Using Artificial Intelligence to Establish Chest X-Ray Image Recognition Model to Assist Crucial Diagnosis in Elder Patients With Dyspnea

Abstract

Pneumonia and pulmonary edema are the most common causes of acute respiratory failure in emergency and intensive care. Airway maintenance and heart function preservation are two foundations for resuscitation. Laboratory examinations have been utilized for clinicians to early differentiate pneumonia and pulmonary edema; however, none can provide results as prompt as radiology examinations, such as portable chest X-ray (CXR), which can quickly deliver results without mobilizing patients. However, similar features between pneumonia and pulmonary edema are found in CXR. It remains challenging for Emergency Department (ED) physicians to make immediate decisions as radiologists cannot be on-site all the time and provide support. Thus, Accurate interpretation of images remains challenging in the emergency setting. References have shown that deep convolutional neural networks (CNN) have a high sensitivity in CXR readings. In this retrospective study, we collected the CXR images of patients over 65 hospitalized with pneumonia or pulmonary edema diagnosis between 2016 and 2020. After using the ICD-10 codes to select qualified patient records and removing the duplicated ones, we used keywords to label the image reports found in the electronic medical record (EMR) system. After that, we categorized their CXR images into five categories: positive correlation, negative correlation, no correlation, low correlation, and high correlation. Subcategorization was also performed to better differentiate characteristics. We applied six experiments includes the crop interference and non-interference categories by GoogLeNet and applied three times of validations. In our best model, the F1 scores for pneumonia and pulmonary edema are 0.835 and 0.829, respectively; accuracy rate: 83.2%, Recall rate: 83.2%, positive predictive value: 83.3%, and F1 Score: 0.832. After the validation, the best accuracy rate of our model can reach up to 73%. The model has a high negative predictive value of excluding pulmonary edema, meaning the CXR shows no sign of pulmonary edema. At the time, there was a high positive predictive value in pneumonia. In that way, we could use it as a clinical decision support (CDS) system to rule out pulmonary edema and rule in pneumonia contributing to the critical care of the elderly.

Use of Electronic Medical Records (EMR) in Gerontology: Benefits, Considerations and a Promising Future

“…22 Table 1 lists the selected articles that demonstrate the use of EMRs in different areas, including diagnosis of specific diseases, pharmacology, and clinical decision support. 17,[23][24][25][26][27][28][29] Notable in the literature review is the large number of retrospective studies using the already collected EMR datasets. 24,28,30,31 There are also numerous pilot studies most often aimed at improving EMR-based diagnostics.…”

“…17,[23][24][25][26][27][28][29] Notable in the literature review is the large number of retrospective studies using the already collected EMR datasets. 24,28,30,31 There are also numerous pilot studies most often aimed at improving EMR-based diagnostics. 16,29,32 EMRs offer rich resources and opportunities to improve research (and clinical care), especially for undiagnosed dementia.…”

“… 17 , 23–29 Notable in the literature review is the large number of retrospective studies using the already collected EMR datasets. 24 , 28 , 30 , 31 There are also numerous pilot studies most often aimed at improving EMR-based diagnostics. 16 , 29 , 32

Figure 2 Summary of benefits and limitations of using electronic medical records.

…”

An Arrhythmia classification approach via deep learning using single-lead ECG without QRS wave detection

“…The traditional method of ECG detection relies on manually selecting features and making subjective judgments to determine heart rhythms. Convolutional neural networks (CNNs) utilize an operation method that eliminates the necessity for manual feature selection, rendering them highly effective in the domain of image processing [ 11 ]. Originally introduced in 1990 for the recognition of handwritten digits, researchers are currently incorporating artificial intelligence powered Computer-Aided Diagnosis (CAD) systems into medical imaging to facilitate early diagnosis and intervention by physicians [ 12 ].…”

An Arrhythmia Classification Approach via Deep Learning Using Single-lead ECG

“…The traditional method of ECG detection relies on manually selecting features and making subjective judgments to determine heart rhythms. Convolutional neural networks (CNNs) use an operation method with no need to select features; thus, CNNs are effective in the image eld [11]. However, the use of CNN architecture allows for the automatic extraction of ECG signal features in a non-linear manner.…”