Researchers in Australia have developed an artificial intelligence model capable of reconstructing detailed plantar pressure maps using foot geometry and only a small number of pressure measurements, potentially opening the way to cheaper and more accessible foot-health assessment technology.
The research, led by The University of Queensland in collaboration with iOrthotics and Healthia Limited, could have particular relevance for diabetic-foot screening, orthotic prescription and gait assessment in regions where conventional plantar-pressure systems remain too expensive or difficult to deploy.
According to Medical Xpress, the system uses a multimodal deep-learning model to predict dense pressure information from anatomical foot data and sparse sensor inputs.
The work has been published in the journal Sensors under the title “Multimodal Feature-Level Fusion CBAM U-Net for Static Plantar Pressure Prediction Using Plantar Geometry and Sparse Anatomical Landmarks.”
Reconstructing pressure maps with fewer sensors
Plantar-pressure assessment is widely used by podiatrists, orthotists and other foot-health professionals to understand how forces are distributed underneath the foot.
The information can help clinicians assess gait, balance and foot function, identify high-pressure areas and inform the design or prescription of foot orthoses.
However, conventional pressure platforms and in-shoe systems can be costly.
In-shoe monitoring systems may also require large numbers of individual sensors, increasing complexity, power consumption and manufacturing cost.
The University of Queensland research examined whether artificial intelligence could reduce that dependence on dense sensor arrays.
Using foot geometry and plantar-pressure measurements collected from 35 study participants, doctoral researcher Chongguang Wang developed an artificial neural network capable of generating high-resolution plantar-pressure maps from a much smaller number of physical measurements.
The model achieved its strongest results using 16 anatomical landmarks on the plantar surface, but importantly, researchers reported comparable reconstruction performance using only two pressure landmarks.
That finding could be particularly significant for future low-cost monitoring devices.
Potential for lower-cost pressure technology
Professor Martin Veidt, emeritus professor at The University of Queensland and an applied mechanics engineer involved in the research, said existing pressure measurement technologies can be difficult to access, particularly for people living in rural and remote areas.
The research suggests that detailed plantar-pressure information may eventually be generated by combining a relatively simple representation of foot shape with only a limited number of physical pressure measurements.
If successfully translated into commercially viable devices, this could reduce the hardware requirements associated with pressure assessment.
Rather than relying on a mat or insole containing a dense array of pressure sensors, future devices could potentially use a smaller number of strategically positioned sensors while AI reconstructs the wider pressure distribution.
For the P&O sector, this could make pressure measurement practical in settings where a full laboratory or sophisticated gait-analysis infrastructure is unavailable.
Relevance to orthotic prescription
Pressure analysis has become increasingly integrated into digital foot-orthotic workflows.
Clinicians can combine pressure information with 3D foot scans, clinical examination and gait assessment to identify regions requiring accommodation, redistribution or support.
For patients with pain or biomechanical problems, this information may influence shell geometry, material selection, cushioning and local offloading within a foot orthosis.
The Australian research could therefore contribute to a wider move towards AI-assisted orthotic assessment.
Instead of AI replacing the clinician, the technology could help produce additional objective data at lower cost, allowing orthotists and podiatrists to make more informed decisions.
This could be especially useful in community clinics, mobile rehabilitation programmes and smaller P&O facilities that cannot justify the cost of high-end pressure measurement equipment.
Important implications for diabetic-foot prevention
One of the most significant potential applications is diabetic-foot care.
Elevated plantar pressure, peripheral neuropathy, deformity and poor tissue health can combine to increase the risk of diabetic foot ulceration.
Identifying areas of excessive loading can help clinicians prescribe footwear, insoles and offloading interventions intended to reduce tissue stress.
The University of Queensland team specifically highlighted diabetic foot ulcers and amputations as major health-system challenges that improved monitoring technologies could help address.
The issue is particularly relevant across the IMEA region.
Many Middle Eastern countries have high rates of diabetes, while large parts of Africa and South Asia face a combination of increasing diabetes prevalence and limited access to specialist podiatry and diabetic-foot services.
In these environments, a lower-cost method of identifying abnormal plantar loading could potentially expand screening beyond specialist diabetic-foot centres.
Earlier detection could reduce downstream burden
The economic argument is also important.
Diabetic foot complications frequently result in repeated hospital admissions, infection, surgery and ultimately minor or major amputation.
A separate 2026 analysis of 500 patients admitted with diabetic-foot complications between 2015 and 2024 found persistent rates of severe disease, repeated admissions and major amputations, reinforcing the importance of earlier screening and preventive intervention.
Technologies that make foot assessment easier to deliver in primary care or community settings could therefore have value well beyond the P&O clinic.
If patients at elevated risk can be identified earlier, they may be referred for footwear modification, orthotic intervention, wound-prevention programmes or specialist diabetic-foot management before ulceration becomes advanced.
Potential role in remote and rural healthcare
One of the most interesting aspects of the research is its possible application away from specialist centres.
Veidt said the approach could eventually allow data collection to take place in isolated communities where health services and outcomes may be limited.
For many countries across Africa, Central Asia and South Asia, geographic accessibility remains one of the major barriers to rehabilitation and foot-health services.
Patients may need to travel significant distances to reach a hospital or prosthetic and orthotic centre equipped with specialist assessment technology.
Simplified sensing technology combined with AI processing could potentially allow basic plantar-pressure data to be collected in satellite clinics or community programmes and reviewed remotely by specialist clinicians.
The same principle could fit within emerging tele-podiatry and digital orthotics models, where assessment data is captured locally while clinical interpretation or device design takes place elsewhere.
Part of a broader smart-orthotics programme
The research forms part of a wider programme being developed by iOrthotics, Healthia and academic partners, investigating how new technologies could make orthotic and foot-health services more accessible.
Healthia Group Chief Education and Research Officer Kerrie Evans said the aim is to develop practical and affordable technologies capable of helping clinicians understand foot function and identify potential problems earlier.
The study remains an early-stage technology demonstration.
Researchers emphasised that additional work is required, particularly testing involving clinical populations and deployment in real-world healthcare environments.
That validation will be particularly important before AI-generated pressure predictions can be relied upon in patients with diabetes, neuropathy, significant deformity or other complex foot conditions.
From measuring pressure to predicting risk
The wider opportunity for the orthotics industry may eventually go beyond reconstructing a pressure map.
As datasets become larger, AI systems could potentially combine plantar pressure with 3D geometry, patient history, diabetes status, neuropathy assessment and previous ulceration to support clinical risk stratification.
The emerging trend is therefore not simply toward cheaper pressure platforms, but towards smarter foot-health assessment systems.
For the IMEA P&O sector, where both advanced digital orthotic laboratories and severely underserved communities exist within the same region, that distinction matters.
High-end systems will remain valuable for specialist clinics, research and complex biomechanical assessment.
But technologies capable of generating clinically useful information using inexpensive hardware could help extend foot-health services to a far wider population.
If the University of Queensland approach can be successfully translated from research into robust clinical technology, AI may help turn plantar-pressure assessment from a specialist laboratory tool into something that can be deployed much closer to the patient.
- Medical Xpress – Study Shows AI Could Support Low-Cost Foot Health Technology
- Sensors – Multimodal Feature-Level Fusion CBAM U-Net for Static Plantar Pressure Prediction
- The University of Queensland
- iOrthotics
- Healthia
- Frontiers in Public Health – Hospital Burden, Amputation Risk and Mortality Trends in Diabetic Foot Patients