Anatomical classification of pharyngeal and laryngeal endoscopic images using artificial intelligence.

BACKGROUND: The entire pharynx should be observed endoscopically to avoid missing pharyngeal lesions. An artificial intelligence (AI) model recognizing anatomical locations can help identify blind spots. We developed and evaluated an AI model classifying pharyngeal and laryngeal endoscopic locations. METHODS: The AI model was trained using 5382 endoscopic images, categorized into 15 anatomical locations, and evaluated using an independent dataset of 1110 images. The main outcomes were model accuracy, precision, recall, and F1-score. Moreover, we investigated focused regions in the input images contributing to the model predictions using gradient-weighted class activation mapping (Grad-CAM) and Guided Grad-CAM. RESULTS: Our AI model correctly classified pharyngeal and laryngeal images into 15 anatomical locations, with an accuracy of 93.3%. The weighted averages of precision, recall, and F1-score were 0.934, 0.933, and 0.933, respectively. CONCLUSION: Our AI model has an excellent performance determining pharyngeal and laryngeal anatomical locations, helping endoscopists notify of blind spots.

A multicenter case-control study of self-expanding metallic stent versus trans-anal colorectal tube for stage II/III non-right-sided obstructive colon cancer.

BACKGROUND: Self-expanding metallic stent (SEMS) and trans-anal colorectal tube (TCT) are alternative treatments to conventional emergency surgery for non-right-sided obstructive colon cancer (NROCC). However, the one with better short- and long-term outcomes remains controversial. Thus, this multicenter case-control study aimed to analyze and compare SEMS and TCT for NROCC. METHODS: Patients with stage II/III NROCC who underwent surgery between January 2010 and December 2019 at either of the eight selected Japanese affiliate hospitals were, retrospectively, reviewed. Baseline characteristics between the SEMS and TCT groups were adjusted by propensity score (PS) matching. RESULTS: Among 239 reviewed patients (SEMS: 76, TCT: 163), 180 were finally included in two well-balanced cohorts through PS: SEMS group (65 patients) and TCT group (115 patients). Technical success, clinical success, morbidity, and short-term mortality were not significantly different between the two groups. SEMS placement achieved significantly higher rates for primary resection/anastomosis without stoma (SEMS: 90.8% vs. TCT: 77.4%, p < 0.001) and laparoscopic surgery (SEMS: 64.6% vs. TCT: 43.5%, p < 0.001) than TCT placement. However, 5-year overall survival (SEMS: 83.7% vs. TCT: 86.4%; p = 0.822) and 5-year relapse-free survival (SEMS: 64.7% vs. TCT: 66.4%; p = 0.854) showed no significant differences between these groups. CONCLUSIONS: Both SEMS and TCT revealed similar long-term outcomes, but SEMS placement was better in achieving primary resection/anastomosis and laparoscopic surgery in patients with stage II/III NROCC.

Acute Acalculous Cholecystitis Caused by Giardia lamblia.

We report a case of a 70-year-old man with acute acalculous cholecystitis caused by Giardia lamblia. Contrast-enhanced computed tomography (CT) showed distention of the gallbladder due to a pericholecystic abscess without gallstones. Magnetic resonance cholangiopancreatography and drip infusion cholecystocholangiography-CT demonstrated a stricture of the hilar bile duct and cystic duct obstruction. We conducted transpapillary bile duct brush cytology and a biopsy of the hilar bile duct stricture; numerous active trophozoites of Giardia lamblia were observed without malignant findings. We considered this bile duct lesion to be biliary giardiasis. Biliary giardiasis should be taken into consideration when diagnosing acute acalculous cholecystitis.