SUPR
AI-driven segmentation for testicular sample evaluation
Dnr:

NAISS 2024/22-1403

Type:

NAISS Small Compute

Principal Investigator:

João Pedro Alves Lopes

Affiliation:

Karolinska Institutet

Start Date:

2024-11-05

End Date:

2025-06-01

Primary Classification:

10604: Cell Biology

Allocation

Abstract

Various diseases, including hematological disorders or cancer, can impact the highly complex process of spermatogenesis directly or indirectly, through the side effects of necessary treatments. As a result, this can lead to compromised fertility or even complete infertility, especially in young boys before and during puberty, who may struggle to produce viable sperm for future use later in life. For these patients, a technique called testicular cryopreservation is employed as a means of preserving fertility. However, given the experimental nature of this approach, the need to minimize invasive procedures, and the patients' right to be informed about the state of their tissue, it is crucial to assess the quality of testis biopsy samples. To achieve this efficiently, our goal is to utilize artificial intelligence to automate the segmentation of images obtained from patients' testicular biopsies. This approach will enable us to extract large amounts of data and provide both patients and doctors with detailed information on the options for fertility preservation.