Performance of a new portable near-infrared spectroscopy device for detection of traumatic intracranial hematoma

Abstract

IntroductionWe report results of a newly developed portable near-infrared spectroscopy (NIRS) based point-of-care device CEREBO® to detect traumatic intracranial hematoma (TICH). Materials and methodsPatients with alleged history of head injury visiting emergency room were enrolled. They were examined consecutively for the presence of TICH using CEREBO® and computed tomography (CT) scans. ResultsA total of 158 participants and 944 lobes were scanned, and 18% of lobes were found to have TICH on imaging with computed tomography of the head. 33.9% of the lobes could not be scanned due to scalp lacerations. The mean depth of hematoma was 0.8 (SD 0.5) cm and the mean volume of the hematoma was 7.8 (11.3) cc. The overall sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV) of CEREBO® to classify subjects as hemorrhagic or non-hemorrhagic were 96% (CI 90 - 99%), 85% (CI 73 - 93%), 92% (CI 86 - 96%), 91% (CI 84 - 96%) and 93% (CI 82 - 98%) whereas to classify the lobes as hemorrhagic and non-hemorrhagic, the sensitivity, specificity, accuracy, PPV and NPV were 93% (CI 88 - 96%), 90% (CI 87 - 92%), 90% (CI 88 - 92%), 66% (CI 61 - 73%), and 98% (CI 97 - 99%) respectively. The sensitivity was highest at 100% (CI 92 - 100%) for the detection of extradural and subdural hematoma. The sensitivity for detecting intracranial hematoma including epidural, subdural, intracerebral and subarachnoid hematomas, of more than 2 cc was 97% (CI 93 - 99%) and the NPV was 100% (CI 99 - 100%). The sensitivity dropped for hematomas less than 2cc in volume to 84% (CI 71 - 92%) and the NPV was 99% (CI 98 - 99%). The sensitivity to detect bilateral hematomas was 94% (CI 74 - 99%). ConclusionThe performance of currently tested NIRS device for detection of TICH was good and can be considered for triaging a patient requiring a CT scan of the head after injury. The NIRS device can efficiently detect traumatic unilateral hematomas as well as those bilateral hematomas where the volumetric difference is greater than 2cc.