THE ROLE OF AI ASSISTANCE IN CONVENTIONAL KARYOTYPING: ENHANCEMENTS, LIMITATIONS, AND THE CONTINUING NEED FOR HUMAN EXPERTISE

Authors

  • Gul E Rehan Islamabad Diagnostic Center
  • Aftab Ahmad Khan Islamabad Diagnostic Centre (IDC)
  • Rizwan Uppal Islamabad Diagnostic Centre (IDC)
  • Hamid Saeed Malik Islamabad Diagnostic Centre (IDC)

DOI:

https://doi.org/10.52764/jms.26.34.2.5

Keywords:

Artificial Intelligence, , Karyotyping, Cytogenetics

Abstract

Objectives:

This study aimed to assess the performance, efficiency, and limitations of AI-assisted karyotyping using Applied Spectral Imaging (ASI) software and to highlight the importance of expert cytogeneticist involvement to ensure that the diagnosis is accurate, reliable, and properly interpreted in light of clinical history and laboratory findings.

Methods:

A comparative observational study was conducted at Islamabad Diagnostic Center from October 2025 to December 2025. A total of 2,305 chromosomes were analyzed using ASI AI-based karyotyping software. Each AI-generated karyogram was reviewed by two cytogeneticists. Parameters evaluated in this study included chromosome count accuracy, overlap detection, segmentation errors requiring joining or separation, and the correctness of chromosome placement. The time required for AI-assisted analysis was compared with that of conventional manual karyotyping.

Results:

The AI software detected a mean of 49.34 ± 2.63 chromosomes per metaphase, indicating a tendency toward over-segmentation. Chromosomal overlap was observed in 2.30% of cases. Manual joining and separation of chromosome segments were required in 7.50% and 4.59% of chromosomes, respectively. Correct placement without human intervention was achieved in 82.93% of chromosomes. AI-assisted analysis, including manual verification, required approximately 2 minutes per metaphase, compared with 10–15 minutes for manual karyotyping.

Conclusion:

AI-assisted karyotyping improves efficiency and reduces turnaround time. Expert review remains essential for resolving errors and interpreting complex patterns. A combined AI–human approach provides the most reliable framework for accurate cytogenetic interpretation.

Keywords (MeSH): Artificial Intelligence, Karyotyping, Cytogenetics, Chromosomal abnormalities

Downloads

Published

2026-06-30

How to Cite

Rehan, G. E., Ahmad Khan, A., Uppal, R., & Malik, H. S. (2026). THE ROLE OF AI ASSISTANCE IN CONVENTIONAL KARYOTYPING: ENHANCEMENTS, LIMITATIONS, AND THE CONTINUING NEED FOR HUMAN EXPERTISE. Journal of Medical Sciences, 34(2), 87–92. https://doi.org/10.52764/jms.26.34.2.5

Most read articles by the same author(s)

Similar Articles

You may also start an advanced similarity search for this article.