2023 ACMG Annual Clinical Genetics Meeting Digital Edition: Poster Gallery-Improved Efficiency of G-band Chromosome Analysis via Deep Neural Network Workflow

Pdf Summary

A study conducted by the Mayo Clinic evaluated the efficiency of G-band chromosome analysis using deep neural network (DNN) workflows in a high-volume cytogenetics laboratory. The researchers compared the accuracy, ergonomics impact, and analysis time of different platforms.<br /><br />First, they compared the MetaSystems V1 platform with CytoVision. The results showed that the average accuracy of chromosome placement using MetaSystems V1 was significantly higher for blood and bone marrow specimens compared to CytoVision. Additionally, the total number of mouse clicks required to complete a karyotype was significantly less using MetaSystems V1. This indicated that the DNN classifier in MetaSystems V1 improved accuracy and reduced the strain on technologists.<br /><br />Next, they compared the MetaSystems V1 with MetaSystems V2 Ikaros platform. The results showed that the average time to complete a chromosome case was significantly less using V2 compared to V1 for both blood and bone marrow specimens. The addition of a metaphase image segmentation DNN classifier in the V2 platform reduced the analysis time.<br /><br />Overall, the study demonstrated that the deployment of DNN workflows in cytogenetics laboratories has the potential to overcome the shortage of experienced technologists. The automation of chromosome identification and analysis using DNN classifiers improved efficiency and accuracy. Future studies are underway to compare the mouse clicks needed for MetaSystems V1 and V2 platforms.<br /><br />This research provides valuable insights into the benefits of using DNN workflows in G-band chromosome analysis, offering potential solutions to the labor-intensive and manual nature of the process. These findings have implications for the genetics workup of congenital conditions and hematologic malignancies.

Asset Subtitle

Presenting Author - Xinjie Xu, PhD; Co-Author - Derek G. Slagter, BA; Co-Author - Todd J. VanDeWalker, BA; Co-Author - Erin C. Theobald, BS; Co-Author - Tina L. Brown, BS; Co-Author - Daniel A. Hoppe, BS; Co-Author - Sheryl L. Castro, BS; Co-Author - Joel T. Denney, BA; Co-Author - Jennifer P. Gardner, MS; Co-Author - Christopher A. Stahl, BA, MBA; Co-Author - Hutton M. Kearney, PhD; Co-Author - Jansen N. Seheult, MB, BCh, MD; Co-Author - Carley Ann Suarez, BS; Co-Author - Harshawardhan Banda, MS; Submitter Only - Xinjie Xu, PhD;

Meta Tag

Cancer Cytogenetics

Clinical Cytogenetics

Cytogenetics

Co-Author Derek G. Slagter, BA

Co-Author Todd J. VanDeWalker, BA

Co-Author Erin C. Theobald, BS

Co-Author Tina L. Brown, BS

Co-Author Daniel A. Hoppe, BS

Co-Author Sheryl L. Castro, BS

Co-Author Joel T. Denney, BA

Co-Author Jennifer P. Gardner, MS

Co-Author Christopher A. Stahl, BA, MBA

Co-Author Hutton M. Kearney, PhD

Co-Author Jansen N. Seheult, MB, BCh, MD

Co-Author Carley Ann Suarez, BS

Co-Author Harshawardhan Banda, MS

Presenting Author Xinjie Xu, PhD

Submitter Only Xinjie Xu, PhD

Keywords

Mayo Clinic

G-band chromosome analysis

DNN workflows

MetaSystems V1

CytoVision

accuracy

analysis time

automation

chromosome identification

efficiency