates mask a rising absolute burden and a growing alcohol-attributable component, underscoring the need for Epstein–Barr virus-informed early detection and region-tailored prevention.