Hostname: page-component-cb9f654ff-k7rjm Total loading time: 0 Render date: 2025-08-03T02:20:34.949Z Has data issue: false hasContentIssue false
  • English
  • Français

Explainable deep learning model WAL-net for individualised assessment of potentially reversible malnutrition in patients with cancer: a multicentre cohort study

Published online by Cambridge University Press:  10 July 2025

Liangyu Yin Open the ORCID record for Liangyu Yin [Opens in a new window] ,
Ning Tong ,
Na Li ,
Jie Liu ,
Wei Li ,
Jiuwei Cui ,
Zengqing Guo ,
Qinghua Yao ,
Fuxiang Zhou  and
Ming Liu
...Show all authors
Liangyu Yin*
Affiliation:
Department of Nephrology, Chongqing Key Laboratory of Prevention and Treatment of Kidney Disease, Chongqing Clinical Research Center of Kidney and Urology Diseases, Xinqiao Hospital, Army Medical University (Third Military Medical University), Chongqing 400037, People’s Republic of China
Ning Tong
Affiliation:
Department of Clinical Nutrition, Daping Hospital, Army Medical University (Third Military Medical University), Chongqing 400042, People’s Republic of China
Na Li
Affiliation:
Department of Clinical Nutrition, Daping Hospital, Army Medical University (Third Military Medical University), Chongqing 400042, People’s Republic of China
Jie Liu
Affiliation:
Department of Clinical Nutrition, Daping Hospital, Army Medical University (Third Military Medical University), Chongqing 400042, People’s Republic of China
Wei Li
Affiliation:
Cancer Center of the First Hospital of Jilin University, Changchun, Jilin 130021, People’s Republic of China
Jiuwei Cui
Affiliation:
Cancer Center of the First Hospital of Jilin University, Changchun, Jilin 130021, People’s Republic of China
Zengqing Guo
Affiliation:
Department of Medical Oncology, Fujian Cancer Hospital, Fujian Medical University Cancer Hospital, Fuzhou, Fujian 350014, People’s Republic of China
Qinghua Yao
Affiliation:
Department of Integrated Chinese and Western Medicine, Cancer Hospital of the University of Chinese Academy of Science (Zhejiang Cancer Hospital), Hangzhou, Zhejiang 310022, People’s Republic of China
Fuxiang Zhou
Affiliation:
Department of Oncology, Zhongnan Hospital of Wuhan University, Wuhan, Hubei 430071, People’s Republic of China
Ming Liu
Affiliation:
Department of Colorectal Surgery, The Fourth Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang 150001, People’s Republic of China
Zhikang Chen
Affiliation:
Department of Colorectal and Anal Surgery, Xiangya Hospital of Central South University, Changsha, Hunan 410008, People’s Republic of China
Huiqing Yu
Affiliation:
Department of Palliative Care and Department of Geriatric Oncology, Chongqing University Cancer Hospital, Chongqing 400030, People’s Republic of China
Tao Li
Affiliation:
Department of Radiation Oncology, Sichuan Cancer Hospital and Institute, Sichuan Cancer Center, School of Medicine, University of Electronic Science and Technology of China, Chengdu, Sichuan 610041, People’s Republic of China
Zengning Li
Affiliation:
Department of Clinical Nutrition, The First Hospital of Hebei Medical University, Shijiazhuang, Hebei 050031, People’s Republic of China
Pingping Jia
Affiliation:
Department of Gastrointestinal Surgery and Department of Clinical Nutrition, Beijing Shijitan Hospital, Capital Medical University, Beijing 100038, People’s Republic of China
Chunhua Song
Affiliation:
Department of Epidemiology, College of Public Health, Zhengzhou University, Zhengzhou, Henan 450001, People’s Republic of China
Hongxia Xu*
Affiliation:
Department of Clinical Nutrition, Daping Hospital, Army Medical University (Third Military Medical University), Chongqing 400042, People’s Republic of China
Hanping Shi*
Affiliation:
Department of Gastrointestinal Surgery and Department of Clinical Nutrition, Beijing Shijitan Hospital, Capital Medical University, Beijing 100038, People’s Republic of China Key Laboratory of Cancer FSMP for State Market Regulation, Beijing 100038, People’s Republic of China Lead Contact and Principal Investigator of the Investigation on Nutrition Status and Its Clinical Outcome of Common Cancers (INSCOC) Project, Beijing, People’s Republic of China
*
Corresponding authors: Liangyu Yin, Emails: liangyuyin1988@tmmu.edu.cn, liangyuyin1988@qq.com; Hongxia Xu, Email: hx_xu2015@163.com; Hanping Shi, Email: shihp@ccmu.edu.cn
Corresponding authors: Liangyu Yin, Emails: liangyuyin1988@tmmu.edu.cn, liangyuyin1988@qq.com; Hongxia Xu, Email: hx_xu2015@163.com; Hanping Shi, Email: shihp@ccmu.edu.cn
Corresponding authors: Liangyu Yin, Emails: liangyuyin1988@tmmu.edu.cn, liangyuyin1988@qq.com; Hongxia Xu, Email: hx_xu2015@163.com; Hanping Shi, Email: shihp@ccmu.edu.cn
Get access Rights & Permissions [Opens in a new window]

