• Parvatham Kavya
• M P Princy Verronica
• Kasabu Pavan
• Surekha Satram

Accurately classifying bone marrow cells is essential for identifying a variety of hematological illnesses in the realm of medical image analysis. However, traditional methods face challenges in distinguishing between different cell types due to their complex morphological features. This work will build a model based on multiscale convolutional neural network that can automatically extract discriminative features from the original photos in order to address these difficulties. In addition, this research presents a novel method to enhance the interpretability and performance of deep learning models for bone marrow cell categorization by using multi-scale cross attention mechanisms and visual attention networks. Results from experiments show that our bone marrow cell classification accuracy exceeds other deep learning techniques as well as conventional machine learning methods. Moreover, the technique demonstrates outstanding reliability and stability.

Bone marrow cell classification, CNNs, Visual attention network, multi-scale cross attention, Medical Image analysis, Deep learning

"Multi-Scale Cross Attention for Improved Bone Marrow Cell Classification", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 4, page no.i359-i365, April-2024, Available :http://www.jetir.org/papers/JETIR2404845.pdf

"Multi-Scale Cross Attention for Improved Bone Marrow Cell Classification", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 4, page no. ppi359-i365, April-2024, Available at : http://www.jetir.org/papers/JETIR2404845.pdf