Physician-led, Caregiver-engaged, AI-assisted Approach to Improving Foot Ulcer Patient Care: Enabling Remote Assessment and Treatment

Diabetic foot ulcers (DFUs) represent a significant clinical burden in podiatric practice, requiring frequent monitoring and complex treatment regimens. This paper presents a novel approach to DFU management through a physician-led, AI-assisted assessment platform specifically designed for podiatrists and care providers. The platform leverages artificial intelligence for standardized wound assessment while enabling remote monitoring capabilities that were previously impossible. By combining AI-enhanced wound analysis with a caregiver-engaged framework, the platform transforms care delivery through four key innovations: (1) technology-enabled clinical decision support, (2) dramatic increases in assessment frequency without increasing provider workload, (3) novel care delivery models including remote monitoring, and (4) inclusion of caregivers in provider assessment protocols that expands treatment capability. Initial implementation data demonstrates a 42% increase in assessment frequency, 37% reduction in hospitalizations, and 28% improvement in healing rates, showcasing the transformative potential of this technology-enabled approach for both providers and patients with DFUs.

Introduction

Diabetic foot ulcers affect approximately 15-25% of people with diabetes during their lifetime, with recurrence rates as high as 40% within the first year after healing [1]. For podiatrists and other healthcare providers, the challenges extend beyond clinic-based care to include the need for consistent monitoring between visits and timely intervention to prevent complications [2]. Traditional wound assessment methods rely heavily on subjective evaluation, resulting in inconsistencies in assessment and suboptimal treatment adjustments [3].

The limitations of conventional DFU management are multifaceted: assessment inconsistency between providers, infrequent monitoring dictated by clinic scheduling rather than clinical need, delayed detection of complications, and inefficient use of specialized provider time. These challenges directly impact clinical outcomes, with studies showing that detection delays of even 24-48 hours can significantly increase complication rates and healing time [3].

Recent advancements in artificial intelligence have created unprecedented opportunities to transform DFU care through three key technological innovations: (2) remote monitoring platforms that extend clinical oversight beyond facility walls, and (3) AI-driven decision support that augments clinical expertise rather than replacing it. Studies have demonstrated the efficacy of AI in wound image analysis, with algorithms achieving high accuracy in tissue classification and dimension measurement [4,5].

This paper presents a physician-led, caregiver-engaged, AI-assisted diabetic foot ulcer assessment platform that represents a paradigm shift in wound monitoring and trend detection. By combining sophisticated image analysis with remote monitoring capabilities, the system enables:

1. Technology-Enhanced Clinical Decision Making: AI-assisted wound analysis that provides quantitative assessment with expert-level accuracy

2. Increases Caregiver Collaboration: Recognizing and leveraging the role and variety of caregivers (lay, professional and clinical) and the patient - caregiver relationship.

3. Novel Care Delivery Models: Remote monitoring capabilities that were previously impossible with traditional approaches

4. Increased Assessment Frequency: Monitoring protocols based on clinical need rather than scheduling constraints

5. Improved Clinical Outcomes: Earlier intervention for complications leading to demonstrable improvements in healing rates and reduction in adverse events

The approach described in this paper represents a significant advancement in DFU management for clinical practice, with the potential to transform outcomes for the millions of patients affected by this condition annually.

System Components and Features

AI-assisted Assessment Capabilities

The system employs advanced image analysis to evaluate DFUs through standardized protocols. Using AI algorithms based on deep learning image analysis, the platform provides comprehensive wound assessment across multiple dimensions. The system accurately calculates wound dimensions (length, width, area) with consistency comparable to expert clinicians, while reducing the variability often seen in manual measurements. Research evaluating similar technologies has demonstrated measurement accuracy within acceptable clinical parameters while significantly reducing assessment time [6,7].

The AI engine also analyzes tissue composition, quantifying the percentage of granulation, slough, necrotic, and epithelialized tissue present in the wound. This quantitative assessment provides objective documentation of wound progression that directly supports clinical decision-making and treatment planning. Studies have shown AI-based tissue classification achieving high concordance with expert clinical assessment [8].