Abstract

Persistent malnutrition is associated with poor clinical outcomes in cancer. However, assessing its reversibility can be challenging. The present study aimed to utilise machine learning (ML) to predict reversible malnutrition (RM) in patients with cancer. A multicentre cohort study including hospitalised oncology patients. Malnutrition was diagnosed using an international consensus. RM was defined as a positive diagnosis of malnutrition upon patient admission which turned negative one month later. Time-series data on body weight and skeletal muscle were modelled using a long short-term memory architecture to predict RM. The model was named as WAL-net, and its performance, explainability, clinical relevance and generalisability were evaluated. We investigated 4254 patients with cancer-associated malnutrition (discovery set = 2977, test set = 1277). There were 2783 men and 1471 women (median age = 61 years). RM was identified in 754 (17·7 %) patients. RM/non-RM groups showed distinct patterns of weight and muscle dynamics, and RM was negatively correlated to the progressive stages of cancer cachexia (r = –0·340, P < 0·001). WAL-net was the state-of-the-art model among all ML algorithms evaluated, demonstrating favourable performance to predict RM in the test set (AUC = 0·924, 95 % CI = 0·904, 0·944) and an external validation set (n 798, AUC = 0·909, 95 % CI = 0·876, 0·943). Model-predicted RM using baseline information was associated with lower future risks of underweight, sarcopenia, performance status decline and progression of malnutrition (all P < 0·05). This study presents an explainable deep learning model, the WAL-net, for early identification of RM in patients with cancer. These findings might help the management of cancer-associated malnutrition to optimise patient outcomes in multidisciplinary cancer care.

Keywords

MalnutritionGLIMCancerMachine learningRecurrent neural network

Information

Type
Research Article
Information
British Journal of Nutrition , First View , pp. 1 - 18
Check for updates
Copyright
© The Author(s), 2025. Published by Cambridge University Press on behalf of The Nutrition Society

