Recent Posts
Connect with:
Tuesday / June 25.
HomeminewsLimited Data Hampers Machine Learning Surgical Insights

Limited Data Hampers Machine Learning Surgical Insights

As machine learning (ML) gains traction in healthcare, researchers are exploring its potential for predicting outcomes after major surgeries. However, a new study cautions that current ML models have constraints due to limited data sources. 

“Machine Learning has great potential for processing ‘big data’ and has proved its undeniable capability, although it is not free of issues,” warned Dr Reza Hashemi from Flinders University’s College of Science and Engineering. “The accuracy of predictive models is dependent on the quality of the data sources, and predictions may be significantly affected by the amount of data and the number of variables included.” 

Dr Hashemi added, “At present, predictive models developed for total hip reconstruction and total knee reconstruction are based mainly on patient-reported factors and imaging variables. Therefore, the output of Machine Learning models in this area needs to be interpreted carefully.” 

Joining Forces to Assess ML’s Role  

In a collaborative effort with the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR), Royal Adelaide Hospital, and the University of South Australia, Flinders researchers evaluated common ML techniques, data sources, applications, limitations, and prediction quality for post-op outcomes of hip and knee replacements. 

“The most widely used Machine Learning approach in medical sciences is ‘supervised learning’, which estimates the mapping function for new input data in order to predict categorised, real values, or time-to-event outputs,” explained Dr Khashayar Ghadirinejad, a co-author from Flinders University. 

While conventional statistical methods rely on predefined equations, ML leverages large datasets to identify relationships between variables autonomously. 

Overcoming Data Scarcity  

Despite ML’s potential, the study underscores the importance of data quality and quantity. “More improvements are needed in Machine Learning orthopaedic applications to translate research aims into clinical practices,” Dr Ghadirinejad stated. 

The researchers recommend evaluating ML models using randomised, real-world cohort studies and trials rather than solely relying on available data. This approach could enhance ML’s clinical utility for tasks like early detection of outlier prostheses based on national joint registry data. 

“Joint registries aim to reduce the revision rates of arthroplasty surgeries by early detection of outlier joint arthroplasty devices,” the authors noted, highlighting the Australian Joint Registry’s efforts to curb harm and costs from underperforming implants. 

Personalised Surgical Decision Support  

Looking ahead, the study authors propose developing ML-driven decision support systems to guide pre-surgical planning tailored to individual patients’ needs. “A future direction for Machine Learning in the domain of joint arthroplasty could be to develop decision-making support systems focused on pre-surgical predictions that enable surgeons to determine what is the best for their patients individually,” they suggest. 

As ML capabilities evolve alongside growing clinical datasets, its role in enhancing surgical care planning and outcomes shows promising potential, albeit with current constraints to address through ongoing research and real-world validation. 



Ghadirinejad K., Milimonfared R., Hashemi R., et al., Supervised machine learning for the prediction of post-operative clinical outcomes of hip and knee replacements: a review. ANZ Journal of Surgery.DOI:10.1111/ans.19003