Automatic maxillary sinus segmentation and pathology classification on cone-beam computed tomographic images using deep learning.

BACKGROUND: Maxillofacial complex automated segmentation could alternative traditional segmentation methods to increase the effectiveness of virtual workloads. The use of DL systems in the detection of maxillary sinus and pathologies will both facilitate the work of physicians and be a support mechanism before the planned surgeries. OBJECTIVE: The aim was to use a modified You Only Look Oncev5x (YOLOv5x) architecture with transfer learning capabilities to segment both maxillary sinuses and maxillary sinus diseases on Cone-Beam Computed Tomographic (CBCT) images. METHODS: Data set consists of 307 anonymised CBCT images of patients (173 women and 134 males) obtained from the radiology archive of the Department of Oral and Maxillofacial Radiology. Bilateral maxillary sinuses CBCT scans were used to identify mucous retention cysts (MRC), mucosal thickenings (MT), total and partial opacifications, and healthy maxillary sinuses without any radiological features. RESULTS: Recall, precision and F1 score values for total maxillary sinus segmentation were 1, 0.985 and 0.992, respectively; 1, 0.931 and 0.964 for healthy maxillary sinus segmentation; 0.858, 0.923 and 0.889 for MT segmentation; 0.977, 0.877 and 0.924 for MRC segmentation; 1, 0.942 and 0.970 for sinusitis segmentation. CONCLUSION: This study demonstrates that maxillary sinuses can be segmented, and maxillary sinus diseases can be accurately detected using the AI model.

Three-dimensional computed tomography analysis of frontoethmoidal cells: A critical evaluation of the International Frontal Sinus Anatomy Classification (IFAC).

OBJECTIVES: Two major classifications of frontoethmoidal cells, Lee and Kuhn and the IFAC (International Frontal Sinus Anatomy Classification), distinguish anterior, posterior and medial cells. The European anatomical position paper includes also lateral cells. According to the IFAC, anterior cells push the frontal sinus drainage pathway (FSDP) medially, posteriorly or posteromedially. Posterior cells push the FSDP anteriorly. The only medial cell, pushing the FSDP laterally is the frontal septal cell, which is attached to or located in the interfrontal sinus septum. The aim of this study was to verify the IFAC and characterise cells, which are inconsistent with this classification. DESIGN: A radioanatomic analysis. SETTING: Tertiary university hospital. PARTICIPANTS: One hundred and three Caucasian adult patients with no inflammatory changes in paranasal sinuses CT. MAIN OUTCOMES MEASURE: Results of assessment of multiplanar reconstruction of thin slice CT. RESULTS: Two types of cells that cannot be classified using the IFAC were found: (a) Lateral cells extending between the skull base and the anterior buttress, pushing the FSDP anteromedially or medially, present in 34 (16.5%) of the sides, (b) Paramedian cells: medially based, not adjacent to the interfrontal septum, pushing the FSDP laterally and posteriorly, present in 33 (16%) of sides. Suprabulla cells and suprabulla frontal cells were found to push the FSDP in directions other than anterior 28% and 31% of the time respectively. CONCLUSIONS: Neglecting lateral and paramedian cells may lead to inconsistent results between radioanatomical studies and impede communication between surgeons. They should be included in existing classifications of frontoethmoidal cells.

Chronic maxillary atelectasis and silent sinus syndrome: two faces of the same clinical entity.

Chronic maxillary atelectasis (CMA) and silent sinus syndrome (SSS) are rare clinical entities characterized by an implosion of the maxillary sinus that may or may not be associated with sinonasal symptoms, and are complicated by ipsilateral enophthalmos. The objective of this article is to discuss the definitions, physiopathology, clinical and radiographic characteristics, and surgical management of these entities. We retrospectively reviewed 18 patients (7 women, 11 men, aged 12-70 years) diagnosed and treated in the ear, nose, and throat departments of four Belgian teaching hospitals between 2000 and 2015. Nine patients had a history of sinus disease. In all cases, a computed tomography scan showed downward displacement of the orbital floor, increased orbital volume, and maxillary sinus contraction. Five patients met the criteria for grade II CMA and 13 for grade III CMA. Four patients met the criteria for SSS. All patients underwent wide endoscopic middle maxillary antrostomy. There were no orbital complications and all patients experienced resolution or a dramatic reduction of their symptomatology. Only one patient asked for an orbital floor reconstruction to correct a persisting cosmetic deformity. Although CMA and SSS are usually regarded as different entities in the literature, we believe that they lie on the same clinical spectrum. Treatment for both conditions is similar, i.e., middle meatal antrostomy to halt or even reverse the pathological evolution and reconstruction of the orbital floor in the event of persistent cosmetic deformity.

