Impact of an AI software on the diagnostic performance and reading time for the detection of cerebral aneurysms on time of flight MR-angiography

PURPOSE

To evaluate the impact of an AI-based software trained to detect cerebral aneurysms on TOF-MRA on the diagnostic performance and reading times across readers with varying experience levels.

METHODS

One hundred eighty-six MRI studies were reviewed by six readers to detect cerebral aneurysms. Initially, readings were assisted by the CNN-based software mdbrain. After 6 weeks, a second reading was conducted without software assistance. The results were compared to the consensus reading of two neuroradiological specialists and sensitivity (lesion and patient level), specificity (patient level), and false positives per case were calculated for the group of all readers, for the subgroup of physicians, and for each individual reader. Also, reading times for each reader were measured.

RESULTS

The dataset contained 54 aneurysms. The readers had no experience (three medical students), 2 years experience (resident in neuroradiology), 6 years experience (radiologist), and 12 years (neuroradiologist). Significant improvements of overall specificity and the overall number of false positives per case were observed in the reading with AI support. For the physicians, we found significant improvements of sensitivity on lesion and patient level and false positives per case. Four readers experienced reduced reading times with the software, while two encountered increased times.

CONCLUSION

In the reading with the AI-based software, we observed significant improvements in terms of specificity and false positives per case for the group of all readers and significant improvements of sensitivity and false positives per case for the physicians. Further studies are needed to investigate the effects of the AI-based software in a prospective setting.

Delayed Internal Carotid Artery Dissection Mimicking Cerebral Vasospasms after Subarachnoid Hemorrhage: A Case Report

BACKGROUND

Delayed cerebral ischemia (DCI) is usually caused by cerebral vasospasm (CVS). To detect DCI and CVS a cranial CT scan will be performed, but cervical vessels are not necessarily displayed.

PATIENT

A 63-year-old female patient who suffered from aneurysmal subarachnoid hemorrhage (SAH) was treated at the authors' institution. After an initially unremarkable clinical course, she developed aphasia on day 11. CT angiography (CTA) and perfusion imaging revealed significant hypoperfusion of the left hemisphere. In addition, the CTA showed a subtotal stenosis of the internal carotid artery (ICA) at the level of the petrous segment, suspicious for a dissection. This was not detectable angiographically in the final control of the intervention and was also not clinically evident until day 11. Cerebral perfusion as well as the clinical symptoms normalized rapidly after stent reconstruction of the ICA.

CONCLUSION

Even though CVS is the most frequent cause of hypoperfusion in patients after SAH, a periinterventional dissection can also lead to relevant stenosis and thus to a disturbed cerebral perfusion and corresponding neurological deficits. The time delay between the intervention and the clinical as well as CT-angiographical manifestation in our case is remarkable.

Sonographic assessment of the optic nerve and the central retinal artery in idiopathic intracranial hypertension

PURPOSE

Transorbital sonography easily detects papilledema and enlarged optic nerve sheath diameters (ONSD) in IIH (idiopathic intracranial hypertension) patients. As the central retinal artery is located within the optic nerve, its hemodynamic properties might be affected by the increased pressure. In this study we assessed the diagnostic usefulness of transorbital sonography in IIH with a special focus on color Doppler imaging of the central retinal artery. IIH patients presented papilledema and enlarged ONSD. ONSD accurately predicted an increased intracranial pressure in IIH (cut-off: 5.8 mm, 81% sensitivity, 80% specificity). 24 h following therapeutic lumbar puncture ONSD diminished significantly, whereas papilledema was not changed. PSV (peak systolic velocity) and Vmean (mean flow velocity) of the central retinal artery were increased in IIH patients compared to controls. PSV accurately predicted an increase of intracranial pressure (cut-off: 11.0 cm/s, 70% sensitivity, 69% specificity). PI (pulsatility index), PSV and Vmean decreased following lumbar puncture. PSV and Vmean decreases were statistically significant for right eyes only in which the values changed to normal. In summary, besides ONSD enlargement and papilledema transbulbar sonography demonstrated an alteration of central retinal artery blood flow in IIH patients. Especially PSV might serve as valuable surrogate marker for intracranial pressure in IIH. Furthermore, the change of intra-individual central retinal arteries PI might be a valuable parameter to demonstrate response to lumbar puncture in IIH patients.

Limbic Encephalitis in Patients with Epilepsy-is Quantitative MRI Diagnostic?

PURPOSE

Limbic encephalitis (LE) is an immune-related disease with limbic symptoms, variable and asymmetric magnetic resonance imaging (MRI) aspects and antibody profiles. This study investigated the diagnostic value of quantitative relaxation times T2 (qT2) and MRI signal intensities (SI) in LE.

METHODS

The prospective 3T-MRI study included 39 epilepsy patients with initially suspected LE and 20 healthy controls. Values and asymmetry indices of qT2, T2-weighted (T2-w) and proton density (PD)-w SI of manually delineated and automatically segmented amygdala and hippocampus were measured. Additionally, two raters made a blinded visual analysis on FLAIR (fluid attenuation inversion recovery) and T2-w images.

RESULTS

According to diagnostic guidelines, 22 patients had probable LE and 17 patients had possible LE. The qT2 was higher (p < 0.01) in patients than in controls (mean ± SD, amygdala 98 ± 7 ms vs. 90 ± 5 ms, hippocampus 101 ± 7 ms vs. 92 ± 3 ms), but was not different between probable and possible LE or between sides (left and right). The PD-w SI and T2-w SI were lower in patients than in controls but were not different between patient subgroups or between sides. Diagnostic performance of visual analysis was relatively poor.

CONCLUSIONS

Epilepsy patients with suspected LE had elevated qT2 in amygdala and hippocampus, whereas the expected T2-w SI increase was not found; however, the diagnostic value of qT2 remains questionable since it did not discriminate probable from possible LE.