Deep Learning Models to Predict Fatal Pneumonia Using Chest X-Ray Images

Table 1

Characteristics of the study patients in the CXR image original dataset for training of deep learning models (n = 1274).

Characteristics	N	(%)
Median age (range), years	75 (15–104)
Gender
Male	750	58.9
Female	524	41.1
Hospitalization for pneumonia treatment
No	282	22.1
Yes	992	77.9
Pneumonia type
CAP	417	32.7
Other than CAP (NHCAP, HAP, VAP)	857	67.3
Complication of congestive heart failure
No	1149	90.2
Yes	125	9.8
Positive results with the sputum culture test	455	35.7
Positive results with the blood culture test	20	1.6
Positive results with the pleural fluid culture test	4	0.3
Positive results with the bronchial lavage fluid culture test	7	0.5
Posteroanterior chest radiographs	841	66.0
Chest radiographs under intubation	15	1.2
Prognosis
Nonfatal	1031	80.9
Fatal	243	19.1

CAP, community-acquired pneumonia; NHCAP, nursing and healthcare-associated pneumonia; HAP, hospital-acquired pneumonia; VAP, ventilator-associated pneumonia.