URGENT X-RAY DIAGNOSTICS OF INTRACRANIAL PYOINFLAMMATORY OTORHINOSINUSOGENIC COMPLICATIONS.

An analysis of the results of X-ray CT and MR-imaging was made in 46 patients aged from 13 to 78 years old. The patients were admitted to multifield hospitals in Kursk at the period from 2005 to 2015. The research included the nasal cavity, paranasal sinuses, mastoid and pyramid of the temporal bones and the brain. The study could be repeated with bolus contrast medium infusion. The condition of the patients was evaluated in dynamics at intervals of 5-7 days and these data was associated with clinical picture. The authors presents a complex of symptoms and an algorithm of differentiated X-ray diagnostics of diseases of the ENT organs and the main nosological forms of pyoinflammatory diseases of arachnoid membrane and substances of the brain.

Frontal sinus mucocoeles: new algorithm for surgical management.

BACKGROUND: Frontal sinus mucocoeles require a structured approach to their surgical management. We share our experience of a novel method of positional classification for frontal mucocoele and corresponding surgical algorithm. METHODS: A retrospective case-note review examined all frontal sinus surgery for mucocoele, spanning three years (2008-2010). Patients had pre-operative nasendoscopy, multi-planar CT and MRI when indicated. Several important variables (position, drainage dimensions, fronto-ethmoidal cells and degree of neo-osteogenesis) were noted. A systematic algorithm was used for surgical drainage based on these characteristics. The three year outcomes using this approach are presented. RESULTS: Thirty-six patients were identified with a total of 43 frontal mucocoeles. Using our classification, 30 mucocoeles were medial; seven were intermediate; six were lateral. Thirty-four patients underwent a primary endoscopic procedure; six required a combined primary osteoplastic-flap (OPF) and endoscopic approach. Six patients required revision surgery for polypoidal change or neo-ostium stenosis. All patients were eventually rendered asymptomatic. CONCLUSION: Implementation of our positional classification and surgical algorithm was successful with a revision rate of 19%. Presence of frontal sinus wall dehiscence and extra-sinus mucocoele extension are invalid indicators for external approach. We feel our classification and treatment algorithm, with its associated indicators for surgical escalation (i.e. limited dimensions of frontal ostium, presence of Type III / IV front-ethmoidal cells etc), are applicable for future management of frontal mucocoeles.

Sinus problems as a cause of headache refractoriness and migraine chronification.

Sinus headache is not a diagnostic term supported by the academia, yet it appears to be understood by the general public and larger medical community. It can be considered both a primary and secondary headache disorder. As a primary headache disorder, most of the patients considered to have sinus headache indeed have migraine (migraine with sinus symptoms). Yet it is also possible that some attacks of sinus headache may represent a unique clinical phenotype of migraine or be a unique clinical entity. Potentially, primary sinus headache can chronify and be refractory through immune-mediated mechanisms or as a catalyst for migraine chronification through ineffective treatment or medication overuse and misuse. As a secondary headache disorder, sinus headache can be associated with a wide range of underlying etiologies such as infection, anatomical abnormalities, trauma, and immunological disease or sleep disorders. It is possible that these underlying pathophysiological processes generate long-standing activation of nociceptive mechanisms involved in headache and can lead to chronification and refractoriness of the headache symptomatology. This article explores some of the potential mechanisms and the available scientific studies that may explain how sinus headache can become chronic and present to the clinician as a refractory headache disorder.

The intranasal ethmoidectomy: an experience with 1,077 procedures.

A series of 1,077 intranasal ethmoidectomies (825 with sphenoid sinusotomies) was performed in 600 patients over a 15-year period at The Mount Sinai Medical Center. The technique is a modification of the classical operation originally proposed by Yankauer. The rate of significant complications was 1.1%. A subset of 90 patients underwent 166 procedures and were followed an average of 3.5 years. The patients were analyzed according to whether the disease was focal or diffuse, infectious or polypoid, and whether asthma was present. The surgical success rate was 88% in nonasthmatics, but dropped to 50% in asthmatic patients despite total sphenoethmoidectomy. This underscores the importance of this condition as a biological modifier of surgical prognosis. Accordingly, a system of classification of sinus diseases is proposed based upon disease extent and type and whether asthma is present.

Sarcoidosis of the sinonasal tract: a new staging system.

Sarcoidosis is a chronic multisystem granulomatous disease that has a predilection for pulmonary and upper respiratory tract involvement. Because the initial signs and symptoms of sarcoidosis may be identical to those of other forms of chronic sinonasal inflammatory disease, these patients will often first seek treatment from an otolaryngologist. We present a series of 28 patients whose primary symptoms was involvement of a sinonasal tract. A new staging system is proposed to categorize the severity and sites of involvement and to guide the aggressiveness of therapy. Sarcoidosis should be considered in the differential diagnosis of inflammatory sinonasal disease.