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Article purchase

Temporarily unavailable

References

Cederholm, T, Barazzoni, R, Austin, P, et al. (2017) ESPEN guidelines on definitions and terminology of clinical nutrition. Clin Nutr 36, 49–64.Google Scholar
Wang, Z, Barazzoni, R & Shi, H (2023) Medical nutrition is one basic human right. Precision Nutr 2, e00026.Google Scholar
Arends, J, Bachmann, P, Baracos, V, et al. (2017) ESPEN guidelines on nutrition in cancer patients. Clin Nutr 36, 11–48.Google Scholar
Yin, L, Song, C, Cui, J, et al. (2021) A fusion decision system to identify and grade malnutrition in cancer patients: machine learning reveals feasible workflow from representative real-world data. Clin Nutr 40, 4958–4970.Google Scholar
Yin, L, Lin, X, Liu, J, et al. (2021) Classification tree-based machine learning to visualize and validate a decision tool for identifying malnutrition in cancer patients. JPEN J Parenteral Enteral Nutr 45, 1736–1748.Google Scholar
Gioulbasanis, I, Baracos, VE, Giannousi, Z, et al. (2011) Baseline nutritional evaluation in metastatic lung cancer patients: mini nutritional assessment v. weight loss history. Ann Oncol: Offic J Eur Soc Med Oncol 22, 835–841.Google Scholar
Almasaudi, AS, McSorley, ST, Dolan, RD, et al. (2019) The relation between Malnutrition Universal Screening Tool (MUST), computed tomography-derived body composition, systemic inflammation, and clinical outcomes in patients undergoing surgery for colorectal cancer. Am J Clin Nutr 110, 1327–1334.Google Scholar
Hu, C, Barazzoni, R & Shi, H (2023) Nutritional care is the first-line therapy for many conditions. Precision Nutr 2, e00059.Google Scholar
Rosato, E, Gigante, A, Colalillo, A, et al. (2023) GLIM-diagnosed malnutrition predicts mortality and risk of hospitalization in systemic sclerosis: a retrospective study. Eur J Intern Med 117, 103–110.Google Scholar
Brown, D, Loeliger, J, Stewart, J, et al. (2023) Relationship between global leadership initiative on malnutrition (GLIM) defined malnutrition and survival, length of stay and post-operative complications in people with cancer: a systematic review. Clin Nutr 42, 255–268.Google Scholar
Arends, J, Baracos, V, Bertz, H, et al. (2017) ESPEN expert group recommendations for action against cancer-related malnutrition. Clin Nutr 36, 1187–1196.Google Scholar
Gyan, E, Raynard, B, Durand, JP, et al. (2018) Malnutrition in patients with cancer: comparison of perceptions by patients, relatives, and physicians-results of the NutriCancer2012 Study. JPEN J Parenteral Enteral Nutr 42, 255–260.Google Scholar
Attar, A, Malka, D, Sabate, JM, et al. (2012) Malnutrition is high and underestimated during chemotherapy in gastrointestinal cancer: an AGEO prospective cross-sectional multicenter study. Nutr Cancer 64, 535–542.Google Scholar
Hebuterne, X, Lemarie, E, Michallet, M, et al. (2014) Prevalence of malnutrition and current use of nutrition support in patients with cancer. JPEN J Parenteral Enteral Nutr 38, 196–204.Google Scholar
Kipouros, M, Vamvakari, K, Kalafati, IP, et al. (2023) The level of adherence to the ESPEN guidelines for energy and protein intake prospectively influences weight loss and nutritional status in patients with cancer. Nutrients 15, 4232.Google Scholar
Cederholm, T, Jensen, GL, Correia, M, et al. (2019) GLIM criteria for the diagnosis of malnutrition - a consensus report from the global clinical nutrition community. Clin Nutr 38, 1–9.Google Scholar
Murnane, LC, Forsyth, AK, Koukounaras, J, et al. (2023) Malnutrition defined by GLIM criteria identifies a higher incidence of malnutrition and is associated with pulmonary complications after oesophagogastric cancer surgery, compared to ICD-10-defined malnutrition. J Surg Oncol 128, 769–780.Google Scholar
