Adult ; Aged ; Anthracosis ; diagnostic imaging ; therapy ; Bronchoalveolar Lavage ; Humans ; Lung ; diagnostic imaging ; Male ; Middle Aged ; Radiography

Objective: To construct and verify a light-weighted convolutional neural network (CNN), and explore its application value for screening the early stage (subcategory 0/1 and stage Ⅰ of pneumoconiosis) of coal workers' pneumoconiosis (CWP) from digital chest radiography (DR) . Methods: A total of 1225 DR images of coal workers who were examined at an Occupational Disease Prevention and Control Institute in Anhui Province from October 2018 to March 2021 were retrospectively collected. All DR images were collectively diagnosed by 3 radiologists with diagnostic qualifications and gave diagnostic results. There were 692 DR images with small opacity profusion 0/- or 0/0 and 533 DR images with small opacity profusion 0/1 to stage Ⅲ of pneumoconiosis. The original chest radiographs were preprocessed differently to generate four datasets, namely 16-bit grayscale original image set (Origin16), 8-bit grayscale original image set (Origin 8), 16-bit grayscale histogram equalized image set (HE16) and 8-bit grayscale histogram equalized image set (HE8). The light-weighted CNN, ShuffleNet, was applied to train the generated prediction model on the four datasets separately. The performance of the four models for pneumoconiosis prediction was evaluated on a test set containing 130 DR images using measures such as the receiver operating characteristic (ROC) curve, accuracy, sensitivity, specificity, and Youden index. The Kappa consistency test was used to compare the agreement between the model predictions and the physician diagnosed pneumoconiosis results. Results: Origin16 model achieved the highest ROC area under the curve (AUC=0.958), accuracy (92.3%), specificity (92.9%), and Youden index (0.8452) for predicting pneumoconiosis, with a sensitivity of 91.7%. And the highest consistency between identification and physician diagnosis was observed for Origin16 model (Kappa value was 0.845, 95%CI: 0.753-0.937, P<0.001). HE16 model had the highest sensitivity (98.3%) . Conclusion: The light-weighted CNN ShuffleNet model can efficiently identify the early stages of CWP, and its application in the early screening of CWP can effectively improve physicians' work efficiency.
Humans ; Retrospective Studies ; Anthracosis/diagnostic imaging* ; Pneumoconiosis/diagnostic imaging* ; Coal Mining ; Neural Networks, Computer ; Coal

OBJECTIVETo study the consistency between DR and high-kV chest radiographs in diagnosis of pneumoconiosis and to explore the feasibility of DR chest radiograph in the diagnosis of pneumoconiosis.

METHODSTwenty five coal miners were examined with DR and high-kV chest radiographs at the same time. Image post-processing parameters (density, contrast, etc.) were set to ensure the quality of DR chest radiograph in Philips Essenta DR machine. In order to avoid the repetitive numbers, 50 chest radiographs were numbered at random. Pneumoconiosis diagnosis was conducted by six independent certified occupational physicians of pneumoconiosis by blind method. The consistency between 2 kinds of chest films was assessed.

RESULTSAll chest radiographs (25 cases, 50 chest films) were excellent. The diagnosis results of six readers on the 15 pairs of DR and high-kV chest radiographies were summarized. For high-kV chest radiographs, the consistency of pneumoconiosis classification for 12 pairs of readers was more than 68%, the consistency of total density for 11 pairs of readers was more than 68%. For DR chest radiographs, the consistency of pneumoconiosis classification for 13 pairs of readers was more than 60%, the consistency of total density for 14 pairs of readers was more than 60%. The consistency of pneumoconiosis classification between two chest radiographs was 72% (value: 0.69, 95% CI: 0.46-0.92), the consistency of total density between two chest radiographs was 80% (value: 0.78, 95% CI: 0.61-0.95).

CONCLUSIONWhen the chest radiograph quality was good and the difference of reading films was low, there was a good consistency of pneumoconiosis diagnosis between DR chest radiographs and high-kV chest radiographs.

Anthracosis ; diagnostic imaging ; Humans ; Radiographic Image Enhancement ; methods ; Radiography, Thoracic ; methods

OBJECTIVETo explore the effects of image post-processing parameters on DR chest radiograph for the diagnosis of pneumoconiosis.

METHODSEighty three coal miners were examined with high-kV and DR chest radiographs at the same time. Image post-processing parameters (density, contrast and so on) were designed in a Philips Essenta DR machine were designed, then differences of image quality between high-kV and DR chest radiographs were compared.

RESULTSAfter regulating image and proceeding the parameters, the OD (optical density) values of high density areas in the upper-middle lung fields, subphrenic and direct exposure areas were 1.58 +/- 0.10, 0.23 +/- 0.02 and 2.80 +/- 0.21, respectively. The quality of chest films met the requirements of diagnostic criteria of pneumoconiosis. The rate of excellent chest films for DR chest radiograph was 95.18%, which was significantly higher than that (80.72%) for high-kV chest radiograph (P < 0.01).

