Virtual endoscopy of laryngeal carcinoma: Is it useful?

Comparison of diagnostic accuracy of computed tomography virtual endoscopy and flexible fibre-optic laryngoscopy in the evaluation of neck anatomic structures and neoplasms

BACKGROUND AND PURPOSE

Computed tomography virtual endoscopy (CT-VE) is a non-invasive technique which allows visualisation of intraluminal surfaces by tridimensional reconstruction of air/soft tissues. The aim of this study was to compare the diagnostic accuracy of CT-VE and flexible fibre-optic laryngoscopy (FFL) in identifying normal neck anatomic structures and pharyngeal and laryngeal lesions.

METHODS

Forty-two patients with a history of neck cancer were assessed by two ENT surgeons using FFL and by one neuroradiologist using CT-VE in order to evaluate the visualisation of the epiglottis, vallecula, glossoepiglottic folds, pyriform sinuses, vocal cords and mass pathology. The visualisation of the structures in both modalities was assessed according to the following score: 0 = not visualised, 1 = partial visualisation, 2 = complete and clear visualisation. A weighted kappa coefficient was used to evaluate the inter-observer agreement. McNemar's test was performed to compare the two diagnostic tests.

RESULTS

The inter-observer agreement between FFL and CT-VE was fair in the assessment of the vocal cords (k = 0.341); moderate in the assessment of the glossoepiglottic folds (k = 0.418), epiglottis (k = 0.513) and pyriform sinuses (k = 0.477); and substantial in the assessment of the vallecula (k = 0.618) and the tumour (0.740). McNemar's test showed no significant difference between the two tests (p<0.05).

CONCLUSION

CT-VE is a non-invasive technique with a diagnostic accuracy comparable to FFL in terms of visualisation of anatomical structures and pharyngeal and laryngeal lesions.

Terminology, Definitions, and Classification in the Imaging of Laryngoceles

OBJECTIVE

We aim to better define the anatomy, criteria, and classification of laryngoceles, and to raise the awareness to potential mimickers and anatomical variants leading to misdiagnosis.

METHODS

A retrospective review of all computed tomography studies, performed over a decade, with the diagnosis of "laryngocele" in the radiological report in a tertiary medical center. All relevant studies were reviewed by two independent readers for the presence of true laryngoceles.

RESULTS

One hundred and twelve patients were included; average age was 54 (±18). Re-read of scans with 3D reconstructions resulted in detecting 58 (52%) true laryngoceles, with 19.5% bilateral laryngoceles. Anatomical variants and laryngocele mimickers formerly misdiagnosed as laryngoceles included 26/54 (48%) prominent ventricles, 19/54 (35%) saccules not meeting criteria for laryngocele, 8/54 (15%) prominent pyriform sinuses and one tracheal diverticulum.

CONCLUSIONS

Intimate knowledge of the laryngeal anatomy, the criteria for a laryngocele and anatomical variants as well mimickers, is the key to avoid radiological misdiagnosis.

Clinical application of virtual endoscopy as a support system for endoscopic sinus surgery

CONCLUSIONS

Computer processing of conventional CT images can provide virtual endoscopic images (VEIs). Surgeons observing these images feel as if they are observing the lesion using an endoscope. Simulation based on animated VEIs before endoscopic sinus surgery (ESS) is useful for the improvement of surgical safety and surgeons' education. Although the production of VEIs requires certain surgical experience, after the production of VEIs, surgeons can repeat simulation of a surgical procedure and have confidence in the actual operation.

OBJECTIVE

We investigated the clinical application of CT-reconstructed VEIs as a support system for ESS.

MATERIALS AND METHODS

A GE Light Speed Ultra 16 as a 16-slice CT scanner and Advantage Workstation 4.2 were used. Software called Navigator in this workstation allows the production of CT-reconstructed VEIs. We applied simulation based on VEIs to endoscopic operations for mucocele or sinusitis.

RESULTS

In nine cases of mucoceles and sinusitis, simulation based on animated VEIs was applied. According to the simulation, surgeries were performed, and the mucoceles and obstructed sinuses could be opened readily and safely. This system was more effective when there were landmarks such as polyps or irregular mucosal surfaces allowing macroscopic confirmation ahead of the viewpoint.

Laryngeal cancer: how does the radiologist help?

The radiologist makes a valuable contribution to the staging of laryngeal cancer and this has a direct influence on treatment planning. This review focuses on the main anatomical concepts, patterns of tumour spread and how to detect this with optimal cross sectional imaging. Issues surrounding the relationship of tumour to the ventricular complex, submucosal laryngeal spaces, anteroposterior extension, laryngeal cartilage involvement and metastatic spread are discussed and illustrated. The impact of these imaging findings on the array of therapeutic options is described.

Imaging in head and neck cancer

The goals of imaging in head and neck cancer are to establish tumor extent and size, to assess nodal disease, to evaluate for perineural tumor spread, and to distinguish recurrent tumor from post-treatment changes. MRI is the preferred modality for assessment of nasopharyngeal, sinonasal, and parotid tumors, because of better contrast resolution, high frequency of perineural spread, and less prominent motion artifacts. MRI is the best modality to delineate the extent of intraorbital and intracranial extension of malignant tumors. Tumors of the oropharynx, larynx, and hypopharynx are frequently primarily imaged with CT, which is less affected by breathing and swallowing artifacts. MRI is also the initial study of choice for tumors confined to the oral tongue, and possibly also for other oral cavity locations because MRI is superior in detection of tumor spread into the bone marrow. There is no clear advantage of CT or MRI for evaluation of nodal disease. Positron emission tomography (PET) is very sensitive for metastatic lymph nodes that are at least 8 mm in size and is the technique of choice in dubious cases. Imaging-guided biopsies are performed whenever needed. For imaging of treated head and neck cancer, PET scans have been found to generally offer higher sensitivity than MRI or CT. Combined PET/CT may be the modality of choice because it almost completely eliminates the false-positive and false-negative PET findings. Patients with head and neck cancer who are referred to tertiary care centers commonly arrive with cross-sectional images obtained at other institutions. Reinterpretation of these studies by dedicated radiologists frequently leads to changes in findings, which alter treatment and affect prognosis.