A Deep Learning Framework for Predicting Testosterone Deficiency using CAE-Adaptive LSTM

DOI

10.2991/978-94-6239-616-6_28How to use a DOI?

Keywords

Testosterone Deficiency; TD Prediction; Artificial Intelligence; SMOTE; CAE; Adaptive LSTM

Abstract

One of the most common endocrine syndromes in males, Testosterone Deficiency (TD), is often accompanied by reduced quality of life, infertility, metabolic dysfunctions, and potentially chronic diseases. Traditional methods rely solely on invasive biochemical tests and clinical evaluations and can consume a lot of time or be subject to excessive variation. Instead, machine learning and deep learning have recently gained popularity as prime candidates for the early diagnosis of TD, with an increasing demand for more prompt, reliable, and automated tools. This study proposes a novel predictive model that uses a Convolutional Autoencoder (CAE) for feature extraction and an Adaptive Long Short-Term Memory (LSTM) network for the prediction and classification of secondary testosterone deficiency. The CAE reduces dimensionality while retaining the discriminative features of the dataset, with the Adaptive-LSTM benefiting classification by learning temporal dependencies and nonlinear patterns dynamically. Z-score normalization is applied during preprocessing to improve balanced learning, along with the Synthetic Minority Over-Sampling Technique (SMOTE). Experimental evaluation demonstrates that the proposed model achieved superior performance compared to conventional machine learning and ensemble-based methods. The model attained an accuracy value of 97.06%, a precision of 96.78%, a recall of 96.14%, a specificity of 97.51%, and an F1-score of 96.42%. The findings highlight the potential of the proposed CAE–Adaptive LSTM model as an effective decision-support system for the accurate prediction of TD.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - P.John William
AU  - E. Ilavarasan
PY  - 2026
DA  - 2026/03/31
TI  - A Deep Learning Framework for Predicting Testosterone Deficiency using CAE-Adaptive LSTM
BT  - Proceedings of the International Conference on Artificial Intelligence and Secure Data Analytics (ICAISDA 2025)
PB  - Atlantis Press
SP  - 339
EP  - 363
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-616-6_28
DO  - 10.2991/978-94-6239-616-6_28
ID  - William2026
ER  -