Extracting interpretable features for pathologists using weakly supervised learning to predict p16 expression in oropharyngeal cancer.

One drawback of existing artificial intelligence (AI)-based histopathological prediction models is the lack of interpretability. The objective of this study is to extract p16-positive oropharyngeal squamous cell carcinoma (OPSCC) features in a form that can be interpreted by pathologists using AI model. We constructed a model for predicting p16 expression using a dataset of whole-slide images from 114 OPSCC biopsy cases. We used the clustering-constrained attention-based multiple-instance learning (CLAM) model, a weakly supervised learning approach. To improve performance, we incorporated tumor annotation into the model (Annot-CLAM) and achieved the mean area under the receiver operating characteristic curve of 0.905. Utilizing the image patches on which the model focused, we examined the features of model interest via histopathologic morphological analysis and cycle-consistent adversarial network (CycleGAN) image translation. The histopathologic morphological analysis evaluated the histopathological characteristics of image patches, revealing significant differences in the numbers of nuclei, the perimeters of the nuclei, and the intercellular bridges between p16-negative and p16-positive image patches. By using the CycleGAN-converted images, we confirmed that the sizes and densities of nuclei are significantly converted. This novel approach improves interpretability in histopathological morphology-based AI models and contributes to the advancement of clinically valuable histopathological morphological features.

Survival times with and without tube feeding in patients with dementia or psychiatric diseases in Japan.

BACKGROUND: It is widely supposed that there has been no evidence of increased survival in patients with advanced dementia receiving enteral tube feeding. However, more than a few studies have reported no harmful outcome from tube feeding in dementia patients compared to in patients without dementia. METHODS: This was a retrospective study. Nine psychiatric hospitals in Okayama Prefecture participated in this survey. All inpatients fulfilling the entry criteria were evaluated. All subjects suffered from difficulty with oral intake. Attending physicians thought that the patients could not live without long-term artificial nutrition. The physicians decided whether to make use of long-term artificial nutrition between January 2012 and December 2014. RESULTS: We evaluated 185 patients. Their mean age was 76.6 ± 11.4 years. Of all subjects, patients with probable Alzheimer's disease (n = 78) formed the biggest group, schizophrenia patients (n = 44) the second, and those with vascular dementia (n = 30) the third. The median survival times were 711 days for patients with tube feeding and 61 days for patients without tube feeding. In a comparison different types of tube feeding, median survival times were 611 days for patients with a nasogastric tube and more than 1000 days for those with a percutaneous endoscopic gastrostomy tube. CONCLUSION: Patients with tube feeding survived longer than those without tube feeding, even among dementia patients. This study suggests that enteral nutrition for patients with dementia prolongs survival. Additionally, percutaneous endoscopic gastrostomy tube feeding may be safer than nasogastric tube feeding among patients in psychiatric hospitals.

Tube feeding decreases pneumonia rate in patients with severe dementia: comparison between pre- and post-intervention.

BACKGROUND: It is widely supposed that there is no benefit, including extended survival and decreased rate of pneumonia, in patients with severe dementia receiving enteral tube feeding (TF). However, there have been few studies comparing the frequency of pneumonia before and after TF in severe dementia. METHODS: Nine psychiatric hospitals in Okayama Prefecture participated in this retrospective survey. All inpatients fulfilling the entry criteria were evaluated. All subjects suffered from difficulty in oral intake. Attending physicians thought that the patients could not live without long-term artificial nutrition, and they decided whether or not to make use of long-term artificial nutrition from January 1, 2014 to December 31, 2014. RESULTS: We evaluated 58 patients including 46 with TF and 12 without. The mean age of all patients was 79.6 ± 9.0 years old. Patients with probable Alzheimer's disease (n = 38) formed the biggest group, and those with vascular dementia the second (n = 14). Median survival times were 23 months among patients with TF and two months among patients without TF. The start of TF decreased the frequency of pneumonia and the use of intravenous antibiotics. CONCLUSIONS: TF decreased pneumonia and antibiotic use, even in patients with severe dementia. The results of this study do not necessarily indicate that we should administer TF to patients with severe dementia. We should consider the quality of life of patients carefully before deciding the use or disuse of TF for patients with severe dementia.

Heterogeneity of patients receiving artificial nutrition in Japanese psychiatric hospitals: a cross-sectional study.

AIM: Artificial nutrition, including tube feeding, continues to be given to dementia patients in numerous geriatric facilities in Japan. However, the clinical characteristics of patients receiving artificial nutrition have not been fully investigated. Therefore, we tried to evaluate the clinical features of those patients in this study. METHODS: Various clinical characteristics of all inpatients at 18 of 20 psychiatric hospitals in Okayama Prefecture, Japan, with a percutaneous endoscopic gastrostomy tube, nasogastric tube, or total parenteral nutrition were evaluated. RESULTS: Two hundred twenty-one patients (5.4% of all inpatients) had been receiving artificial nutrition for more than 1 month, and 187 (130 women, 57 men; 84.6% of 221 patients) were fully investigated. The mean age was 78.3 years old, and the mean duration of artificial nutrition was 29.8 months. Eighty-four patients (44.7% of 187 patients) were receiving artificial nutrition for more than 2 years. Patients with Alzheimer's disease (n = 78) formed the biggest group, schizophrenia (n = 37) the second, and vascular dementia (n = 26) the third. CONCLUSION: About one-fifth of the subjects receiving artificial nutrition were in a vegetative state. More than a few patients with mental disorders, including schizophrenia, also received long-term artificial nutrition. We should pay more attention to chronic dysphasia syndrome in mental disorders.