CONCLUSIONAppropriate parameters of image post-processing can make DR chest radiograph to meet the requirements of chest radiograph quality for the diagnosis of pneumoconiosis.

Anthracosis ; diagnostic imaging ; Humans ; Image Processing, Computer-Assisted ; Radiographic Image Enhancement ; methods ; Radiography, Thoracic ; methods

Aged ; Anthracosis ; complications ; diagnostic imaging ; Humans ; Male ; Middle Aged ; Pulmonary Emphysema ; complications ; diagnostic imaging ; Tomography, Spiral Computed ; Trachea ; diagnostic imaging ; pathology

OBJECTIVETo study the diagnostic value of CT/HRCT for the coal workers' pneumoconiosis (CWP) with large opacities.

METHODSSixty-two patients with CWP were examined with both chest radiograph and CT/HRCT scan. Comparison was done.

RESULTSLarge opacities were found in 19 patients in the chest radiograph of 62 patients with CWP. The detection rate was 30.6%. Thirty large opacities were found using chest radiograph. Forty-three patients with large opacities were found using the CT/HRCT. The detection rate was 69.4%. One hundred and twenty-three large opacities were found using CT/HRCT. Ninety-three more large opacities were found by CT/HRCT than by chest radiograph. The total detection rate of large opacities by CT was 4.1 times than by chest radiograph. Both differences were statistically significant (chi2 = 18.58, P < 0.01). CT/HRCT found all patients with large opacities detected in the chest radiograph. Seventeen patients with Stage III large opacities (>or= 1.0 cm x 2.0 cm) were found by the chest radiograph. The detection rate was 27.4% (17/62). Twenty-seven large opacities were found using chest radiograph. Thirty-four patients with Stage III large opacities were found using the CT/HRCT. The detection rate was 54.8%. The found Stage III large opacities added up to 67. The found Stage III large opacities by CT/HRCT were 40 more than using chest radiograph. Both differences were statistically significant (chi2 = 9.63, P < 0.01). CT/HRCT was reliable to show the morphology, the inside structure, the tiny lesion and the special location of the large opacities of CWP. It could provide more important information for differential diagnosis.

CONCLUSIONCT/HRCT is significantly better than the chest radiograph in the diagnosis of complicated CWP. It can find the large opacities that the chest radiograph can not. It has a great diagnostic and practical value as an assistant examination method. CT/HRCT could be considered as the reference and evidence for staging progression in diagnosis of pneumoconiosis.

Adult ; Aged ; Aged, 80 and over ; Anthracosis ; diagnostic imaging ; Coal Mining ; Humans ; Male ; Middle Aged ; Tomography, X-Ray Computed ; methods

Adult ; Aged ; Aged, 80 and over ; Anthracosis ; complications ; Diagnostic Imaging ; Humans ; Lung Neoplasms ; epidemiology ; etiology ; Middle Aged ; Retrospective Studies

OBJECTIVETo explore the sensitivity and accuracy of CT and HRCT in early detection of coal-worker's pneumoconiosis (CWP) associated micronodules, nodule coalescence and emphysema from coal miners with and without radiographic CWP evidence.

METHODContinuously Thorax-Vol. CT scanning and High resolution CT scanning were performed on 113 coal miners with or without radiographic CWP evidence and 37 health controls by the use of Multi Spiral Computed Tomography (MSCT). The CT and HRCT images were evaluated for lung parenchyma opacity profusion category, nodule coalescence, emphysema index and mean lung attenuation, and then compared with that on radiographs.

RESULTGood agreement for CWP associated opacity profusion grading was achieved between radiographs and CT scans (Kappa = 0.771). The sensitivity of CT and HRCT were 98.70% for CWP diagnosis when compared with radiographs. 8 (22.22%) cases out of 36 radiographic CWP negative coal miners were categorized as grade 1 opacity profusion in consistent with CWP on CT and HRCT scans, and 26 (35.62%) cases out of 73 radiographic grade 1-2 CWP patients were detected to have at least grade 3 nodular profusion on CT and HRCT scans. Among 113 coal miners, 36 (31.86%) were recognized as emphysema positive on CT; whereas only 7 (6.19%) on radiographs. According to CT opacity profusion grading, mean lung attenuation in coal miners with grade 1-2 opacity profusion classification was significantly increased than that in health control, CWP negative coal miner and grade 4 opacity profusion groups (F1 = -45.73, F2 = -23.00 and F3 = 52.72, P < 0.01 and P < 0.05).

CONCLUSIONCT and HRCT acquired from MSCT showed high sensitive and accurate for the early detection of micronodules, nodule coalescence and emphysema in coal miners. It could be used as semi-quantitative and quantitative method in early diagnosis of CWP and its complications.

Adult ; Aged ; Anthracosis ; diagnostic imaging ; Case-Control Studies ; Early Diagnosis ; Humans ; Lung ; diagnostic imaging ; Male ; Middle Aged ; Pulmonary Emphysema ; diagnostic imaging ; Sensitivity and Specificity ; Tomography, Spiral Computed ; Tomography, X-Ray Computed ; methods