Pneumonia Detection Using Deep Transfer Learning in Gender Specific Chest X-ray Images

Chest X-rays are one of the most commonly used image examination tools that are beneficial for diagnosing and managing various diseases, including pneumonia. In this study, state-of-the-art deep learning models are evaluated to detect pneumonia from chest X-ray images. It has been evaluated if the same pre-trained model with transfer learning approach performs differently on male or female-only or in mixed-gender datasets. After separating the Chest X-ray dataset from the Radiological Society of North America (RSNA) into gender-specific groups, several deep learning models are fine-tuned and trained on the dataset groups. Our study found that some models perform better on the male dataset, whereas a few others work better on the female dataset, and some perform well in the mixed-gender dataset. Health care centers working with only one type of gender, like maternity clinics, will benefit from our study by selecting the right model for a dataset instead of trial and error for pneumonia detection, by which clinicians and patients will be benefited in the long run.