Fungal diseases of the paranasal sinuses.

Fungal diseases of the paranasal sinuses can be categorized into the invasive and noninvasive varieties. The invasive form has been classified as acute fulminant fungal sinusitis, granulomatous invasive fungal sinusitis, and chronic invasive fungal sinusitis. The noninvasive form can be classified as the fungus ball and the allergic fungal sinusitis. The following review discusses the various types of fungal sinusitis with a special emphasis on the CT and MR imaging features.

The effectiveness of guidelines in reducing inappropriate CT scans of the paranasal sinuses.

An effective system for scoring pathological changes on CT scans of the paranasal sinuses has been developed by Lund & Mackay. We have performed an audit using 100 outpatients with nasal symptoms and found that adherence to guidelines prior to ordering CT scans of the paranasal sinuses correlates with an increased average Lund & Mackay score. Using these guidelines has also reduced the number of inappropriate CT scan requests.

Fungal infections of the paranasal sinuses.

Fungal infections can be mainly grouped into four types. The invasive forms are acute sinusitis (fulminant), chronic sinusitis (indolent), whereas the non-invasive forms are mycetoma and allergic fungal sinusitis. From December 1993 to December 1997, 27 cases of fungal sinusitis, 22 of which were noninvasive forms, and 5 of which were invasive forms, were treated and are presented in this study. When we classified the patients with fungal sinusitis, 11 were diagnosed as mycetoma, 9 as allergic fungal sinusitis, 3 as acute fulminant sinusitis and 2 as chronic indolent sinusitis, while 2 patients were not included in our four groups of sinusitis. In all mycetoma cases the active agent was Aspergillus. Patients with non invasive forms of sinusitis were all treated with endoscopic sinus surgery. 2 of the patients with invasive forms of sinusitis underwent maxillectomy and they were given Amphotericin-B. With a mean follow up of 20 months, only 3 recurrences were seen. The infection recurred in 2 patients with allergic fungal sinusitis and 1 patient with chronic invasive sinusitis. However, 2 patients with acute fulminant invasive sinusitis died before they were operated on, and 1 patient died postoperatively.

Sinonasal sarcoidosis: A new system of classification acting as a guide to diagnosis and treatment.

BACKGROUND: The signs and symptoms of sinonasal sarcoidosis are diverse and nonspecific. It easily mimics more common sinonasal disorders such as allergic rhinitis, bacterial sinusitis, and atrophic rhinitis. The purpose of this study was to develop a classification system of sinonasal sarcoidosis that will help serve as a guide for both diagnosis and treatment. METHODS: A retrospective chart review was performed of all patients diagnosed with sinonasal sarcoidosis from 1974 to 2013. A PubMed literature review of all published case series of sinonasal sarcoidosis was also reviewed. RESULTS: The charts of 14 patients with biopsy-proven sinonasal sarcoidosis were reviewed. Follow-up time ranged from 6 months to 28 years. Presenting signs included hypertrophic mucosa, nasal polyps, crusting, nasal enlargement, and destructive processes. Eight patients underwent medical management alone and six patients underwent endoscopic sinus surgery. Five of the six patients were successfully treated with surgery. On review of the literature, 256 cases of sinonasal sarcoidosis were found, of which 43 patients underwent surgery. Clinical outcomes of the patients who underwent surgery were not consistently reported. CONCLUSION: Based on critical analysis of the clinical presentation of sinonasal sarcoidosis in our case series and on review of the literature, sinonasal sarcoidosis can be classified into four subgroups: atrophic, hypertrophic, destructive, and nasal enlargement. Each subgroup responds differently to treatment and has its own differential diagnosis. Surgery is only indicated for a select group of patients and the vast majority of patients benefit from medical management alone.

Sinonasal symptoms in adults with ear disease.

BACKGROUND: The association between eustachian tube dysfunction and middle-ear effusion is well established. Studies have also demonstrated pathological changes affecting the middle-ear mucosa associated with chronic sinonasal inflammation. No previous studies have evaluated symptoms related to sinonasal inflammatory disease in different ear diseases. OBJECTIVE: To assess the presence of sinonasal symptoms in ear diseases using the Dundee Rhinogram. METHODS: Data were collected prospectively in the period February-October 2011. Sinonasal symptoms were graded using the Dundee Rhinogram. Student's t-test analyses were performed to identify any statistically significant associations. RESULTS: In total, 164 patients were assessed. There was a statistically significant association between sinonasal symptoms and mucosal middle-ear diseases (p < 0.005). The mean sinonasal symptoms score for mucosal middle-ear disease patients was 5.94 (range, 0-32). CONCLUSION: Assessment of sinonasal symptoms is paramount in patients presenting with an ear symptom; inflammatory sinonasal disease treatment may become necessary in the management of middle-ear mucosal disease for better patient outcome.