Matsui, R, Rifu, K, Watanabe, J, et al. (2023) Impact of malnutrition as defined by the GLIM criteria on treatment outcomes in patients with cancer: a systematic review and meta-analysis. Clin Nutr 42, 615–624.Google Scholar
Kiss, N, Steer, B, de van der Schueren, M, et al. (2023) Machine learning models to predict outcomes at 30-days using Global Leadership Initiative on Malnutrition combinations with and without muscle mass in people with cancer. J Cachexia, Sarcopenia Muscle 14, 1815–1823.Google Scholar
Yin, L, Chong, F, Huo, Z, et al. (2022) GLIM-defined malnutrition and overall survival in cancer patients: a meta-analysis. JPEN J Parenteral Enteral Nutr 47, 207–219.Google Scholar
Yin, L, Lin, X, Li, N, et al. (2021) Evaluation of the global leadership initiative on malnutrition criteria using different muscle mass indices for diagnosing malnutrition and predicting survival in lung cancer patients. JPEN J Parenteral Enteral Nutr 45, 607–617.Google Scholar
Milanez, DSJ, Razzera, EL, Lima, J, et al. (2023) Feasibility and criterion validity of the GLIM criteria in the critically ill: a prospective cohort study. JPEN J Parenteral Enteral Nutr 47, 754–765.Google Scholar
Yin, L, Fan, Y, Lin, X, et al. (2021) Fat mass assessment using the triceps skinfold thickness enhances the prognostic value of the Global Leadership Initiative on Malnutrition criteria in patients with lung cancer. Br J Nutr 1–11.Google Scholar
Yin, L, Cheng, N, Chen, P, et al. (2021) Association of malnutrition, as defined by the PG-SGA, ESPEN 2015, and GLIM criteria, with complications in esophageal cancer patients After Esophagectomy. Front Nutr 8, 632546.Google Scholar
Peng, D, Zong, K, Yang, H, et al. (2022) Malnutrition diagnosed by the Global Leadership Initiative on Malnutrition criteria predicting survival and clinical outcomes of patients with cancer: a systematic review and meta-analysis. Front Nutr 9, 1053165.Google Scholar
Liu, C, Lu, Z, Li, Z, et al. (2021) Influence of malnutrition according to the GLIM criteria on the clinical outcomes of hospitalized patients with cancer. Front Nutr 8, 774636.Google Scholar
Barazzoni, R, Jensen, GL, Correia, M, et al. (2022) Guidance for assessment of the muscle mass phenotypic criterion for the Global Leadership Initiative on Malnutrition (GLIM) diagnosis of malnutrition. Clin Nutr 41, 1425–1433.Google Scholar
O’Donoghue, N, Shrotriya, S, Aktas, A, et al. (2019) Clinical significance of weight changes at diagnosis in solid tumours. Support Care Cancer: Offic J Multinational Assoc Support Care Cancer 27, 2725–2733.Google Scholar
Hsu, WH, Ko, AT, Weng, CS, et al. (2023) Explainable machine learning model for predicting skeletal muscle loss during surgery and adjuvant chemotherapy in ovarian cancer. J Cachexia, Sarcopenia Muscle 14, 2044–2053.Google Scholar
Xu, H, Song, C, Yin, L, et al. (2022) Extension protocol for the Investigation on Nutrition Status and Clinical Outcome of Patients with Common Cancers in China (INSCOC) study: 2021 update. Precision Nutr 1, e00014.Google Scholar
Kondrup, J, Allison, SP, Elia, M, et al. (2003) ESPEN guidelines for nutrition screening 2002. Clin Nutr 22, 415–421.Google Scholar
Ottery, FD (1994) Rethinking nutritional support of the cancer patient: the new field of nutritional oncology. Semin Oncol 21, 770–778.Google Scholar
Chen, C, Lu, FC & Department of Disease Control Ministry of Health PRC (2004) The guidelines for prevention and control of overweight and obesity in Chinese adults. Biomed Environ Sci: BES 17, 1–36.Google Scholar
Fearon, K, Strasser, F, Anker, SD, et al. (2011) Definition and classification of cancer cachexia: an international consensus. Lancet Oncol 12, 489–495.Google Scholar
