Canal-wall up cholesteatoma surgery with mastoid obliteration leads to lower rates of disease recurrence without affecting hearing outcomes.

Objectives: The primary objective was to determine whether obliteration of the epitympanic area and mastoid cavity during canal wall up (CWU) cholesteatoma surgery reduces the rate of recurrent and residual cholesteatoma compared to not obliterating the same area. The secondary objective was to compare postoperative hearing outcomes between both techniques. Methods: A retrospective cohort study was conducted in a tertiary referral center. One-hundred-fourty-three ears were included of patients (≥18y) who underwent a CWU tympanomastoidectomy for cholesteatoma with or without bony obliteration between January 2015 and March 2020 in the University Medical Center Utrecht. The median follow-up was respectively 1.4 (IQR 1.1-2.2) vs. 2.0 years (IQR 1.2-3.1) (p = 0.013). Interventions: All patients underwent CWU tympanomastoidectomy for cholesteatoma. For 73 ears bone dust, Bonalive® or a combination was used for obliteration of the mastoid and epitympanic area, the rest of the ears (n = 70) were not obliterated. In accordance with the Dutch protocol, included patients are planned to undergo an MRI scan with diffusion-weighted imaging (DWI) one, three and five years after surgery to detect recurrent or residual cholesteatoma. Main outcome measures: The primary outcome measure was recurrent and residual cholesteatoma as evaluated by MRI-DWI and/or micro-otoscopy and confirmed by micro-otoscopy and/or revision surgery. The secondary outcome measure was the postoperative hearing. Results: In this cohort, the group treated with canal wall up tympanomastoidectomy with subsequent bony obliteration (73 ears, 51.0%) had significantly lower recurrent (4.1%) and residual (6.8%) cholesteatoma rates than the group without obliteration (70 ears, 25.7% and 20.0%, respectively; p < 0.001). There was no significant difference between both groups in postoperative bone conduction thresholds (mean difference 2.7 dB, p = 0.221) as well as the mean air-bone gap closure 6 weeks after surgery (2.3 dB in the non-obliteration and 1.5 dB in the obliteration group, p = 0.903). Conclusions: Based on our results, a canal wall up tympanomastoidectomy with bony obliteration is the treatment of choice, since the recurrent and residual disease rate is lower compared to the group without obliteration. The bony obliteration technique does not seem to affect the perceptive or conductive hearing results, as these are similar between both groups.

[Ulcers on a hairy scalp]. / Ulcera op het behaarde hoofd.

A 74-year-old woman presented herself at the outpatient clinic for dermatology with three temporoparietal ulcers on her scalp which are diagnosed as complications of temporal arteritis. The core symptoms of temporal arteritis are often non-specific, causing diagnostic delay. This dermatological complication can be an important clinical clue urging the physician to start treatment.

Deep Learning for Identification of Acute Illness and Facial Cues of Illness.

Background: The inclusion of facial and bodily cues (clinical gestalt) in machine learning (ML) models improves the assessment of patients' health status, as shown in genetic syndromes and acute coronary syndrome. It is unknown if the inclusion of clinical gestalt improves ML-based classification of acutely ill patients. As in previous research in ML analysis of medical images, simulated or augmented data may be used to assess the usability of clinical gestalt. Objective: To assess whether a deep learning algorithm trained on a dataset of simulated and augmented facial photographs reflecting acutely ill patients can distinguish between healthy and LPS-infused, acutely ill individuals. Methods: Photographs from twenty-six volunteers whose facial features were manipulated to resemble a state of acute illness were used to extract features of illness and generate a synthetic dataset of acutely ill photographs, using a neural transfer convolutional neural network (NT-CNN) for data augmentation. Then, four distinct CNNs were trained on different parts of the facial photographs and concatenated into one final, stacked CNN which classified individuals as healthy or acutely ill. Finally, the stacked CNN was validated in an external dataset of volunteers injected with lipopolysaccharide (LPS). Results: In the external validation set, the four individual feature models distinguished acutely ill patients with sensitivities ranging from 10.5% (95% CI, 1.3-33.1% for the skin model) to 89.4% (66.9-98.7%, for the nose model). Specificity ranged from 42.1% (20.3-66.5%) for the nose model and 94.7% (73.9-99.9%) for skin. The stacked model combining all four facial features achieved an area under the receiver characteristic operating curve (AUROC) of 0.67 (0.62-0.71) and distinguished acutely ill patients with a sensitivity of 100% (82.35-100.00%) and specificity of 42.11% (20.25-66.50%). Conclusion: A deep learning algorithm trained on a synthetic, augmented dataset of facial photographs distinguished between healthy and simulated acutely ill individuals, demonstrating that synthetically generated data can be used to develop algorithms for health conditions in which large datasets are difficult to obtain. These results support the potential of facial feature analysis algorithms to support the diagnosis of acute illness.