Search papers, labs, and topics across Lattice.
This study introduces a deep learning pipeline for the patient-level diagnosis of acute myeloid leukemia (AML) using bone marrow smear images, addressing the labor-intensive nature of manual cell interpretation. By employing a YOLO-based segmentation module and an EfficientNet-B0 classifier, the model predicts a composite category of cells termed Composite Blast-like Cells (CBLC), which aggregates various cellular observations into a single diagnostic metric. The approach achieved high ensemble weighted F1-scores, demonstrating robust internal validation and external generalization across multiple patient cohorts.
Achieving an ensemble F1-score of 0.9124, this deep learning model redefines AML diagnosis by aggregating complex cellular data into actionable insights.
Bone marrow smear review remains important for acute myeloid leukemia (AML) assessment, but manual single-cell interpretation is labor-intensive and patient-level diagnosis requires aggregation of many cellular observations. We present a cell-to-patient deep learning pipeline for AML-assisted diagnosis from bone marrow smear images. The study included 258 patients from six anonymized centers, including a main cohort of 169 patients from Centers 1-3 and an external validation cohort of 89 patients from Centers 4-6. A 16-category cell annotation vocabulary was used to describe the global cellular composition, including granulocytic, monocytic, erythroid, lymphoid, eosinophilic, and other cells. Rather than identifying strict AML blasts or leukemic blasts, the model targets an expert-defined composite category termed Composite Blast-like Cells (CBLC), comprising N, N1, M, M1, R, R1, J, and J1 according to the project-wide morphological standard. A fixed YOLO-based segmentation module detected cells, predicted contours were matched to expert polygon annotations by contour IoU, and standardized single-cell crops were generated. An EfficientNet-B0 classifier was trained through a two-stage GT-to-YOLO and YOLO-to-YOLO strategy with class-imbalance correction, center-border regularization, and morphology-assisted supervision. Cell-level predictions were aggregated into patient-level CBLC ratios for AML-oriented diagnostic support. The pipeline achieved stable internal validation and maintained external generalization, with ensemble weighted F1-scores of 0.9076, 0.8696, and 0.9124 on Centers 4, 5, and 6, respectively.
