Statistical detection of nonvisible isodense subdural fluid collections

Area of Science:

• Radiology
• Medical Imaging Analysis
• Neurological Diagnostics

Background:

• Isodense subdural hematomas (ISDH) pose diagnostic challenges on computed tomography (CT) scans due to similar attenuation values.
• Accurate detection of ISDH is crucial for timely patient management and treatment.

Purpose of the Study:

• To present and validate an analytic method for detecting isodense subdural fluid collections using CT scan pixel attenuation data.
• To evaluate the accuracy of this method in identifying isodense subdural hematomas (ISDH).

Main Methods:

• Development of an analytic method based on computed tomography (CT) scan pixel attenuation data.
• Analysis of 44 CT scan levels from patients with confirmed ISDH and 50 CT scan levels from patients without ISDH.
• Restriction of analysis to noncontrast CT scans exhibiting subcalvarial bilateral cortical attenuation symmetry.

Main Results:

• The proposed analytic method demonstrated a 6% false positive rate.
• A 3% false negative rate was observed in the detection of isodense subdural hematomas (ISDH).
• The method showed high accuracy when applied to noncontrast CT scans with specific symmetry characteristics.

Conclusions:

• The presented analytic method offers a reliable approach for detecting isodense subdural fluid collections on CT scans.
• This technique, particularly with noncontrast scans showing symmetry, can aid in the diagnosis of isodense subdural hematomas (ISDH) with minimal errors.