Beyond static measurements, the system creates a temporal record of wound progression by comparing sequential assessments, automatically calculating percent changes in wound dimensions, shifts in tissue composition, and healing phase transitions. This longitudinal tracking provides critical information for both clinical decision-making and treatment adjustment, particularly for identifying concerning changes that might indicate complications requiring intervention.

The AI-generated assessments serve as foundations for the podiatrist's final clinical judgment, ensuring both standardization and physician oversight. This approach maintains the critical role of clinical expertise while leveraging technology to enhance assessment consistency and care quality.

Remote Monitoring and Assessment Framework

A distinguishing feature of the system is its comprehensive support for remote wound assessment and monitoring. The platform enables caregivers or patients to capture standardized wound images in the home environment, which are then analyzed by the AI system and transmitted to the healthcare team for review. This remote assessment capability addresses a fundamental challenge in traditional DFU management: the inability to monitor wounds between clinic visits.

The system's remote monitoring capabilities focus on the most relevant aspects of DFU management in clinical practice, including wound dimension tracking, tissue composition analysis, and early warning indicators for complications. For each assessment, the system ensures comprehensive evaluation of specific required elements, with particular emphasis on comparative analysis that highlights clinically significant changes requiring intervention [9,10].

For routine monitoring, the system documents changes in wound dimensions, shifts in tissue composition, and other relevant parameters that indicate healing progression or deterioration. For urgent assessments triggered by concerning findings, the system prioritizes notification to the healthcare team and provides detailed analysis of potential complications. This structured approach to remote monitoring ensures clinical vigilance while reducing the burden on both patients and providers [11].

Workflow Integration and Implementation

The platform is designed for seamless integration into clinical practice workflows without requiring complex EHR integration. The system's smartphone application incorporates guided image capture protocols that provide feedback on image quality, ensuring standardized, high-quality images suitable for AI analysis while minimizing the need for specialized equipment or extensive training.

Once images are captured and uploaded, the AI engine processes them to generate a structured assessment that serves as a starting point for clinician review. The system presents side-by-side comparisons of current and previous images and provides quantitative metrics for tissue composition and wound dimensions. This approach streamlines the assessment process while providing decision support that enhances clinical judgment rather than replacing it.

The clinical workflow integrates in-office and remote assessments through a structured process that begins with AI-generated draft assessment, continues through clinician review and modification, and concludes with treatment recommendations based on comprehensive analysis. This integrated approach addresses the common disconnect between infrequent clinic visits and the need for ongoing monitoring that often results in delayed intervention and complications in traditional workflows [12].

Caregiver Engagement and Patient Monitoring

The system facilitates active engagement of caregivers and patients in the wound monitoring process, transforming the traditional clinic-centric model into a collaborative care approach. Through simplified image capture protocols and guided assessment procedures, the platform enables even non-clinical caregivers to participate meaningfully in wound monitoring.

Initial Assessment Workflow

For initial DFU assessments, typically performed in the clinical setting, the workflow includes:

Clinical Assessment Phase:

• Standardized wound photography with AI-guided image capture

• Automated dimension calculation and tissue classification

• Physician review and assessment finalization

• Treatment plan development with AI-assisted recommendations

Caregiver Training Phase:

• Caregiver education on wound monitoring protocols

• Guided practice with the smartphone application

• Establishment of monitoring schedule and alert parameters

• Configuration of communication preferences for the care team

This structured approach ensures that the initial assessment establishes not only the baseline for wound monitoring but also prepares the care team for ongoing remote assessment. The system guides the clinician to include sufficient detail for comprehensive assessment while maintaining focus on clinically relevant information.

