Problem
Research questions and friction points this paper is trying to address.
Detects chromosome abnormalities causing blood cancers
Reduces manual analysis time to 15 seconds per image
Enables zero-shot detection of rare genetic aberrations
Karyotype AI for Precision Oncology
🤖 AI Summary
Chromosomal abnormality detection in hematologic malignancies relies heavily on costly and time-consuming manual karyotyping, hindering rapid clinical diagnosis. Method: This paper proposes an end-to-end automated framework for chromosomal abnormality identification from metaphase spread microscopic images. It integrates self-supervised pretraining, fine-tuned Vision Transformers (ViTs), latent-space embedding modeling, and zero-shot transfer learning. A novel pretraining-finetuning strategy is introduced to mitigate the small-sample bottleneck. Contribution/Results: On real-world metaphase images, the framework achieves 94% AUC for critical abnormalities including del(5q) and t(9;22). Processing time per image is reduced to 15 seconds—over 100× faster than manual analysis. To our knowledge, it is the first method enabling zero-shot recognition of rare chromosomal aberrations. Furthermore, the system provides clinically interpretable outputs and supports scalable deployment in diagnostic workflows.
📝 Abstract
We present a machine learning method capable of accurately detecting chromosome abnormalities that cause blood cancers directly from microscope images of the metaphase stage of cell division. The pipeline is built on a series of fine-tuned Vision Transformers. Current state of the art (and standard clinical practice) requires expensive, manual expert analysis, whereas our pipeline takes only 15 seconds per metaphase image. Using a novel pretraining-finetuning strategy to mitigate the challenge of data scarcity, we achieve a high precision-recall score of 94% AUC for the clinically significant del(5q) and t(9;22) anomalies. Our method also unlocks zero-shot detection of rare aberrations based on model latent embeddings. The ability to quickly, accurately, and scalably diagnose genetic abnormalities directly from metaphase images could transform karyotyping practice and improve patient outcomes. We will make code publicly available.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Vision Transformers for chromosome abnormality detection
15-second analysis per metaphase image
Zero-shot detection of rare genetic aberrations
🔎 Similar Papers
No similar papers found.
