Latika Giri

Dr. Latika Giri


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 June 2025 in Journal of Cardiovascular Imaging
Verified
Journal Article
Open access
PDF file
LGLatika Giri
Latika Giri
RSRoshan Singh
Roshan Singh
AMAhmed Marey
Ahmed Marey
YLYupeng Li
Yupeng Li
BVBharath Ambale Venkatesh
Bharath Ambale Venkatesh
MUMuhammad Umair
Muhammad Umair
Background: Cardiovascular magnetic resonance mapping parameters—native T1 mapping, T2 mapping, and extracellular volume (ECV)—are key for diagnosing acute myocarditis under the modified 2018 Lake Louise Criteria (mLLC). This systematic review and meta-analysis evaluated their diagnostic performance and established optimal thresholds for acute myocarditis. Methods: We reviewed articles published in the past decade utilizing parametric mapping for myocarditis diagnosis. Data on sensitivity, specificity, and area under the curve (AUC) were extracted. Quality assessment was conducted using the QUADAS-2 tool by two independent reviewers. Results: Eleven studies with 677 patients were included. Native T1 mapping showed sensitivity of 83%, specificity of 86%, diagnostic odds ratio (DOR) of 39, and an AUC of 0.91. T2 mapping had sensitivity of 81%, specificity of 86%, DOR of 25, and an AUC of 0.89. ECV demonstrated sensitivity of 71%, specificity of 81%, DOR of 13, and an AUC of 0.83. Mean control values were 1,039 ± 39.23 ms for native T1 mapping, 57 ± 5.18 ms for T2 mapping, and 31% ± 5.60% for ECV. Optimal thresholds were 1,021 ms for native T1 mapping, 52 ms for T2 mapping, and 28% for ECV based on receiver operating characteristic curves analysis based on 1.5-T scanner value. Native T1 mapping showed the highest diagnostic accuracy. Subgroup analysis found no significant sensitivity differences based on biopsy or clinical criteria. Conclusions: Parametric mapping, particularly native T1, demonstrated strong diagnostic performance for acute myocarditis compared to T2 mapping and ECV within the modified 2018 Lake Louise Criteria framework. Incorporating these cardiovascular magnetic resonance parameters may improve diagnostic accuracy. Further research is recommended to refine these findings and optimize diagnostic strategies.

24 December 2025 in Diagnostics
Verified
Journal Article
Open access
PDF file
LGLatika Giri
Latika Giri
PRPradeep Raj Regmi
Pradeep Raj Regmi
GGGhanshyam Gurung
Ghanshyam Gurung
GGGrusha Gurung
Grusha Gurung
SAShova Aryal
Shova Aryal
MUMuhammad Umair
Muhammad Umair
Background: Chest radiography is the most widely used diagnostic imaging modality globally, yet its interpretation is hindered by a critical shortage of radiologists, especially in low- and middle-income countries (LMICs). The interpretation is both time-consuming and error-prone in high-volume settings. Artificial Intelligence (AI) systems trained on public data may lack generalizability to multi-view, real-world, local images. Deep learning tools have the potential to augment radiologists by providing real-time decision support by overcoming these. Objective: We evaluated the diagnostic accuracy of a deep learning-based convolutional neural network (CNN) trained on multi-view, hybrid (public and local datasets) for detecting thoracic abnormalities in chest radiographs of adults presenting to a tertiary hospital, operating in offline mode. Methodology: A CNN was pretrained on public datasets (Vin Big, NIH) and fine-tuned on a local dataset from a Nepalese tertiary hospital, comprising frontal (PA/AP) and lateral views from emergency, ICU, and outpatient settings. The dataset was annotated by three radiologists for 14 pathologies. Data augmentation simulated poor-quality images and artifacts. Performance was evaluated on a held-out test set (N = 522) against radiologists’ consensus, measuring AUC, sensitivity, specificity, mean average precision (mAP), and reporting time. Deployment feasibility was tested via PACS integration and standalone offline mode. Results: The CNN achieved an overall AUC of 0.86 across 14 abnormalities, with 68% sensitivity, 99% specificity, and 0.93 mAP. Colored bounding boxes improved clarity when multiple pathologies co-occurred (e.g., cardiomegaly with effusion). The system performed effectively on PA, AP, and lateral views, including poor-quality ER/ICU images. Deployment testing confirmed seamless PACS integration and offline functionality. Conclusions: The CNN trained on adult CXRs performed reliably in detecting key thoracic findings across varied clinical settings. Its robustness to image quality, integration of multiple views and visualization capabilities suggest it could serve as a useful aid for triage and diagnosis.

