Computers in biology and medicine · 2025
AI-driven rapid identification of bacterial and fungal pathogens in blood smears of septic patients.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Sroka-Oleksiak Agnieszka, Pardyl Adam, Rymarczyk Dawid, Olechowska-Jarząb Aldona, Biegun-Drożdż Katarzyna, Ochońska Dorota, Wronka Michał, Borowa Adriana, Gosiewski Tomasz, Adamczyk Miłosz, Telega Henryk, Zieliński Bartosz, Brzychczy-Włoch Monika
Аннотация
BACKGROUND: Sepsis is a life-threatening condition which requires rapid diagnosis and treatment. Traditional microbiological methods are time-consuming and expensive. In response to these challenges, deep learning algorithms were developed to identify 14 bacterial and 3 fungal species from microscopic images of Gram-stained smears of blood samples from sepsis patients.
METHOD: ology: A total of 16 637 Gram-stained microscopic images were used in the study. Novelty of our technological workflow lies in combination of the Cellpose 3 model for robust and precise segmentation of microbial cells, followed by Attention-based Deep Multiple Instance Learning for sophisticated and context-aware classification. This comprehensive approach allows effective analysis of complex image data, overcoming limitations of standard single-instance classification in heterogeneous samples.
RESULTS: Our model achieved high classification accuracy: 77.15 % for bacteria and 71.39 % for fungi, with ROC AUC values of 0.97 and 0.88, respectively. Most accurate results, reaching up to 96.2 %, were obtained for: Cutibacterium acnes, Enterococcus faecium, Stenotrophomonas maltophilia, and Nakaseomyces glabratus. We observed some classification difficulties in closely related species, such as Staphylococcus hominis and Staphylococcus haemolyticus, and within Candida albicans.
KEY CONTRIBUTIONS: Despite these challenges, the study confirms the model's significant potential. Our approach achieves microscopy image-based diagnosis while reducing the microbial diagnostic time. By identifying specific bacteria or fungal infection species, this technique improves sepsis diagnosis and its subsequent treatment. This is a simple and accessible alternative to current laborious methods and achieves high performance in species-specific classification. These results indicate the benefits of using artificial intelligence (AI) in the clinical diagnosis of sepsis.
METHOD: ology: A total of 16 637 Gram-stained microscopic images were used in the study. Novelty of our technological workflow lies in combination of the Cellpose 3 model for robust and precise segmentation of microbial cells, followed by Attention-based Deep Multiple Instance Learning for sophisticated and context-aware classification. This comprehensive approach allows effective analysis of complex image data, overcoming limitations of standard single-instance classification in heterogeneous samples.
RESULTS: Our model achieved high classification accuracy: 77.15 % for bacteria and 71.39 % for fungi, with ROC AUC values of 0.97 and 0.88, respectively. Most accurate results, reaching up to 96.2 %, were obtained for: Cutibacterium acnes, Enterococcus faecium, Stenotrophomonas maltophilia, and Nakaseomyces glabratus. We observed some classification difficulties in closely related species, such as Staphylococcus hominis and Staphylococcus haemolyticus, and within Candida albicans.
KEY CONTRIBUTIONS: Despite these challenges, the study confirms the model's significant potential. Our approach achieves microscopy image-based diagnosis while reducing the microbial diagnostic time. By identifying specific bacteria or fungal infection species, this technique improves sepsis diagnosis and its subsequent treatment. This is a simple and accessible alternative to current laborious methods and achieves high performance in species-specific classification. These results indicate the benefits of using artificial intelligence (AI) in the clinical diagnosis of sepsis.