The two-stage detection-after-segmentation model improves the accuracy of identifying subdiaphragmatic lesions

Abstract

Chest X-rays (CXRs) are primarily used to detect lung lesions. While the abdominal portion of CXRs can sometimes reveal critical conditions, research in this area is limited. To address this, we introduce a two-stage architecture that separates the abdominal region from the CXR and detects abdominal lesions using a specialized dataset. We compared the performance of our method on whole CXRs versus isolated abdominal regions. First, we created masks for the right upper quadrant (RUQ), left upper quadrant (LUQ), and upper abdomen (ABD) regions and trained corresponding segmentation models for each area. For detecting abdominal lesions, we curated a dataset of 5,996 images, categorized into 19 classes based on anatomical locations, air patterns, and levels of stomach or bowel dilation. The detection process was initially conducted on the original images, followed by the three regional areas, RUQ, LUQ, and ABD, extracted by the segmentation models. The results showed that the detection model trained on the entire ABD region achieved the highest accuracy, followed closely by the RUQ and LUQ models. In contrast, the CXR model had the lowest accuracy. This study highlights that the two-stage architecture effectively manages distinct segmentation and detection tasks in CXRs, offering a promising avenue for more accurate diagnoses. It also suggests that an optimal ratio between the sizes of the target lesions and the input images may exist.

Fig. 1

An overview of dataset preparation, model training, and evaluation in the two-stage detection-after-segmentation architecture. The figure on the right illustrates the first stage: (A) We used the image from National Institutes of Health Chest X-ray Dataset (NIHCXR) image_001 to create three masks: right upper quadrant (RUQ), left upper quadrant (LUQ), and upper abdomen (ABD). These masks were combined with their corresponding images to form a dataset of mask/image pairs, split into training and validation sets in a ratio of 0.8:0.2. (B) The UNet architecture was employed to train the segmentation model. After each training epoch, the models’ performance was evaluated using the mean Intersection over Union (mIoU) to determine the optimal threshold and number of epochs. (C) The subplot on the right shows the mIoU-threshold curve derived from 200 test samples after a particular epoch. The y-axis represents mIoU values, and the x-axis represents thresholds, ranging from 0.1 to 0.9. The subplot on the left illustrates the mIoU-threshold curve over the training epochs, showing that the mIoU curves peaked at approximately 15–20 epochs, with no further improvement observed beyond that point. The curve generally forms a dome shape, peaking between thresholds 0.5, 0.6, 0.7, and 0.8. The final segmentation model used 0.5 as the cutoff threshold for the prediction mask. In the workflows from image_002 and image_003, we move to the detection stage: (D) The original images were extracted from NIHCXR image_002 and image_003, focusing solely on the abdominal portion of the CXR. Based on anatomical locations, air patterns, and levels of stomach or bowel dilation, a subdiaphragmatic dataset was established, comprising 19 classes and a total of 5,996 images. (E) This labeled subdiaphragmatic dataset was partitioned into training/validation and testing datasets. It was then input into the previously trained segmentation models, generating three labeled subdiaphragmatic training sets (RUQ, ABD, LUQ) alongside the full CXR as a control group for comparison during training. (F) A neural network model with four CNN layers was used. Multiple models were trained depending on the region of interest. For RUQ lesion detection, RUQ, ABD, and whole CXR models were trained. For LUQ lesion detection, LUQ, ABD, and CXR models were trained. For bilateral subdiaphragmatic air detection, ABD and CXR models were used. (G) In the detection stage, we evaluated the models using the area under the receiver-operating characteristic curve (ROC AUC), along with TensorFlow’s built-in evaluation metrics such as ROC, PRC, accuracy, recall (sensitivity), specificity, and F1-scores.