AI automates sister chromatid exchange counting, improving diagnosis of Bloom syndrome

AI automates sister chromatid exchange counting, improving diagnosis of Bloom syndrome

Researchers at Tokyo Metropolitan University have developed a series of algorithms to automate the counting of sister chromatid exchanges (SCE) in chromosomes under the microscope. Conventional analysis requires trained personnel and time, which varies from person to person. The team's machine learning-based algorithm has 84% ​​accuracy and provides a more objective measurement. This could be crucial for diagnosing disorders associated with an abnormal number of SCE, such as Bloom's syndrome.

DNA, the blueprint of life for all living organisms, is packaged in complex structures called chromosomes. When DNA is replicated, two identical strands are created, called sister chromatids, each carrying exactly the same genetic information. Unlike meiosis, sister chromatids do not have to undergo recombination during mitosis and are in most cases transferred intact to the daughter cells. However, if damage to the DNA occurs, the organism attempts to repair the lesion by using the remaining undamaged DNA as a template. During this repair process, it often happens that certain sections of the sister chromatids are exchanged with each other. During this repair process, it often happens that certain sections of the sister chromatids are exchanged with each other. This “sister chromatic exchange” (SCE) is not harmful in itself, but too many can be a good indicator of some serious disorders. Examples of this are Bloom's syndrome: those affected may have a predisposition to cancer.

To count SCEs, in normal methods, experienced clinicians examine stained chromosomes under a microscope and attempt to identify the telltale "swapped" segments of sister chromatids. Not only is this labor-intensive and slow, but it can also be subjective and dependent on how the human eye perceives features. Fully automated analysis of microscope images would save time and provide objective measurements of the number of SCEs to enable more consistent diagnoses in different clinical settings.

Now a team led by professors Kiyoshi Nishikawa and Kan Okubo from Tokyo Metropolitan University has developed a set of algorithms that use machine learning to count SCEs in images. They combined different methods: one to identify individual chromosomes, another to determine whether SCEs are present, and finally another to group and count these chromosomes, resulting in an objective, fully automated measurement of the number of SCEs in a microscope image. They determined an accuracy of 84.1%, a value that is sufficient for practical applications. To see how it works with real data, they collected images of chromosomes from artificially knocked-out cellsBLMGene, the type of suppression seen in patients with Bloom syndrome. The team's algorithm was able to produce counts for SCEs that matched those of human counters.

Work is currently underway to utilize the vast amounts of available clinical data to train the algorithm, with further refinements to come. The team believes that replacing manual counting with full automation will help enable faster and more objective clinical analysis than ever before, and that this is just the beginning of what AI can do for medical research.

This work was supported by JSPS KAKENHI Grant Numbers 22H05072, 25K09513 and 22K12170.

Sources: