Multi-instance learning based artificial intelligence model to assist vocal fold leukoplakia diagnosis: A multicentre diagnostic study.

OBJECTIVE: To develop a multi-instance learning (MIL) based artificial intelligence (AI)-assisted diagnosis models by using laryngoscopic images to differentiate benign and malignant vocal fold leukoplakia (VFL). METHODS: The AI system was developed, trained and validated on 5362 images of 551 patients from three hospitals. Automated regions of interest (ROI) segmentation algorithm was utilized to construct image-level features. MIL was used to fusion image level results to patient level features, then the extracted features were modeled by seven machine learning algorithms. Finally, we evaluated the image level and patient level results. Additionally, 50 videos of VFL were prospectively gathered to assess the system's real-time diagnostic capabilities. A human-machine comparison database was also constructed to compare the diagnostic performance of otolaryngologists with and without AI assistance. RESULTS: In internal and external validation sets, the maximum area under the curve (AUC) for image level segmentation models was 0.775 (95 % CI 0.740-0.811) and 0.720 (95 % CI 0.684-0.756), respectively. Utilizing a MIL-based fusion strategy, the AUC at the patient level increased to 0.869 (95 % CI 0.798-0.940) and 0.851 (95 % CI 0.756-0.945). For real-time video diagnosis, the maximum AUC at the patient level reached 0.850 (95 % CI, 0.743-0.957). With AI assistance, the AUC improved from 0.720 (95 % CI 0.682-0.755) to 0.808 (95 % CI 0.775-0.839) for senior otolaryngologists and from 0.647 (95 % CI 0.608-0.686) to 0.807 (95 % CI 0.773-0.837) for junior otolaryngologists. CONCLUSIONS: The MIL based AI-assisted diagnosis system can significantly improve the diagnostic performance of otolaryngologists for VFL and help to make proper clinical decisions.

Multi-Instance Learning for Vocal Fold Leukoplakia Diagnosis Using White Light and Narrow-Band Imaging: A Multicenter Study.

OBJECTIVES: Vocal fold leukoplakia (VFL) is a precancerous lesion of laryngeal cancer, and its endoscopic diagnosis poses challenges. We aim to develop an artificial intelligence (AI) model using white light imaging (WLI) and narrow-band imaging (NBI) to distinguish benign from malignant VFL. METHODS: A total of 7057 images from 426 patients were used for model development and internal validation. Additionally, 1617 images from two other hospitals were used for model external validation. Modeling learning based on WLI and NBI modalities was conducted using deep learning combined with a multi-instance learning approach (MIL). Furthermore, 50 prospectively collected videos were used to evaluate real-time model performance. A human-machine comparison involving 100 patients and 12 laryngologists assessed the real-world effectiveness of the model. RESULTS: The model achieved the highest area under the receiver operating characteristic curve (AUC) values of 0.868 and 0.884 in the internal and external validation sets, respectively. AUC in the video validation set was 0.825 (95% CI: 0.704-0.946). In the human-machine comparison, AI significantly improved AUC and accuracy for all laryngologists (p < 0.05). With the assistance of AI, the diagnostic abilities and consistency of all laryngologists improved. CONCLUSIONS: Our multicenter study developed an effective AI model using MIL and fusion of WLI and NBI images for VFL diagnosis, particularly aiding junior laryngologists. However, further optimization and validation are necessary to fully assess its potential impact in clinical settings. LEVEL OF EVIDENCE: 3 Laryngoscope, 2024.

Pan-cancer analysis of the prognostic and immunological role of GJB2: a potential target for survival and immunotherapy.

Background: GJB2 plays an essential role in the growth and progression of several cancers. However, asystematic pan-cancer analysis of GJB2 is lacking. Therefore, in this study, we performed a comprehensive pan-cancer analysis to determine the potential role of GJB2 in prognostic prediction and cancer immunotherapy response. Methods: The differential expression of GJB2 in the tumor and adjacent normal tissues of various cancer types was analyzed using the TIMER, GEPIA, and Sangerbox databases. GEPIA and Kaplan-Meier plotter databases were used to analyze the survival outcomes based on GJB2 expression levels in pan-cancer. Furthermore, the association of GJB2 expression with the immune checkpoint (ICP) genes, tumor mutational load (TMB), microsatellite instability (MSI), neoantigens, and tumor infiltration of immune cells was analyzed using via the Sangerbox database. The cBioPortal database was used to determine the characteristics of GJB2 gene alterations in the cancer tissues. The STRING database was used to identify the GJB2-binding proteins. GEPIA database was used to identify the GJB2 co-expressed genes. DAVID was used to perform the functional enrichment analysis of gene ontology (GO) terms and KEGG pathways associated with GJB2. Finally, the mechanistic role of GJB2 in pancreatic adenocarcinoma (PAAD) was analyzed using the LinkedOmics database. Results: The GJB2 gene was highly expressed in a variety of tumors. Furthermore, GJB2 expression levels showed significant positive or negative association with the survival outcomes in various cancers. GJB2 expression levels cor related with tumor mutational burden, microsatellite instability, neoantigens, and tumor infiltration of immune cells in multiple cancers. This suggested that GJB2 played a critical role in the tumor microenvironment. Functional enrichment analysis showed that the biological role of GJB2 in tumors included modulation of gap junction-mediated intercellular transport, regulation of cell communication by electrical coupling, ion transmembrane transport, autocrine signaling, apoptotic signaling pathway, NOD-like receptor signaling pathway, p53 signaling pathway, and PI3K-Akt signaling pathway. Conclusions: Our study demonstrated that GJB2 played a significant role in tumorigenesis and tumor immunity in multiple cancers. Furthermore, GJB2 is a potential prognostic biomarker and a promising therapeutic target in multiple types of cancers.

