Semi-Supervised Deep Learning Semantic Segmentation for 3D Volumetric Computed Tomographic Scoring of Chronic Rhinosinusitis: Clinical Correlations and Comparison with Lund-Mackay Scoring

Paranasal sinus occupancy assessed from magnetic resonance images-associations with clinical indicators in patients with systemic lupus erythematosus

OBJECTIVES

Nasal, paranasal sinus and mucosal disorders are common symptoms in autoimmune rheumatic diseases. Soft tissue changes and fluid accumulation in the osteomeatal complexes and paranasal sinuses manifest as opaqueness on radiological images which can be assessed using visual scoring and computational methods on CT scans, but their results do not always correlate. Using MRI, we investigate the applicability of different image analysis methods in SLE.

METHODS

We assessed paranasal sinus opaqueness on MRI from 51 SLE patients, using three visual scoring systems and expert-delineated computational volumes, and examined their association with markers of disease activity, inflammation, endothelial dysfunction and common small vessel disease (SVD) indicators, adjusting for age and sex-at-birth.

RESULTS

The average paranasal sinus volume occupation was 4.55 (6.47%) [median (interquartile range) = 0.67 (0.25-2.65) ml], mainly in the maxillary and ethmoid sinuses. It was highly correlated with Lund-Mackay (LM) scores modified at 50% opaqueness cut-off (Spearman's ρ: 0.71 maxillary and 0.618 ethmoids, P < 0.001 in all), and with more granular variations of the LM system. The modified LM scores were associated with SVD scores (0: B = 5.078, s.e. = 1.69, P = 0.0026; 2: B = -0.066, s.e. = 0.023, P = 0.0045) and disease activity (anti-dsDNA: B = 4.59, s.e. = 2.22, P = 0.045; SLEDAI 3-7: 2.86 < B < 4.30; 1.38 < s.e. < 1.63; 0.0083 ≤ P ≤ 0.0375). Computationally derived percent opaqueness yielded similar results.

CONCLUSION

In patients with SLE, MRI computational assessment of sinuses opaqueness and LM scores modified at a 50% cut-off may be useful tools in understanding the relationships among paranasal sinus occupancy, disease activity and SVD markers.

Blood Eosinophil Percentage and Improved Sinus CT Score as Diagnostic Tools for ECRS

Objective

Differentiating 2 types of chronic rhinosinusitis with nasal polyps (CRSwNP) is important for the treatment. The current diagnostic methods using single indicators, including peripheral blood eosinophils and traditional sinus computed tomography (CT) scores, are not accurate. In this study, we aimed to investigate the diagnostic value of combining peripheral blood eosinophils and improved sinus CT scores for eosinophic chronic rhinosinusitis (ECRS).

Study Design

Retrospective cohort.

Setting

Tertiary medical center.

Methods

We conducted a study involving 81 patients with CRSwNP. Peripheral blood samples were collected from the non-ECRS and ECRS groups. Improved three-dimensional volume image analysis and Lund-Mackay scoring system were performed to quantify the thickening of sinus mucosa. Multivariate binary logistic regression analysis was carried out to detect the predictive value of the scoring indicators. For significant indexes, receiver operating characteristic (ROC) curve analysis was applied.

Results

The ECRS group had higher levels of blood eosinophil percentage and count, ethmoid sinus score, total sinus score, the ratio of ethmoid sinus score and maxillary sinus score, and the difference between ethmoid and maxillary score, compared to the non-ECRS group (P < 0.05). Binary logistic regression analysis demonstrated that both blood eosinophil percentage and the improved E - M score (subtraction of ethmoid and maxillary sinus scores) were significant predictors of ECRS diagnosis (P < .01). ROC curve analysis indicated that the combination of improved E - M score and blood eosinophil percentage had a higher diagnostic value compared to either factor alone (area under the curve = 0.874).

Conclusion

Our study suggested the combination of improved total ethmoid sinus-maxillary score and blood eosinophil percentage is more accurate in predicting the diagnosis of ECRS.

A Systematic Literature Review of 3D Deep Learning Techniques in Computed Tomography Reconstruction

Computed tomography (CT) is used in a wide range of medical imaging diagnoses. However, the reconstruction of CT images from raw projection data is inherently complex and is subject to artifacts and noise, which compromises image quality and accuracy. In order to address these challenges, deep learning developments have the potential to improve the reconstruction of computed tomography images. In this regard, our research aim is to determine the techniques that are used for 3D deep learning in CT reconstruction and to identify the training and validation datasets that are accessible. This research was performed on five databases. After a careful assessment of each record based on the objective and scope of the study, we selected 60 research articles for this review. This systematic literature review revealed that convolutional neural networks (CNNs), 3D convolutional neural networks (3D CNNs), and deep learning reconstruction (DLR) were the most suitable deep learning algorithms for CT reconstruction. Additionally, two major datasets appropriate for training and developing deep learning systems were identified: 2016 NIH-AAPM-Mayo and MSCT. These datasets are important resources for the creation and assessment of CT reconstruction models. According to the results, 3D deep learning may increase the effectiveness of CT image reconstruction, boost image quality, and lower radiation exposure. By using these deep learning approaches, CT image reconstruction may be made more precise and effective, improving patient outcomes, diagnostic accuracy, and healthcare system productivity.

The efficacy of supervised learning and semi-supervised learning in diagnosis of impacted third molar on panoramic radiographs through artificial intelligence model

OBJECTIVES

The aim of the study was to evaluate the efficacy of traditional supervised learning (SL) and semi-supervised learning (SSL) in the classification of mandibular third molars (Mn3s) on panoramic images. The simplicity of preprocessing step and the outcome of the performance of SL and SSL were analyzed.

METHODS

Total 1625 Mn3s cropped images from 1000 panoramic images were labeled for classifications of the depth of impaction (D class), spatial relation with adjacent second molar (S class), and relationship with inferior alveolar nerve canal (N class). For the SL model, WideResNet (WRN) was applicated and for the SSL model, LaplaceNet (LN) was utilized.

RESULTS

In the WRN model, 300 labeled images for D and S classes, and 360 labeled images for N class were used for training and validation. In the LN model, only 40 labeled images for D, S, and N classes were used for learning. The F1 score were 0.87, 0.87, and 0.83 in WRN model, 0.84, 0.94, and 0.80 for D class, S class, and N class in the LN model, respectively.

CONCLUSIONS

These results confirmed that the LN model applied as SSL, even utilizing a small number of labeled images, demonstrated the satisfactory of the prediction accuracy similar to that of the WRN model as SL.

A Decade of Clinical Advances in Chronic Rhinosinusitis: 2012-2022

The field of chronic rhinosinusitis (CRS) is constantly evolving. In the past 10 years, key advancements in basic and translational research as well as clinical studies have improved our understanding and management of CRS. Notably, treatment options have expanded to include novel therapeutic drugs, devices, and surgical techniques. Assessments of patient symptoms and their impact on quality of life have become more standardized. Progress has also been made in both determining the true prevalence of CRS and recognizing comorbidities that can impact CRS severity. Practice guidelines have also shifted from expert opinion to more data-driven analyses. This review highlights major clinical advancements made in the field of CRS over the past 10 years as well as identifies current gaps in knowledge that can form the basis for new areas of study over the next decade.