8 January 2026 in Frontiers in Radiology
Verified
Journal Article
Open access
PDF file
EKEhsan Karimialavijeh
Ehsan Karimialavijeh
LGLatika Giri
Latika Giri
EBEduardo Baettig
Eduardo Baettig
MUMuhammad Umair
Muhammad Umair
Acute myocarditis is an inflammatory condition of the myocardium, often triggered by viral infections, autoimmune diseases, or toxins. It can lead to arrhythmias, heart failure, and sudden cardiac death. Early and accurate diagnosis is crucial for timely management and preventing complications. It poses a significant diagnostic challenge in emergency departments (EDs) due to nonspecific symptoms, overlapping features with conditions like acute coronary syndrome, and limitations of conventional diagnostics. Cardiac magnetic resonance imaging (CMR) is the gold standard for noninvasive diagnosis, using the 2018 Modified Lake Louise Criteria (mLLC). However, high-field CMR (1.5–3T) faces barriers in EDs, such as longer scan times, higher cost, lack of accessibility, and contraindications in patients with implantable devices, severe kidney disease, or hemodynamic instability. Low-field MRI (<1.5T) offers advantages in portability, safety, and cost while reducing susceptibility artifacts. Recent advances in AI-driven image reconstruction (e.g., LoHiResGAN, U-net) address low signal-to-noise ratios, enabling cine imaging, strain analysis, and parametric mapping at 0.55T. Studies show that low-field CMR can detect subclinical myocarditis and predict outcomes, with ECV measurements at 0.55T strongly correlating with 1.5T ( r = 0.91), demonstrating comparable reliability. By integrating low-field CMR into ED protocols, clinicians can improve early detection of occult myocarditis, guide risk stratification, and reduce long-term morbidity and healthcare costs. Standardization of imaging workflows and AI-enhanced protocols will further bridge diagnostic gaps, particularly in resource-limited settings. This review highlights low-field CMR's potential to redefine acute myocarditis management, balancing diagnostic precision with practicality in emergency care.

3 November 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
LGLatika Giri
Latika Giri
PRPradeep Raj Regmi
Pradeep Raj Regmi
GGGhanshyam Gurung
Ghanshyam Gurung
GGGrusha Gurung
Grusha Gurung
SAShova Aryal
Shova Aryal
MUMuhammad Umair
Muhammad Umair
Background: Chest radiography is the most widely used diagnostic imaging modality. Yet its interpretation is challenged by limited radiology workforce, especially in low- and middle-income countries (LMICs). The interpretation is both time consuming and error-prone in high volume settings. Artificial Intelligence (AI) systems trained on public data may lack generalizability to multi-view, real-world, local images. Deep learning tools have the potential to augment radiologists by providing real-time decision support by overcoming these. Objective: We evaluated the diagnostic accuracy of a deep learning-based convolutional neural network (CNN) trained on multi-view, hybrid (public and local datasets) for detecting thoracic abnormalities in chest radiographs of adults presenting to a tertiary hospital. Methods: A CNN was pretrained on large public datasets (VinBig, NIH) and fine-tuned on adult chest radiographs both frontal [posteroanterior (PA) and anteroposterior (AP)] from Tribhuvan University Teaching Hospital TUTH, Nepal. The dataset included emergency (ER), ICU, and outpatient (OPD) radiographs. Data augmentation simulated poor-quality images and artifacts (ECG wires, text labels, rotation, low exposure). Fourteen thoracic pathologies were annotated by 3 radiologists. Bounding boxes were refined using Weighted Boxes Fusion (WBF) and displayed in 14 unique colors for interpretability. The system was evaluated on a held-out test set (N=522) against radiologist consensus. Primary outcomes included AUC, sensitivity, specificity, mean average precision (mAP), and reporting time. Deployment feasibility was tested on Picture Archiving and Communication System (PACS) and in offline standalone mode. Results: The CNN achieved an overall AUC of 0.86 across 14 abnormalities, with 68% sensitivity,99% specificity, and 0.93 mAP. Colored bounding boxes improved clarity when multiple pathologies co-occurred (e.g., cardiomegaly with effusion). The system performed effectively on PA, AP, and lateral views, including poor-quality ER/ICU images. Deployment testing confirmed seamless PACS integration and offline functionality. Conclusion: The CNN trained on adult CXRs performed reliably in detecting key thoracic findings across varied clinical settings. Its robustness to image quality, integration of multiple views and visualization capabilities suggest it could serve as a useful aid for triage and diagnosis.

16 November 2022
Verified
Preprint
Open access
PDF file
LGLatika Giri
Latika Giri
SASuyash Acharya
Suyash Acharya
NBNiroj Bhandari
Niroj Bhandari
GSGrishma Sharma
Grishma Sharma
PGPrabesh Gautam
Prabesh Gautam
Leprosy affects multiple systems causing morbidity.1 However with attainment of global elimination and prevalence being 0.89/10000 at national level; it is seldom considered a diagnosis when it doesn’t manifest telltale signs. 2,3 Leprosy can masquerade various dermatological, orthopedic or neurological problems thus delaying the diagnosis.

2 November 2026 in Clinical Case Reports
Verified
Journal Article
Open access
IKIswor Karki
Iswor Karki
SJSagar Kumar Jha
Sagar Kumar Jha
SPShankar Poudel
Shankar Poudel
STSabitra Thapa
Sabitra Thapa
LGLatika Giri
Latika Giri
KAK. C. Ajay
K. C. Ajay
A 5‐year‐old girl with prior patent ductus arteriosus device closure presented with abdominal pain, vomiting, and fever. Despite a normal leukocyte count and nonvisualization of the appendix on ultrasonography, computed tomography identified perforation. Emergency appendectomy confirmed gangrenous appendicitis. This case underscores the importance of serial examination and timely diagnostic imaging.