How Many Private Data Are Needed for Deep Learning in Lung Nodule Detection on CT Scans? A Retrospective Multicenter Study

Abstract

Early detection of lung nodules is essential for preventing lung cancer. However, the number of radiologists who can diagnose lung nodules is limited, and considerable effort and time are required. To address this problem, researchers are investigating the automation of deep-learning-based lung nodule detection. However, deep learning requires large amounts of data, which can be difficult to collect. Therefore, data collection should be optimized to facilitate experiments at the beginning of lung nodule detection studies. We collected chest computed tomography scans from 515 patients with lung nodules from three hospitals and high-quality lung nodule annotations reviewed by radiologists. We conducted several experiments using the collected datasets and publicly available data from LUNA16. The object detection model, YOLOX was used in the lung nodule detection experiment. Similar or better performance was obtained when training the model with the collected data rather than LUNA16 with large amounts of data. We also show that weight transfer learning from pre-trained open data is very useful when it is difficult to collect large amounts of data. Good performance can otherwise be expected when reaching more than 100 patients. This study offers valuable insights for guiding data collection in lung nodules studies in the future.

Multiround Transfer Learning and Modified Generative Adversarial Network for Lung Cancer Detection

“…Related works [13][14][15][16] only reported the results after applying the transfer learning model. Te percentage improvement of the accuracy is 7.80% [9], 3.77% [10], 1.34% [11], 5.88% [12], and 6.85-9.92% (our work).…”

“…(i) Source domain and target domain: the related works [9][10][11][12] formulated the transfer learning problem using a similar source domain and target domain whereas other works [13][14][15][16] considered the distant source and target domains. Our work considered 10 benchmark datasets to evaluate the MTL using similar and distant sources and target domains.…”

“…Related Works. Although existing works [9][10][11][12][13][14][15][16] formulated the transfer learning problems with a single source domain and single target domain, the discussion has merit as these works fell into the same research area, i.e., transfer learning for LCD. In the following, two common types of formulations will be discussed: (i) transfer learning between the similar source domain and target domain [9][10][11][12] and (ii) transfer learning between the distant source domain and target domain [13][14][15][16].…”

“…Transfer learning enhanced the accuracy of the model from 87.5% to 90.8%. To transfer the knowledge from LUNA16 to the target domain of the Gangneung Asan Hospital for LCD, a YOLOX algorithm was used [11]. Results showed a slight enhancement of the model's accuracy from 89.7% to 90.9%.…”

A robust model training strategy using hard negative mining in a weakly labeled dataset for lymphatic invasion in gastric cancer

“…Hard negative mining is a technique employed in semisupervised learning to improve the performance of deep learning models in object detection and classification tasks, as demonstrated in numerous studies [8][9][10][11]14,15]. This approach involves the identification and selection of challenging negative samples incorrectly classified by the current model; thereafter, these samples are employed to train the model, thereby yielding a robust model.…”