A retrospective, blineded, multicenter study of the InferRead CT Pneumonia.AI to evaluate the performance in identifying non-contrast chest CT scans containing pneumonia findings.

This is a retrospective, blinded, multicenter study for the InferRead CT Pneumonia.AI. The primary endpoint will evaluate the software's performance in identifying non-contrast chest CT scans containing pneumonia. It will evaluate InferRead CT Pneumonia.AI in terms of sensitivity and specificity with respect to a ground truth, as established by trained thoracic radiologists, in the detection of penumonia. In addition, the study will compare time to notification for the InferRead CT Pneumonia.AI software with respect to the time to notification for the standard of care as established from available radiological reports.

1. To evaluate the performance of InferRead CT Pneumonia.AI.
2. To evaluate the effect of different factors such as CT manufacturer, CT scanning protocols and patient characteristics on the performance of InferRead CT Pneumonia.AI

Inclusion Criteria:

1. Non-contrast chest CT DICOM scans
2. Contains both lung lobes
3. Slice thickness is less than 3 mm.

Exclusion Criteria:

1. Incomplete scan or corrupted scan
2. Irregular scanning, increased intrapulmonary density due to insufficient inspiration, and respiratory artifacts affecting the physician's judgment.
3. Altered or absent lung morphology in the postoperative patient.

This retrospective study will be used to assess the performance of the subject device. The principal investigator will collect DICOM data and relevant clinical information based on the selection criteria. DICOM will be de-identified of PHI and stored in a secure location.

The validation dataset will contain non-contrast chest CT scans from patients in the United States that are either negative or have confirmed pneumonia (viral or bacterial). Data will come from at least 3 hospitals and cover a variety of data attributes.