AI-based body composition analysis of CT data has the potential to predict disease course in patients with multiple myeloma

Franz Wegner; Malte Maria Sieren; Hanna Grasshoff; Lennart Berkel; Christoph Rowold; Marcel Philipp Röttgerding; Soleiman Khalil; Sam Mogadas; Felix Nensa; René Hosch; Gabriela Riemekasten; Anna Franziska Hamm; Nikolas von Bubnoff; Jörg Barkhausen; Roman Kloeckner; Cyrus Khandanpour; Theo Leitner

doi:10.1038/s41598-025-11560-3

AI-based body composition analysis of CT data has the potential to predict disease course in patients with multiple myeloma

Franz Wegner^*, Malte Maria Sieren, Hanna Grasshoff, Lennart Berkel, Christoph Rowold, Marcel Philipp Röttgerding, Soleiman Khalil, Sam Mogadas, Felix Nensa, René Hosch, Gabriela Riemekasten, Anna Franziska Hamm, Nikolas von Bubnoff, Jörg Barkhausen, Roman Kloeckner, Cyrus Khandanpour^*, Theo Leitner^*

^*Corresponding author for this work

Abstract

The aim of this study was to evaluate the benefit of a volumetric AI-based body composition analysis (BCA) algorithm in multiple myeloma (MM). Therefore, a retrospective monocentric cohort of 91 MM patients was analyzed. The BCA algorithm, powered by a convolutional neural network, quantified tissue compartments and bone density based on routine CT scans. Correlations between BCA data and demographic/clinical parameters were investigated. BCA-endotypes were identified and survival rates were compared between BCA-derived patient clusters. Patients with high-risk cytogenetics exhibited elevated cardiac marker index values. Across Revised-International Staging System (R-ISS) categories, BCA parameters did not show significant differences. However, both subcutaneous and total adipose tissue volumes were significantly lower in patients with progressive disease or death during follow-up compared to patients without progression. Cluster analysis revealed two distinct BCA-endotypes, with one group displaying significantly better survival. Furthermore, a combined model composed of clinical parameters and BCA data demonstrated a higher predictive capability for disease progression compared to models based solely on high-risk cytogenetics or R-ISS. These findings underscore the potential of BCA to improve patient stratification and refining prognostic models in MM.

Original language	English
Article number	26455
Journal	Scientific Reports
Volume	15
Issue number	1
ISSN	2045-2322
DOIs	https://doi.org/10.1038/s41598-025-11560-3
Publication status	Published - 12.2025

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being
SDG 9 Industry, Innovation, and Infrastructure

Research Areas and Centers

Centers: Center for Artificial Intelligence Luebeck (ZKIL)
Academic Focus: Center for Infection and Inflammation Research (ZIEL)
Academic Focus: Biomedical Engineering
Research Area: Luebeck Integrated Oncology Network (LION)

DFG Research Classification Scheme

2.21-05 Immunology
2.22-07 Medical Informatics and Medical Bioinformatics
2.22-32 Medical Physics, Biomedical Technology
2.22-18 Rheumatology
2.22-14 Hematology, Oncology
4.43-04 Artificial Intelligence and Machine Learning Methods

Access to Document

10.1038/s41598-025-11560-3

Cite this

Wegner, F., Sieren, M. M., Grasshoff, H., Berkel, L., Rowold, C., Röttgerding, M. P., Khalil, S., Mogadas, S., Nensa, F., Hosch, R., Riemekasten, G., Hamm, A. F., von Bubnoff, N., Barkhausen, J., Kloeckner, R., Khandanpour, C., & Leitner, T. (2025). AI-based body composition analysis of CT data has the potential to predict disease course in patients with multiple myeloma. Scientific Reports, 15(1), Article 26455. https://doi.org/10.1038/s41598-025-11560-3

@article{18b6639c01fd4b1fa85073804cee168a,

title = "AI-based body composition analysis of CT data has the potential to predict disease course in patients with multiple myeloma",

abstract = "The aim of this study was to evaluate the benefit of a volumetric AI-based body composition analysis (BCA) algorithm in multiple myeloma (MM). Therefore, a retrospective monocentric cohort of 91 MM patients was analyzed. The BCA algorithm, powered by a convolutional neural network, quantified tissue compartments and bone density based on routine CT scans. Correlations between BCA data and demographic/clinical parameters were investigated. BCA-endotypes were identified and survival rates were compared between BCA-derived patient clusters. Patients with high-risk cytogenetics exhibited elevated cardiac marker index values. Across Revised-International Staging System (R-ISS) categories, BCA parameters did not show significant differences. However, both subcutaneous and total adipose tissue volumes were significantly lower in patients with progressive disease or death during follow-up compared to patients without progression. Cluster analysis revealed two distinct BCA-endotypes, with one group displaying significantly better survival. Furthermore, a combined model composed of clinical parameters and BCA data demonstrated a higher predictive capability for disease progression compared to models based solely on high-risk cytogenetics or R-ISS. These findings underscore the potential of BCA to improve patient stratification and refining prognostic models in MM.",

