Neural network reveals platelet age from fluorescence microscopy images
Johan A Slotman1, Maurice Swinkels2, Sophie Hordijk2
1Optical Imaging Center, Department of Pathology, Erasmus MC, University Medical Centre Rotterdam, Rotterdam, The Netherlands.
Insights
Scientists developed an AI model to predict platelet age from microscopic images, achieving over 97% accuracy. This breakthrough could improve platelet transfusion safety and diagnose platelet disorders.
Area of Science:
- Hematology
- Artificial Intelligence
- Medical Imaging
Background:
- Platelets are crucial for hemostasis and thrombus formation.
- Platelet age correlates with clinical outcomes, but no reliable in vitro tests exist to predict age or assess transfusion product fitness.
- Current methods lack the ability to determine platelet age in vivo or during storage.
Purpose of the Study:
- To develop a convolutional neural network (CNN) model capable of predicting the chronological age of platelets from confocal microscopic images.
- To validate the model's accuracy in predicting platelet age both in vitro during storage and in vivo.
Main Methods:
- A CNN model was trained using confocal microscopic images of platelets stored in platelet-rich plasma (up to 8 hours) and routine platelet concentrates (up to 10 days).
- The model's performance was evaluated by predicting the chronological age of stored platelets.
- The model was further tested in vivo on a cohort of acute myeloid leukemia patients undergoing chemotherapy-induced thrombocytopenia.
Main Results:
- The CNN model achieved >97% accuracy in predicting the chronological age of stored platelets.
- The model successfully distinguished between younger and older platelets in vivo in patients during chemotherapy treatment.
- This demonstrates the model's capability for both in vitro and in vivo platelet age determination.
Conclusions:
- The developed AI model can accurately predict platelet chronological age from microscopic images.
- This technology holds significant potential for improving clinical transfusion medicine by assessing platelet viability.
- The model may also aid in the diagnosis and management of platelet disorders.
Abstract:
Platelets are small, anucleate cells with a primary physiological role in vascular damage repair (hemostasis) and initiation of thrombus formation in response to vascular injury. Platelets circulate approximately 7-10 days, slowly undergoing age-related changes in molecular composition, morphology, activation capacity, function, and surface receptor density. As older platelets are associated with poor clinical outcome, no in vitro tests are available to predict platelet age, or to determine the fitness of platelet transfusion products. In this study, we developed a convolutional neural network model that could determine platelets' chronological age from confocal microscopic images. The model was trained using platelets stored in platelet-rich plasma up to 8 hours and using routine platelet concentrates up to 10 days. The model predicted chronological age of stored platelets with >97% accuracy. To test our model in vivo, we analyzed a cohort of patients with acute myeloid leukemia, experiencing thrombocytopenia due to chemotherapy. Our model could reliably distinguish in vivo between samples with younger and older platelets during the course of treatment. This study demonstrates the ability to predict platelets' chronological age both in vitro during storage and in vivo, which may impact clinical transfusion medicine and the diagnosis and treatment of patients with platelet disorders.
