Chromosome Classification and Straightening Based on an Interleaved and Multi-Task Network
- 25 February 2021
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 25 (8) , 3240-3251
Abstract
Karyotyping is the gold standard in the detection of chromosomal abnormalities. To facilitate the diagnostic process, in this paper, a method for chromosome classification and straightening based on an interleaved and multi-task network is proposed. This method consists of three stages. In the first stage, multi-scale features are learned via an interleaved network. In the second stage, high-resolution features from the first stage are input to a convolution neural subnetwork for chromosome joint detection, and other features are fused and fed to two multi-layer perceptron subnetworks for chromosome type and polarity classification. In the third stage, the bent chromosome is straightened with the help of detected joints by two steps: first the chromosome is separated, rotated and assembled according to the detected joints; then the areas around the bending points are recovered by replacing the gaps formed in the first step with the sampled intensities from the bent chromosome. The classification of type and polarity can expedite the process of producing karyograms, which is an important step for chromosome diagnosis in clinical practice. Straightening makes the banding information of the chromosome easier to read. Classification results of the 5-fold cross validation on our dataset with 32 810 chromosomes achieve average accuracy of 98.1% for type classification and 99.8% for polarity classification. The straightening results show consistency in intensity and length of the chromosome before and after straightening.Keywords
Funding Information
- National Key R&D Program of China (2018YFC0116402)
- Zhejiang Province Science and Technology (2017C03029)
This publication has 33 references indexed in Scilit:
- Straightening of highly curved human chromosome for cytogenetic analysisMeasurement, 2014
- Analysis of human chromosome classification using centromere positionMeasurement, 2014
- AN AUTOMATIC ALGORITHM FOR IDENTIFICATION AND STRAIGHTENING IMAGES OF CURVED HUMAN CHROMOSOMESBiomedical Engineering: Applications, Basis and Communications, 2012
- A modular framework for the automatic classification of chromosomes in Q-band imagesComputer Methods and Programs in Biomedicine, 2012
- Automatic Pattern Extraction and Classification for Chromosome ImagesJournal of Infrared, Millimeter, and Terahertz Waves, 2010
- Automated classification of metaphase chromosomes: Optimization of an adaptive computerized schemeJournal of Biomedical Informatics, 2009
- A novel algorithm for straightening highly curved images of human chromosomePattern Recognition Letters, 2008
- A rule-based computer scheme for centromere identification and polarity assignment of metaphase chromosomesComputer Methods and Programs in Biomedicine, 2008
- Medial axis transform-based features and a neural network for human chromosome classificationPattern Recognition, 1995
- Automatically inferred markov network models for classification of chromosomal band pattern structuresCytometry, 1990