Arnaud Arindra Adiyoso Setio

A deep learning risk assessment model for pulmonary nodules that outperforms established clinical risk scores (Lung-RADS, Mayo Clinic) in a multi-center lung cancer screening cohort.

Systematic review and meta-analysis evaluating the diagnostic accuracy of AI models for classifying infectious keratitis.

A large-scale evaluation comparing deep learning systems against 11 radiologists for lung cancer detection on screening CT, showing competitive AI performance.

A deep learning approach that aggregates aortic centerline features to efficiently predict hemodynamic parameters, replacing costly CFD simulations.

A deep learning model that outperforms radiologists and existing risk models in estimating malignancy risk of pulmonary nodules at CT screening.

Integrates prior medical knowledge with noisy label learning to robustly classify abnormalities in chest X-rays.

Creates a large synthetic dataset of aortic morphology and hemodynamics to address data scarcity in cardiovascular AI research.

Efficient multi-label CNN approach for fast and accurate localization of multiple organs in thorax-abdomen CT.

A comprehensive survey covering deep learning applications across medical image analysis tasks including classification, detection, and segmentation.

The LUNA16 grand challenge: benchmarking state-of-the-art algorithms for pulmonary nodule detection across 888 CT scans.

Uses convolutional networks to detect and prevent leakage in automated airway segmentation from chest CT.

Applies deep learning to automatically segment breast and fibroglandular tissue in MRI for density assessment.

A deep learning system for automated pulmonary nodule management following lung cancer screening guidelines.

A multi-view CNN approach that significantly reduces false positives in automated pulmonary nodule detection.

An automated method for detecting large solid pulmonary nodules in thoracic CT using multi-scale Hessian filtering and a classification pipeline.

Explores customization strategies for radiology report generation models to accommodate different clinical workflows and institutional requirements.

Proposes an entity-level evaluation framework for chest X-ray report generation to better assess clinical safety of AI-generated radiology reports.

Uses lung inpainting to extract and transfer nodule features for data augmentation in nodule detection models.

A class-aware GAN for synthesizing realistic lung nodules in CT to augment training data for detection and classification.

Multi-label CNNs for simultaneous detection and localization of multiple organs in thorax-abdomen CT.

Applies deep CNNs to automatically score coronary artery calcification in low-dose chest CT screening.

Demonstrates that pre-trained CNN features can be effectively repurposed for pulmonary nodule detection in CT.

Combines multiple nodule detection systems with a tumor risk model for improved lung cancer CAD.

A memory-centric hardware accelerator design for efficient CNN inference.

Evaluates textural feature representations for early-stage cancer detection in endoscopy images.

An optimized algorithm and VLSI architecture for LTE turbo encoding.

PhD thesis on deep learning methods for automated detection and characterization of pulmonary nodules in CT for lung cancer screening.