Journal of virological methods · 2026
Artificial intelligence-assisted technology to reduce turnaround time for rapid diagnosis of infectious diseases.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
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.
Zeng Qihang, Liu Lunjing, Wang Tong, Zhang Chengli, Wang Jingru, Fang Liu, Yang Delong, Fu Yu
Аннотация
BACKGROUND: Quantitative polymerase chain reaction (qPCR) is a fundamental tool for disease detection; however, the relatively large number of amplification cycles can restrict its responsiveness in time-critical situations such as infectious disease outbreaks. Artificial intelligence (AI) offers the potential to enable earlier interpretation of amplification curves and thereby accelerate qPCR-based diagnosis, particularly for COVID-19.
METHODS: A quick qPCR (q-qPCR) algorithm based on a bidirectional long short-term memory (BiLSTM) network is introduced to predict fluorescence signals from early-cycle qPCR measurements. The model was trained and validated using datasets derived from standard low-concentration COVID-19 detection kit samples. Comparative experiments were carried out against recurrent neural network (RNN) and LSTM baseline models.
RESULTS: The BiLSTM-based model generated fluorescence predictions that closely aligned with experimentally measured values and demonstrated the best performance among the tested models. Importantly, the model enabled early qualitative classification using only 27 amplification cycles, compared with the conventional 45 cycles, corresponding to a 40% reduction in required cycles. The classification accuracy on the test dataset reached about 92%.
CONCLUSIONS: The proposed q-qPCR algorithm can be incorporated into conventional qPCR workflows and may contribute to shortening assay runtime and improving sample turnaround time (TAT). By enabling earlier preliminary interpretation, this approach could support more timely triage and diagnostic response in time-sensitive infectious disease testing scenarios.
METHODS: A quick qPCR (q-qPCR) algorithm based on a bidirectional long short-term memory (BiLSTM) network is introduced to predict fluorescence signals from early-cycle qPCR measurements. The model was trained and validated using datasets derived from standard low-concentration COVID-19 detection kit samples. Comparative experiments were carried out against recurrent neural network (RNN) and LSTM baseline models.
RESULTS: The BiLSTM-based model generated fluorescence predictions that closely aligned with experimentally measured values and demonstrated the best performance among the tested models. Importantly, the model enabled early qualitative classification using only 27 amplification cycles, compared with the conventional 45 cycles, corresponding to a 40% reduction in required cycles. The classification accuracy on the test dataset reached about 92%.
CONCLUSIONS: The proposed q-qPCR algorithm can be incorporated into conventional qPCR workflows and may contribute to shortening assay runtime and improving sample turnaround time (TAT). By enabling earlier preliminary interpretation, this approach could support more timely triage and diagnostic response in time-sensitive infectious disease testing scenarios.