Papers › Lung nodule detection and classification from Thorax CT-scan using RetinaNet with...

Lung nodule detection and classification from Thorax CT-scan using RetinaNet with transfer learning

8 Apr 2020Journal of King Saud University - Computer and Information Sciences 2020 4archive 2025-07-28

Ivan William Harsono, Suryadiputra Liawatimena, Tjeng Wawan Cenggoro

Lung malignancy is one of the most common causes of death in the world caused by malignant lung nodules which commonly diagnosed radiologically by radiologists. Unfortunately, the continuous flow of medical images in hospitals drives radiologists to prioritize quantity over quality. This work condition allows misinterpretation especially on ambiguous anatomical structures that resemble lung nodule for example enlarged lymph nodes and resulting in decreasing sensitivity and accuracy of malignant lung nodule detections and late diagnosis proven to be fatal to patients. To address the problem, this paper proposed a novel lung nodule detection and classification model using one stage detector called as “I3DR-Net.” The model was formed by combining pre-trained natural images weight of Inflated 3D ConvNet (I3D) backbone with feature pyramid network to multi-scale 3D Thorax Computed tomography scan (CT-scan) dataset. I3DR-Net able to produce remarkable results on lung nodule texture detection task with mAP 49.61% and 22.86%, and area under curve (AUC) 81.84% and 70.36% for public and private dataset. Additionally, I3DR-Net successfully outperform previous state-of-the-art Retina U-Net and U-FRCNN + mean average precision (mAP) by 7.9% and 7.2% (57.71% VS 49.8% VS 50.5%) for malignant nodule detection and classification task.

PaperPDFCode

Syntology This paper has no arXiv id in the archive, and this site joins papers to Syntology's graph by arXiv id, so this page does not say whether the graph holds it.

Code

ivanwilliammd/I3DR-Net-Transfer-Learning officialpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset (arXiv:1705.07750; that paper's own run record: 16 ran (of which 7 constructed an object rather than computing a result; 10 with no instrument failure: 0 honoured, 0 violated, 10 with no contract checked; 6 where Syntology's instrument failed) · 12 unverified); nothing checks that they implement this paper's method. report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Lung Nodule 3D ClassificationLung Nodule 3D DetectionLung Nodule ClassificationLung Nodule DetectionTransfer Learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Lung Nodule 3D Classification LIDC-IDRI I3DR-Net AUC 81.84 #1 of 1 Archive leaderboard report
Lung Nodule 3D Detection LIDC-IDRI I3DR-Net AUC 81.84 #1 of 1 Archive leaderboard report
Lung Nodule 3D Detection LIDC-IDRI I3DR-Net Sensitivity 94.12 #1 of 1 Archive leaderboard report
Lung Nodule 3D Detection LIDC-IDRI I3DR-Net mean average precision 49.61 #1 of 1 Archive leaderboard report
Lung Nodule Classification LIDC-IDRI I3DR-Net AUC 81.84 #8 of 8 Archive leaderboard report
Lung Nodule Detection LIDC-IDRI I3DR-Net AUC 81.84 #1 of 1 Archive leaderboard report
Lung Nodule Detection LIDC-IDRI I3DR-Net Sensitivity 94.12 #1 of 1 Archive leaderboard report
Lung Nodule Detection LIDC-IDRI I3DR-Net mean average precision 49.61 #1 of 1 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

1x1 ConvolutionConcatenated Skip ConnectionConvolutionFPNFocal LossInception ModuleMax PoolingReLURetinaNetU-Net

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections