Mayo Clinic Proceedings. Digital health最新文献

Objective

To identify geographic locations with high numbers of medically attended falls (ie, hotspots) by older adults and to test the associations between fall hotspots and resident/housing characteristics.

Patients and Methods

In this cohort study, we retrospectively reviewed adults who were 65 years or older, lived in a single-family detached dwelling, and had a medically attended fall in Olmsted County, MN, between April 1, 2012, and December 31, 2014. We identified medically attended falls by using billing codes and confirmed by manual review of the electronic health records. We performed geospatial analysis to identify fall hotspots and evaluated the association between fall hotspots and resident or housing characteristics with logistic regression models, adjusting for age, sex, socioeconomic status, chronic health conditions, and/or a history of falls.

Results

Among 12,888 residents living in single-family detached dwellings in our community, 587 residents (4.6%) had documented accidental falls. Falls were more common in older residents and in women. Residents who had more chronic diseases, lower socioeconomic status, and a history of falls also had higher odds of a fall. Geospatial analysis identified 2061 (16.0%) residents who lived in a fall hotspot. Houses in hotspots were more likely to have more stories with fewer stairs (split level) (odds ratio [OR], 1.75; 95% CI, 1.57-1.94, for split level vs 1-story houses), smaller square feet (OR, 0.29; 95% CI, 0.24-0.35, for largest vs smallest houses), and in the highest quartile for age (OR, 1.46; 95% CI, 1.26-1.70, for oldest built vs newest built houses).

Conclusion

Falls were more common in locations in our community that had older, smaller homes and lower housing-based socioeconomic status. These findings can be used by clinicians to identify residents who are at higher risk for falls.

Navigating clinical guidelines can be complex for real-time health care decision making. Our study evaluates the chat generative prerained transformer (ChatGPT)-4 in improving responses to clinical questions by integrating guidelines on Clostridioides difficile infection and colon polyp surveillance. We assessed ChatGPT-4’s responses to questions before and after guideline integration, noting a clear improvement in accuracy. ChatGPT-4 provided guideline-aligned answers consistently. Further analysis showed its ability to summarize information from conflicting guidelines, highlighting its utility in complex clinical scenarios. The findings suggest that large language models such as ChatGPT-4 can enhance clinical decision making and patient education by providing quick, conversational, and accurate responses. This approach opens a path for using artificial intelligence to deliver reliable responses in health care, supporting clinicians in real-time decision making and improving patient care.

Objective

To assess the performance of the current foundation models in histopathology.

Patients and Methods

The assessment involves a comprehensive evaluation of some foundation models, such as the CLIP derivatives, namely PLIP and BiomedCLIP, which were fine-tuned on data scraped from the internet. The comparison is performed against simpler and nonfoundational histology models that are trained on well-curated data, eg, the cancer genome atlas. All models are evaluated on 8 datasets, 4 of which are internal histology datasets collected and curated at Mayo Clinic, and 4 well-known public datasets: PANDA, BRACS, CAMELYON16, and DigestPath. Evaluation metrics include accuracy and macro-averaged F1 score, using a majority vote among top-k (eg, MV@5) at the whole slide image/patch levels. Moreover, all models are evaluated in classification settings. This detailed analysis allows for a deep understanding of each model’s performance across various datasets.

Results

In various evaluation tasks, domain-specific (and nonfoundational) models like DinoSSLPath and KimiaNet outperform general-purpose foundation models. The DinoSSLPath excels in whole slide image-level retrieval for internal colorectal cancer and liver datasets with MV@5 macro-averaged F1 scores of 63% and 74%, respectively. The KimiaNet leads in breast and skin cancer tasks with respective Top-1 and MV@5 scores of 56% and 70%, respectively and scores 75% on the public CAMELYON16 dataset. Similar trends are observed in patch-level metrics, highlighting the advantage of using specialized datasets like the cancer genome atlas for histopathological analysis.

Conclusion

To enable effective vision-language foundation models in biomedicine, high-quality, multi-modal medical datasets are essential. These datasets serve as the substrate for training models capable of translating research into clinical practice. Of importance, the alignment (correspondence) between textual and visual data—often diagnostic—is critical and requires validation by domain experts. Thus, advancing foundation models in this field necessitates collaborative efforts in data curation and validation.