Chen, LK, Woo, J, Assantachai, P, et al. (2020) Asian Working Group for Sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Directors Assoc 21, 300–307.e2.Google Scholar
Yin, L, Li, N, Lin, X, et al. (2025) Early identification of potentially reversible cancer cachexia using explainable machine learning driven by body weight dynamics: a multicenter cohort study. Am J Clin Nutr 121, 535–547.Google Scholar
Yin, L, Zhang, L, Li, L, et al. (2024) Exploring the optimal indicator of short-term peridiagnosis weight dynamics to predict cancer survival: a multicentre cohort study. J Cachexia, Sarcopenia Muscle 15, 1177–1186.Google Scholar
Yin, L, Liu, J, Liu, M, et al. (2023) Ensemble learning system to identify nutritional risk and malnutrition in cancer patients without weight loss information. Sci China Life Sci 66, 1200–1203.Google Scholar
Yin, L, Cui, J, Lin, X, et al. (2022) Identifying cancer cachexia in patients without weight loss information: machine learning approaches to address a real-world challenge. Am J Clin Nutr 116, 1229–1239.Google Scholar
Bouloubasi, Z, Karayiannis, D, Pafili, Z, et al. (2023) Re-assessing the role of peri-operative nutritional therapy in patients with pancreatic cancer undergoing surgery: a narrative review. Nutr Res Rev 1–10.Google Scholar
Yaceczko, S & Baltz, J (2023) Evaluation of nutrition components within prehabilitation programs in gastrointestinal cancers: is prehab worth the hype? Nutr Clin Pract: Offic Publ Am Soc Parenteral Enteral Nutr 39, 117–128.Google Scholar
Zhou, T, Yang, K, Thapa, S, et al. (2017) Differences in symptom burden among cancer patients with different stages of cachexia. J Pain Symptom Manage 53, 919–926.Google Scholar
Nakajima, N (2021) Differential diagnosis of cachexia and refractory cachexia and the impact of appropriate nutritional intervention for cachexia on survival in terminal cancer patients. Nutrients 13, 915.Google Scholar
McLuskie, A, Bowers, M, Bayly, J, et al. (2025) Nutritional interventions in randomised clinical trials for people with incurable solid cancer: a systematic review. Clin Nutr 44, 201–219.Google Scholar
Baguley, BJ, Edbrooke, L, Denehy, L, et al. (2024) A rapid review of nutrition and exercise approaches to managing unintentional weight loss, muscle loss, and malnutrition in cancer. The Oncol (Epublication ahead of print version 07 October 2024).Google Scholar
Muscaritoli, M, Imbimbo, G, Jager-Wittenaar, H, et al. (2023) Disease-related malnutrition with inflammation and cachexia. Clin Nutr 42, 1475–1479.Google Scholar
Penet, MF & Bhujwalla, ZM (2015) Cancer cachexia, recent advances, and future directions. Cancer J 21, 117–122.Google Scholar
Roeland, EJ, Bohlke, K, Baracos, VE, et al. (2020) Management of cancer cachexia: ASCO guideline. J Clin Oncol: Offic J Am Soc Clin Oncol 38, 2438–2453.Google Scholar
Evans, WJ, Morley, JE, Argiles, J, et al. (2008) Cachexia: a new definition. Clin Nutr 27, 793–799.Google Scholar
Argiles, JM, Lopez-Soriano, FJ, Toledo, M, et al. (2011) The cachexia score (CASCO): a new tool for staging cachectic cancer patients. J Cachexia, Sarcopenia Muscle 2, 87–93.Google Scholar
Schuetz, P, Fehr, R, Baechli, V, et al. (2019) Individualised nutritional support in medical inpatients at nutritional risk: a randomised clinical trial. Lancet 393, 2312–2321.Google Scholar
Wen, X, Wang, M, Jiang, CM, et al. (2011) Anthropometric equation for estimation of appendicular skeletal muscle mass in Chinese adults. Asia Pac J Clin Nutr 20, 551–556.Google Scholar
Supplementary material: File

Yin et al. supplementary material

Yin et al. supplementary material
Download Yin et al. supplementary material(File)
File 3.5 MB