author = "Franz Wegner and Sieren, \{Malte Maria\} and Hanna Grasshoff and Lennart Berkel and Christoph Rowold and R{\"o}ttgerding, \{Marcel Philipp\} and Soleiman Khalil and Sam Mogadas and Felix Nensa and Ren{\'e} Hosch and Gabriela Riemekasten and Hamm, \{Anna Franziska\} and \{von Bubnoff\}, Nikolas and J{\"o}rg Barkhausen and Roman Kloeckner and Cyrus Khandanpour and Theo Leitner",

note = "Publisher Copyright: {\textcopyright} The Author(s) 2025.",

year = "2025",

month = dec,

doi = "10.1038/s41598-025-11560-3",

language = "English",

volume = "15",

journal = "Scientific Reports",

issn = "2045-2322",

publisher = "Springer Science and Business Media, LLC",

number = "1",

}

Wegner, F , Sieren, MM , Grasshoff, H , Berkel, L, Rowold, C, Röttgerding, MP, Khalil, S, Mogadas, S, Nensa, F, Hosch, R, Riemekasten, G , Hamm, AF , von Bubnoff, N , Barkhausen, J , Kloeckner, R , Khandanpour, C & Leitner, T 2025, 'AI-based body composition analysis of CT data has the potential to predict disease course in patients with multiple myeloma', Scientific Reports, vol. 15, no. 1, 26455. https://doi.org/10.1038/s41598-025-11560-3

TY - JOUR

T1 - AI-based body composition analysis of CT data has the potential to predict disease course in patients with multiple myeloma

AU - Wegner, Franz

AU - Sieren, Malte Maria

AU - Grasshoff, Hanna

AU - Berkel, Lennart

AU - Rowold, Christoph

AU - Röttgerding, Marcel Philipp

AU - Khalil, Soleiman

AU - Mogadas, Sam

AU - Nensa, Felix

AU - Hosch, René

AU - Riemekasten, Gabriela

AU - Hamm, Anna Franziska

AU - von Bubnoff, Nikolas

AU - Barkhausen, Jörg

AU - Kloeckner, Roman

AU - Khandanpour, Cyrus

AU - Leitner, Theo

PY - 2025/12

Y1 - 2025/12

N2 - The aim of this study was to evaluate the benefit of a volumetric AI-based body composition analysis (BCA) algorithm in multiple myeloma (MM). Therefore, a retrospective monocentric cohort of 91 MM patients was analyzed. The BCA algorithm, powered by a convolutional neural network, quantified tissue compartments and bone density based on routine CT scans. Correlations between BCA data and demographic/clinical parameters were investigated. BCA-endotypes were identified and survival rates were compared between BCA-derived patient clusters. Patients with high-risk cytogenetics exhibited elevated cardiac marker index values. Across Revised-International Staging System (R-ISS) categories, BCA parameters did not show significant differences. However, both subcutaneous and total adipose tissue volumes were significantly lower in patients with progressive disease or death during follow-up compared to patients without progression. Cluster analysis revealed two distinct BCA-endotypes, with one group displaying significantly better survival. Furthermore, a combined model composed of clinical parameters and BCA data demonstrated a higher predictive capability for disease progression compared to models based solely on high-risk cytogenetics or R-ISS. These findings underscore the potential of BCA to improve patient stratification and refining prognostic models in MM.

AB - The aim of this study was to evaluate the benefit of a volumetric AI-based body composition analysis (BCA) algorithm in multiple myeloma (MM). Therefore, a retrospective monocentric cohort of 91 MM patients was analyzed. The BCA algorithm, powered by a convolutional neural network, quantified tissue compartments and bone density based on routine CT scans. Correlations between BCA data and demographic/clinical parameters were investigated. BCA-endotypes were identified and survival rates were compared between BCA-derived patient clusters. Patients with high-risk cytogenetics exhibited elevated cardiac marker index values. Across Revised-International Staging System (R-ISS) categories, BCA parameters did not show significant differences. However, both subcutaneous and total adipose tissue volumes were significantly lower in patients with progressive disease or death during follow-up compared to patients without progression. Cluster analysis revealed two distinct BCA-endotypes, with one group displaying significantly better survival. Furthermore, a combined model composed of clinical parameters and BCA data demonstrated a higher predictive capability for disease progression compared to models based solely on high-risk cytogenetics or R-ISS. These findings underscore the potential of BCA to improve patient stratification and refining prognostic models in MM.

UR - https://www.scopus.com/pages/publications/105011165147

U2 - 10.1038/s41598-025-11560-3

DO - 10.1038/s41598-025-11560-3

M3 - Journal articles

C2 - 40691492

AN - SCOPUS:105011165147

SN - 2045-2322

VL - 15

JO - Scientific Reports

JF - Scientific Reports

IS - 1

M1 - 26455

ER -

AI-based body composition analysis of CT data has the potential to predict disease course in patients with multiple myeloma

Abstract

UN SDGs

Research Areas and Centers

DFG Research Classification Scheme

Access to Document

Other files and links

Fingerprint

Cite this