Diagnostic and staging accuracy of magnetic resonance imaging for the assessment of sinonasal disease.

BACKGROUND: Concerns over radiation exposure with paranasal sinus computed tomography (CT) have recently evolved. Magnetic resonance imaging (MRI) may be more advantageous in this regard, but the diagnostic assessment of MRI in chronic rhinosinusitis (CRS) has not been reported. The purpose of this study was to determine the correlation between CT- and MRI-based staging and diagnosis of CRS. METHODS: Paired CT and MRI scans of 89 adult patients who were imaged by both modalities within a 3-month time period for evaluation of pituitary disease were scored for sinus disease using the Lund-Mackay system in a randomized and blinded fashion. The Lund scores were compared for similarity, correlation, and diagnostic classification between modalities. RESULTS: The mean Lund scores were 2.3 +/- 0.6 (95% CI) for CT-based staging and 2.1 +/- 0.5 for MRI-based staging with a median time interval between scans of 3 days. The difference means was not statistically significant (p = 0.444, paired t-test). Correlation analysis revealed a significant association between CT- and MRI-based scores (Pearson's r = 0.837, p < 0.001). Disease classification agreement analysis using published Lund score cutoffs (3 versus 4) for the likelihood of true sinus disease revealed that CT- and MRI-based scoring agreed on 76 cases (85.4%). Disagreement occurred in 13 cases (kappa: 0.557, p < 0.001). Sensitivity, specificity, positive predictive value, and negative predictive value were 66.7, 90.1, 63.2, and 91.4%, respectively. CONCLUSION: Lund-Mackay staging of sinus disease by MRI is closely correlated to corresponding staging based on CT. MRI does not significantly overstage or overclassify patients with sinus disease.

Destructive noninvasive paranasal sinus aspergillosis: component of a spectrum of disease.

A chronic, destructive form of paranasal sinus aspergillosis may exist without evidence of tissue fungal invasion. In this state, the pathogen results in progressive, chronic inflammation intermediate between previously described saprophytic and invasive states. We report three such cases. This variety of aspergillosis and an analysis of the clinical and histopathologic findings of previously reported cases support the concept that infection with the organism may produce a spectrum of pathologic changes and is akin to the range of aspergillosis disease states described affecting the lung. We classify paranasal and nasal aspergillosis as (1) noninvasive, either an aspergilloma or allergic type, (2) destructive, noninvasive, and (3) invasive, either slowly progressive or fulminant. Any form may progress to, or be associated with, more aggressive disease.

Mucosal cysts of the maxillary sinus: a literature review.

The aim of this article is to review the clinical, radiographic and microscopic features, as well as the etiopathogenesis, the differential diagnosis and the treatment of mucosal cysts of the maxillary sinus.

Genotype-phenotype correlations for the paranasal sinuses in cystic fibrosis.

Genotype-phenotype correlations in cystic fibrosis (CF) have been found for lung and pancreatic function, but not for paranasal sinus disease. Because such correlations may have pathophysiological and clinical implications, the correlation of mutations, in particular DeltaF508, with paranasal sinus disease was investigated in 113 CF patients with known genotype. The clinical importance of paranasal sinus disease was evaluated using three parameters: polyps, overall clinical severity of upper airway problems, and surgery. Polyps were evaluated by nasal endoscopy and graded on a five-point scale. Four severity groups were distinguished based on history, clinical records, and examination: no upper airway problems; more problems than in control subjects; severe, recurrent or chronic problems; and paranasal sinus surgery cases. DeltaF508 homozygosity correlated with clinical severity (p < 0.02) and with the presence of polyps on endoscopy (p < 0.05). The relative risk for paranasal sinus surgery in DeltaF508 homozygous CF patients was 2.33. In conclusion, there are genotype-phenotype correlations for paranasal sinus disease in CF. DeltaF508 homozygosity is a risk factor for paranasal sinus disease in CF.

The CT classification of ethmoid mucoceles.

The computed tomography (CT) appearance of mucoceles involving the ethmoid sinuses is presented. A classification based on the growth patterns of these mucoceles is proposed, and a differential diagnosis for each group is suggested. The unique CT appearance of the ethmoid polypoid--mucocele is described for the first time, and the radiographic definition of a mucocele is discussed.