Max Gordon

Max Gordon

Adjunct Senior Lecturer | Docent
Visiting address: Entrévägen 2, 18257 Danderyd
Postal address: D1 Kliniska vetenskaper, Danderyds sjukhus, D1 Ortopedi, 182 88 Stockholm
Part of:

About me

  • I am a docent and an orthopaedic surgeon at Danderyd University Hospital. In
    addition to my clinical profile I have a background in programming. I've been
    toying around with different projects in a multitude of languages since mid
    90:s where the statistical programming language R lies very close to heart as
    I've been in the Stockholm R-User Group [1] since 2013 and have 4 published
    R-packages. My most downloaded package, htmlTable [2], is downloaded more
    than 100 000 times/month. You can find all of my open-source contributions on
    GitHub [3]. I also have a research blog, G-forge [4], where I try to post
    random things that I find interesting whenever I find the time.
    [1] http://www.meetup.com/StockholmR/
    [2] https://github.com/gforge/htmlTable
    [3] https://github.com/gforge/
    [4] http://gforge.se

Research

  • My main research interests are currently epidemiology and machine learning
    where I'm one of the co-founders of teh Clinical Artificial Intelligence
    Laboratory [1] (CAIR-Lab). The lab is a unique environment for clinicians to
    develop clinical AI-driven applications. It is a meeting place for all
    medical fields with the sole purpose of taking AI-research all the way from
    idea to the patient.
    I wrote my thesis [2] in 2014 on how patient factors influence outcomes after
    total hip replacements. It was a wonderful collaboration with the Swedish Hip
    Arthroplasty Register [3] that work spurred my interest in advanced
    statistics and that then led me towards AI and deep learning. My current
    focus is primarily on interpreting radiographs using deep learning where try
    to find features beyond just fracture detection that can be used to guide
    clinicians and help patients.
    My primary motivator is to bridge the gap between research and implementation
    - too many good papers fail too reach the benefit of the patients. I am
    certain of that this is not a lack of will but a lack of proper tooling. AI
    has the potential to change this and I hope that my research will be a part
    of this change.
    [1] https://ki.se/en/kids/cair-lab
    [2] http://publications.ki.se/xmlui/handle/10616/42012
    [3] https://shpr.registercentrum.se/

Articles

All other publications

Grants

  • Swedish Research Council
    1 January 2023 - 31 December 2025
    Early diagnosis of cancer is key for improving patient outcomes, but cancers are hard to diagnose if patients present with unspecific symptoms. The principal objective of the MEDECA study is to identify a multi-analyte blood test that can detect and map cancer within a mixed population presenting with serious but unspecific symptoms. We are including 1.000 patients referred to the Diagnostic Center at Danderyd Hospital (DC DS), a multidisciplinary diagnostic center referral pathway. Blood samples are collected prior to a standardized and comprehensive cancer diagnostic work-up including an expanded panel of biochemical analyses and extensive imaging such as CT or MR investigations. In close international collaboration with world-leading scientists within the field of "liquid biopsy", the blood samples will be analyzed for a panel of novel and established blood biomarkers predicitve of an underlying cancer. These include markers of neutrophil extracellular traps, circulating tumor DNA, platelet mRNA profiling, affinity-based proteomics and nuclear resonance metabolomics. The diagnostic accuracy of the blood biomarkers with respect to cancer detection will be analyzed through machine learning, with the overarching goal to develop a multi-analyte blood test that can detect an occult cancer. This may lead to an earlier diagnosis, prompt cancer treatment and reduced morbidity and mortality. This would also lead to avoidance of unnecessary and costly investigations.
  • Swedish Research Council
    1 January 2022 - 31 December 2025
    Rational for conducting the study: The most commonly used treatment of displaced femoral neck fracture in the elderly is hemiarthroplasty which provides a reliable clinical result. Deep periprosthetic joint infection (PJI) is a severe complication associated with an increased morality and often leads to major and repeated reoperations. PJI has been reported to occur in 2-4% of hemiarthroplasties after fracture. The aim of the trial is to investigate whether the risk of PJI is reduced after the use of dual impregnated antibiotic laden bone cement.Study design: Register-based, cluster randomized controlled trial.Population: Number of patients: 7,312. Inclusion criteria: Patients aged ≥60 year, with a displaced femoral neck fracture and eligible for hemiarthroplasty according to local guidelines. Exclusion criteria: Previous inclusion of contralateral hip. Pathological or stress fracture of the femoral neck or fracture adjacent to a previous ipsilateral hip implant.Intervention: Dual-impregnated antibiotic laden bone cement.Control: Single-impregnated antibiotic laden bone cement.Outcome: Primary outcome variable is the incidence of PJI within one year. Secondary outcome variables include the occurrence of re-operations for any reason, mortality,and health care costs. Future benefit and clinical application: A reduction of PJI could decrease the morbidity and mortality of this elderly population and reduce the need for reoperations and antibiotic treatment.

Employments

  • Adjunct Senior Lecturer, Department of Clinical Sciences, Danderyd Hospital, Karolinska Institutet, 2024-2026

Degrees and Education

  • Docent, Karolinska Institutet, 2022
  • Degree Of Doctor Of Philosophy, Department of Clinical Sciences, Danderyd Hospital, Karolinska Institutet, 2014

News from KI

Events from KI