Background

Due to the highly variable clinical phenotype, Klinefelter Syndrome (KS) is underdiagnosed.

Objective

Assessment of supervised machine learning (sML)-based prediction models for identification of KS among azoospermic patients, and comparison to expert clinical evaluation.

Materials and methods

Retrospective patient data (karyotype, age, height, weight, testis volume, FSH, LH, testosterone, estradiol, prolactin, semen pH and semen volume) collected between January 2005 and June 2019 were retrieved from a patient data bank of a University Centre. Models were trained, validated and benchmarked based on different sML algorithms. Models were then tested on an independent, prospectively acquired set of patient data (between July 2019 and July 2020). Benchmarking against physicians was performed in addition.

Results

Based on average performance, support vector machines and CatBoost were particularly well-suited models, with 100% sensitivity and >93% specificity on the test dataset. Compared to a group of 18 expert clinicians, the ML models had significantly better median sensitivity (100% vs 87.5%, p = 0.0455) and fared comparably with regards to specificity (90% vs 89.9%, p = 0.4795), thereby possibly improving diagnosis rate. A KS Score Calculator based on the prediction models is available on http://klinefelter-score-calculator.uni-muenster.de.

Discussion

Differentiating KS patients from azoospermic patients with normal karyotype (46,XY) is a problem that can be solved with sML techniques, improving patient care.

Conclusions

Machine learning could improve the diagnostic rate of KS among azoospermic patients, even more for less experienced physicians.

This article is protected by copyright. All rights reserved