Efficacy and safety of triamcinolone acetonide in the prevention of esophageal stricture after endoscopic submucosal dissection: a meta-analysis.

AIM: The role of triamcinolone acetonide (TA) in the prevention of esophageal stricture is not well established. This meta-analysis aimed to evaluate its safety and efficacy for the prevention of esophageal stricture after endoscopic submucosal dissection (ESD). METHODS: A comprehensive search was performed in electronic databases including PubMed, the Cochrane Library, Embase for possible controlled studies. The primary outcomes were stenosis rate and endoscopic balloon dilatation (EBD) sessions required, and secondary outcome included complications. Random effects were used to calculate the pooled outcome. Sensitivity analysis and publication bias were conducted to verify the robustness and reliability of the results. Results: Ten studies containing 499 patients were obtained. In the pooled analysis, statistical significance was found in triamcinolone acetonide injection reduced the incidence of stenosis (OR = 0.29, 95% CI [0.11, 0.80], P < 0.05) and the number of endoscopic balloon dilation (MD = -3.33, 95% CI [-4.15, -2.50], P < 0.0001) compared with control. Triamcinolone acetonide injection therapy did not increase the risk of complications (OR = -0.77%, CI [-1.62, 0.09], P = 0.08). Subgroup analysis indicated that the single injection of triamcinolone acetonide after endoscopic submucosal dissection significantly reduced the incidence of stenosis compared with without any prophylaxis. Different concentrations and single session volume of triamcinolone acetonide reduced the incidence of stenosis. It also showed that the dose according to the size of the lesion was more effective than the fixed dose in preventing esophageal stricture. Conclusion: Triamcinolone acetonide injection can reduce the incidence of stricture formation as well as the need for EBD sessions without increasing complications.

Submucosal oesophageal squamous cell carcinoma with lymph node metastasis: a case report and literature review.

BACKGROUND: Submucosal oesophageal squamous cell carcinoma is a quite infrequent and special type of oesophageal cancer. Its endoscopic manifestations are similar to those of submucosal oesophageal lesions, so it is easily ignored or misdiagnosed. Thus, the exact and timely diagnosis of oesophageal subepithelial lesions (SELs) is crucial. Endoscopic submucosal dissection (ESD) improves the diagnosis rate of malignant SELs without specific endoscopic presentations. CASE PRESENTATION: We report a 63-year-old patient with submucosal lesions of the oesophagus under endoscopy, but CT suggested mediastinal lymphadenectasis. Thus, there was a contradiction between them. After multidisciplinary consultation, endoscopic submucosal dissection (ESD) resection was finally recommended. The lesion was completely resected by endoscopic submucosal dissection. Postoperative pathology reported poorly differentiated squamous cell carcinoma, and subsequent PET-CT examination provided clarity, revealing mediastinal lymph node metastasis. CONCLUSIONS: Not all oesophageal SELs are benign, and a small number of SELs can be malignant. Submucosal oesophageal squamous cell carcinoma is a rare disease that may be characterized by oesophageal subepithelial lesions (SELs). Therefore, the precise and timely diagnosis of SELs is essential. If it is necessary to obtain lesion tissue for a definite diagnosis, ESD with less invasiveness is an excellent choice.

Efficacy of EGFR-TKI sequential therapy in patients with EGFR exon 19 insertion-positive non-small-cell lung cancer: A case report.

BACKGROUND: Insertions in exon 19 in the epidermal growth factor receptor gene (EGFR) is a rarely seen mutation in non-small cell lung cancer. These patients have been effectively treated with sequential EGFR tyrosine kinase inhibitors (TKIs). CASE SUMMARY: Here, we presented a case of non-small cell lung cancer, stage IIIB, with EGFR exon 19 insertion mutation as detected in the right lower lobe by next-generation sequencing. The patient was sequentially treated with first, second, and third-generation EGFR TKIs after the surgical operation. The overall survival of the patient was 21.3 mo. There was no dynamic analysis of drug resistance mechanisms in targeted therapy. CONCLUSION: This case emphasized the importance of following the guidelines. In patients with EGFR mutations, repeated and dynamic next-generation sequencing monitoring is necessary to prescribe a personalized treatment plan.

Synchronous multiple carcinoma with small intestine and pulmonary neuroendocrine involvement: A case report.

RATIONALE: In clinical work, neuroendocrine synchronous multiplicity carcinoma was relatively rare. Most were confirmed by the pathological diagnosis of a certain part of the body combined with the imaging of the whole body, while cases that had both pathological and immunohistochemistry diagnosis were few. PATIENT CONCERNS: A patient who presented with abdominal pain visited our hospital, and was diagnosed with lesions in both the small intestine and lung. DIAGNOSES: Both were considered primary tumors by imaging, and diagnosed as neuroendocrine carcinomas by pathology. INTERVENTIONS: The intestinal lesion was surgically resected, and the lung tumor treated by chemoradiotherapy. OUTCOMES: The survival time of this patient exceeded 24 months. LESSONS: The diagnosis relied on clinical, imaging, pathological, and immunohistochemical features, which confirmed a synchronous multiple carcinoma. Treatment was based on the pathological types. Through this case report, the clinical and pathological data of neuroendocrine synchronous multiplicity carcinoma could be enriched.