Hanbiao Chen

Method: This paper focuses on chromosome classification because it is critical for chromosome karyotyping. In recent years, deep learning-based methods are the most promising methods for solving the tasks of chromosome classification. Although the deep learning-based Inception architecture has yielded state-of-the-art performance in the 2015 ILSVRC challenge, it has not been used in chromosome classification tasks so far. Therefore, we develop an automatic chromosome classification approach named CIR-Net based on Inception-ResNet which is an optimized version of Inception. However, the classification performance of origin Inception-ResNet on the insufficient chromosome dataset still has a lot of capacity for improvement. Further, we propose a simple but effective augmentation method called CDA for improving the performance of CIR-Net. Results: The experimental results show that our proposed method achieves 95.98% classification accuracy on the clinical G-band chromosome dataset whose training dataset is insufficient. Moreover, the proposed augmentation method CDA improves more than 8.5% (from 87.46% to 95.9%) classification accuracy compared to other methods. In this paper, the experimental results demonstrate that our proposed method is recent the most effective solution for solving clinical chromosome classification problems in chromosome auto-karyotyping on the condition of the insufficient training dataset. Code and Dataset are available at https://github.com/CloudDataLab/CIR-Net.

Pallister-Killian syndrome (PKS) (OMIM:#601803) is a rare sporadic genetic disorder characterized by multi-malformations which is caused by the presence of the extra isochromosome 12p. PKS is featured by the tissue-limited mosaicism of the isochromosome 12p [i(12p)]. There were a wide spectrum of prenatal ultrasound findings of PKS, which made it difficult to be found in first or second trimester. Polyhydramnios, diaphragmatic hernia, and rhizomelic limb shortening were the most common prenatal ultrasound abnormalities in PKS. This study retrospectively analyzed the ultrasound findings and molecular cytogenetic results of four PKS fetuses diagnosed by using cord blood samples. The ultrasound anomalies of four PKS fetuses are described as follows: fetal macrosomia, cerebral ventriculomegaly, increased NT thickness, rhizomelic limbs shortening, polyhydramnios. Biparietal diameter (BPD), head circumference (HC), abdominal circumference (AC), femur length (FL) measurements were above the mean in three fetuses,while one fetus showed rhizomelic limbs shortening. Combined with this study and previous literature, polyhydramnios was the most frequent anomaly observed in prenatal ultrasound examination of PKS, which accounted for 48% (94/194). Fetal macrosomia was present in 15% (29/194), cerebral ventriculomegaly in 13% (25/194), thickened nuchal fold in 9% (18/194), rhizomelic limbs shortening in 26% (51/194). I(12p) was found in the karyotype analysis of cultured cord blood lymphocytes and the mosaic ratios ranged from 2 to 5%. Single nucleotide polymorphisms array (SNP-array) results suggested that the whole short arm of chromosome 12 was duplicated with 2~3 copies. Fluorescence in situ hybridization (FISH) was performed to confirm the results of karyotype and SNP-array. In case non-specific indicators such as fetal macrosomia, polyhydramnios and rhizomelic limbs shortening are observed meanwhile in prenatal ultrasound, targeted detection of PKS should be considered. In the prenatal diagnosis of PKS, the combination of SNP-array and FISH with conventional karyotype are the key to seek i(12p) and for precise diagnosis.