What type of study is this?

November 1, 2025American Journal of Clinical PathologyOpen Access

367 Deep Learning-Based Classification of AML Subtypes Using Cytomorphology Images: A Pathology-Centered Workflow with Grad-CAM Interpretability

View Full Paper

Ask AI

Bookmark

Authors

SSSaanie Sulley

Discussion

Member takes

Implication

Analysis reveals improved interpretability and performance for acute myeloid leukemia subtype classification, highlighting class imbalance challenges.

Key Points

Validation accuracy reached 55%, showcasing robust sensitivity for minority acute myeloid leukemia subtypes.
Convolutional neural network optimized with a class-balanced sampling strategy addressed the issue of class imbalance.
Grad-CAM visualizations provided insights into critical nuclear and cytoplasmic regions for diagnostics in the study.
Findings underscore the importance of visual analytics in AI-assisted hematopathology workflows for precision diagnostics.

Cite This Study

Saanie Sulley (2025) studied this question.

synapsesocial.com/papers/6925436cc0ce034ddc358727 https://doi.org/10.1093/ajcp/aqaf121.365

View Full Paper

Ask AI

Bookmark

Implication

Analysis reveals improved interpretability and performance for acute myeloid leukemia subtype classification, highlighting class imbalance challenges.

367 Deep Learning-Based Classification of AML Subtypes Using Cytomorphology Images: A Pathology-Centered Workflow with Grad-CAM Interpretability

Discussion

Member takes

Implication

Key Points

Cite This Study

367 Deep Learning-Based Classification of AML Subtypes Using Cytomorphology Images: A Pathology-Centered Workflow with Grad-CAM Interpretability

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider