Citation: Lusnig, L.; Sagingalieva, A.;
Surmach, M.; Protasevich, T.; Michiu,
O.; McLoughlin, J.; Mansell, C.; de’
Petris, G.; Bonazza, D.; Zanconati,
F.; et al. Hybrid Quantum Image
Classification and Federated Learning
for Hepatic Steatosis Diagnosis.
Diagnostics 2024, 14, 558.
https://doi.org/10.3390/
diagnostics14050558
Academic Editor: Dechang Chen
Received: 10 January 2024
Revised: 17 February 2024
Accepted: 26 February 2024
Published: 6 March 2024
Copyright: © 2024 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
diagnostics
Article
Hybrid Quantum Image Classification and Federated
Learning for Hepatic Steatosis Diagnosis
Luca Lusnig
1,2,
*, Asel Sagingalieva
1
, Mikhail Surmach
1
, Tatjana Protasevich
1
, Ovidiu Michiu
1
,
Joseph McLoughlin
1
, Christopher Mansell
1
, Graziano de’ Petris
3
, Deborah Bonazza
4
, Fabrizio Zanconati
4
,
Alexey Melnikov
1
and Fabio Cavalli
2,
*
1
Terra Quantum AG, 9000 St. Gallen, Switzerland; jm@terraquantum.swiss (J.M.);
cm@terraquantum.swiss (C.M.); ame@terraquantum.swiss (A.M.)
2
Research Unit of Paleoradiology and Allied Sciences, Laboratorio di Telematica Sanitaria-Struttura Complessa
Informatica e Telecomunicazioni, Azienda Sanitaria Universitaria Giuliana Isontina, 34149 Trieste, Italy
3
Laboratorio di Telematica Sanitaria-Struttura Complessa Informatica e Telecomunicazioni, Azienda Sanitaria
Universitaria Giuliana Isontina, 34149 Trieste, Italy; graziano.depetris@asugi.sanita.fvg.it
4
Department of Medical, Surgical and Health Sciences, University of Trieste, Cattinara Academic Hospital,
34149 Trieste, Italy; deborah.bonazza@asugi.sanita.fvg.it (D.B.)
* Correspondence: ll@terraquantum.swiss (L.L.); fabio.cavalli@asugi.sanita.fvg.it (F.C.)
Abstract: In the realm of liver transplantation, accurately determining hepatic steatosis levels is
crucial. Recognizing the essential need for improved diagnostic precision, particularly for optimizing
diagnosis time by swiftly handling easy-to-solve cases and allowing the expert time to focus on more
complex cases, this study aims to develop cutting-edge algorithms that enhance the classification of
liver biopsy images. Additionally, the challenge of maintaining data privacy arises when creating
automated algorithmic solutions, as sharing patient data between hospitals is restricted, further
complicating the development and validation process. This research tackles diagnostic accuracy by
leveraging novel techniques from the rapidly evolving field of quantum machine learning, known
for their superior generalization abilities. Concurrently, it addresses privacy concerns through the
implementation of privacy-conscious collaborative machine learning with federated learning. We
introduce a hybrid quantum neural network model that leverages real-world clinical data to assess
non-alcoholic liver steatosis accurately. This model achieves an image classification accuracy of 97%,
surpassing traditional methods by 1.8%. Moreover, by employing a federated learning approach that
allows data from different clients to be shared while ensuring privacy, we maintain an accuracy rate
exceeding 90%. This initiative marks a significant step towards a scalable, collaborative, efficient, and
dependable computational framework that aids clinical pathologists in their daily diagnostic tasks.
Keywords: medical image classification; hybrid quantum ResNet; quantum machine learning; hybrid
quantum neural network; federated learning; computer-aided diagnosis; histology
1. Introduction
In addressing the global challenge of determining liver viability for transplantation,
this study tackles two main issues: the accuracy of hepatic steatosis diagnostics and the
preservation of patient data privacy. The accurate classification of liver biopsy images
is vital for assessing transplant viability, yet it is hampered by the substantial data re-
quirements for training sophisticated machine learning models and the inherent privacy
concerns associated with sensitive patient data. Federated learning (FL) offers a solution
to these privacy concerns by enabling collaborative model training across multiple clients
without centralizing sensitive data, thus adhering to data protection regulations such as
the GDPR [1]. This approach is further underscored by upcoming regulations, like the EU
AI Law [2], emphasizing the need for models that balance accuracy with privacy.
Diagnostics 2024, 14, 558. https://doi.org/10.3390/diagnostics14050558 https://www.mdpi.com/journal/diagnostics