{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/a-probabilistic-disease-progression-model-for","title":"A Probabilistic Disease Progression Model for Predicting Future Clinical Outcome","arxiv_id":"1803.05011","date":"2018-03-13","proceeding":null,"authors":["Yingying Zhu","Mert R. Sabuncu"],"abstract":"In this work, we consider the problem of predicting the course of a\nprogressive disease, such as cancer or Alzheimer's. Progressive diseases often\nstart with mild symptoms that might precede a diagnosis, and each patient\nfollows their own trajectory. Patient trajectories exhibit wild variability,\nwhich can be associated with many factors such as genotype, age, or sex. An\nadditional layer of complexity is that, in real life, the amount and type of\ndata available for each patient can differ significantly. For example, for one\npatient we might have no prior history, whereas for another patient we might\nhave detailed clinical assessments obtained at multiple prior time-points. This\npaper presents a probabilistic model that can handle multiple modalities\n(including images and clinical assessments) and variable patient histories with\nirregular timings and missing entries, to predict clinical scores at future\ntime-points. We use a sigmoidal function to model latent disease progression,\nwhich gives rise to clinical observations in our generative model. We\nimplemented an approximate Bayesian inference strategy on the proposed model to\nestimate the parameters on data from a large population of subjects.\nFurthermore, the Bayesian framework enables the model to automatically\nfine-tune its predictions based on historical observations that might be\navailable on the test subject. We applied our method to a longitudinal\nAlzheimer's disease dataset with more than 3000 subjects [23] and present a\ndetailed empirical analysis of prediction performance under different\nscenarios, with comparisons against several benchmarks. We also demonstrate how\nthe proposed model can be interrogated to glean insights about temporal\ndynamics in Alzheimer's disease.","url_abs":"http://arxiv.org/abs/1803.05011v1","url_pdf":"http://arxiv.org/pdf/1803.05011v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"a-probabilistic-disease-progression-model-for","repo_url":"https://github.com/zyy123jy/kdd","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"bayesian-inference","task_name":"Bayesian Inference"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}