Clinical Validation and Extension of an Automated, Deep Learning-Based Algorithm for Quantitative Sinus CT Analysis

BACKGROUND AND PURPOSE

Sinus CT is critically important for the diagnosis of chronic rhinosinusitis. While CT is sensitive for detecting mucosal disease, automated methods for objective quantification of sinus opacification are lacking. We describe new measurements and further clinical validation of automated CT analysis using a convolutional neural network in a chronic rhinosinusitis population. This technology produces volumetric segmentations that permit calculation of percentage sinus opacification, mean Hounsfield units of opacities, and percentage of osteitis.

MATERIALS AND METHODS

Demographic and clinical data were collected retrospectively from adult patients with chronic rhinosinusitis, including serum eosinophil count, Lund-Kennedy endoscopic scores, and the SinoNasal Outcomes Test-22. CT scans were scored using the Lund-Mackay score and the Global Osteitis Scoring Scale. CT images were automatically segmented and analyzed for percentage opacification, mean Hounsfield unit of opacities, and percentage osteitis. These readouts were correlated with visual scoring systems and with disease parameters using the Spearman ρ.

RESULTS

Eighty-eight subjects were included. The algorithm successfully segmented 100% of scans and calculated features in a diverse population with CT images obtained on different scanners. A strong correlation existed between percentage opacification and the Lund-Mackay score (ρ = 0.85, P < .001). Both percentage opacification and the Lund-Mackay score exhibited moderate correlations with the Lund-Kennedy score (ρ = 0.58, P < .001, and ρ = 0.58, P < .001, respectively). The percentage osteitis correlated moderately with the Global Osteitis Scoring Scale (ρ = 0.48, P < .001).

CONCLUSIONS

Our quantitative processing of sinus CT images provides objective measures that correspond well to established visual scoring methods. While automation is a clear benefit here, validation may be needed in a prospective, multi-institutional setting.

Comparison of Super Stuff and paraffin wax bolus in radiation therapy of irregular surfaces

Irregular facial contours can make radiation therapy of head and neck tumors difficult. Isodose lines become skewed, making treatment planning complex. A traditional solution to this problem is the paraffin wax box bolus. Such a bolus is made to fit the irregular surface compensating for the topology and creating an even surface. The fabrication of a wax bolus can be a difficult and time-consuming process. A method that is simple and efficient has been devised. Super Stuff bolus can be easily molded and has approximately the same effect as a similar paraffin wax bolus. This was verified by irradiating a Rando head phantom with both a paraffin wax bolus and a Super Stuff bolus. Doses to various points of interest were measured with thermoluminescent dosimetry (TLD) chips (LiF). The particular case addressed is malignant melanoma of the nasal septum, but the technique described can be useful in the treatment of other sites as well.

Markers of serious illness in infants under 6 months old presenting to a children's hospital

Six hundred and eighty two assessments were performed on 641 babies under 6 months of age who presented to the emergency department of the Royal Children's Hospital, Melbourne, to try and determine the best markers of serious illness in young infants. Detailed, specific questions that quantified a baby's functional response to illness gave the most useful information. As a group, the six most common predictive symptoms of serious illness were: taking less than half the normal amount of feed over the preceding 24 hours, breathing difficulty, having less than four wet nappies in the preceding 24 hours, decreased activity, drowsiness, and a history of being both pale and hot. The presence of the corresponding sign on examination increased the predictive value of the symptom by 10-20%. Specific, highly predictive (though less common) signs included moderate to severe chest wall recession, respiratory grunt, cold calves, and a tender abdomen. A list of low, medium, and high risk symptoms has been constructed and the five measurements that were most useful in predicting serious illness in young infants have been detailed.