Table 1.
Cluster | Number of Papers | Central Paper/s | Imaging Mode | Connecting Theme | Data Source |
---|---|---|---|---|---|
1 | 36 | [40,41] | CXR | Connected by the use of JSRT [40] and PLCO [41] datasets as a lung nodule dataset, along with a deep learning approach for image analysis and nodule classification. | JSRT PLCO |
2 | 29 | [42,43] | CT/CXR | Connected using local feature analysis, linear filtering, clustering techniques, and other non-deep learning techniques. | LIDC |
3 | 24 | [44] | CXR | Artificial intelligence and machine learning methods, including ANN, SVM, and KNN. Typically used the JSRT database. | JSRT |
4 | 23 | [45] | CXR | Rib/bone suppression and image enhancement techniques, including wavelet transform methods. | JSRT |
5 | 17 | [46,47] | CT | Use of deep learning and shape analysis to diagnose lung cancer from chest CT images. | Luna16 |
6 | 12 | [48,49] | CXR | KNN classification of nodules as blobs. Used stratification of JSRT to train/calibrate schemes to reduce false positive detection by algorithms. | JSRT |
7 | 12 | [50] | CXR | A set of older papers using various techniques to detect nodules and reduce false-positive detections | Private Data JSRT |