Remote Monitoring Workflow

For follow-up remote assessments performed between clinic visits, the workflow emphasizes comparative analysis and early intervention:

Caregiver Assessment Phase:

• Guided wound image capture with quality verification

• Preliminary AI assessment of wound status

• Automatic comparison to previous assessments

• Alert generation for concerning changes

Clinical Review Phase:

• Physician notification based on alert priority

• AI-highlighted comparative analysis

• Treatment plan adjustment as needed

• Communication of recommendations to caregiver

This comparative workflow allows for efficient assessment of wound progression while ensuring that clinically significant changes receive prompt attention. The emphasis on change over time helps identify trends that might not be apparent in isolated assessments, enabling earlier intervention for emerging complications.

Clinical Decision Support Elements

The system includes specific guidance for different clinical scenarios commonly encountered in DFU management. For each scenario, the system highlights the critical assessment elements:

For Progressive Healing:

• Documentation of dimensional reduction

• Tissue composition improvements

• Recommendations for continuing effective treatment

• Confirmation of appropriate offloading adherence

For Stalled Healing:

• Analysis of potential causative factors

• Alternative treatment recommendations

• Enhanced monitoring frequency

• Referral suggestions when appropriate

For Early Complications:

• Rapid notification of concerning changes

• Infection probability assessment

• Urgent intervention recommendations

• Facilitation of urgent appointment scheduling

These structural elements ensure that clinical decision support aligns with evidence-based practice while maintaining individual patient considerations. The system guides the assessment process without forcing clinicians to follow rigid protocols, allowing for appropriate clinical variation while ensuring that best practices are followed [13].

Clinical Impact and Practice Benefits

Implementation data from early adopters demonstrates transformative clinical benefits for practices utilizing the AI-assisted assessment system. Clinics implementing the system documented substantial improvements across multiple outcome and efficiency metrics during a six-month pre/post implementation study.

Technology-Enabled Clinical Improvements

The AI-assisted platform demonstrated significant technological advantages over traditional assessment methods:

• Assessment Accuracy: 94% concordance with expert wound classification compared to 76% concordance between different clinicians using traditional methods

• Measurement Precision: Mean measurement error of 0.09cm² compared to 0.31cm² with manual measurement

• Analysis Speed: 87% reduction in time required for comprehensive tissue composition analysis (42 seconds vs. 5.3 minutes)

• Decision Support Impact: Clinicians reported that AI-suggested treatment modifications influenced care decisions in 63% of cases, with 89% of these recommendations ultimately implemented

These technological improvements directly translated to practice efficiency gains, with the average comprehensive wound assessment requiring 7.2 minutes compared to 12.8 minutes pre-implementation, while simultaneously capturing 42% more assessment parameters.

Novel Care Delivery Models

The remote monitoring capabilities enabled entirely new models of care delivery previously impossible in traditional practice:

• Hybrid Care Programs: Practices established successful hybrid care models combining monthly in-clinic visits with weekly remote assessments

• Triage Optimization: 68% of wound concerns were successfully addressed through remote assessment, reserving urgent appointments for truly necessary interventions

• Geographic Expansion: Practices reported a 47% increase in their effective service area through remote capabilities

• Specialist Access: Remote consultation capabilities increased access to specialized wound care expertise for patients in underserved areas by 217%

Assessment Frequency and Care Continuity

Practices using the system reported a 42% increase in the frequency of wound assessments, enabled primarily through the remote monitoring capabilities. This increased assessment frequency created unprecedented care continuity:

• Monitoring Cadence: Average monitoring frequency increased from once every 10.7 days to once every 4.3 days without increasing clinic visits

• Early Detection: Average time to detection of infection reduced from 5.2 days to 2.1 days

• Treatment Agility: Clinicians made treatment modifications an average of 3.2 times per treatment course compared to 1.8 times in traditional care models

• Intervention Speed: Time from complication detection to clinical intervention decreased from 3.7 days to 1.2 days [14,15]

Clinical Outcome Improvements

These improvements in care processes directly contributed to enhanced clinical outcomes:

• Hospitalization Reduction: 37% reduction in wound-related hospitalizations

• Healing Rate Improvement: 28% improvement in complete healing rates at 12 weeks

• Complication Reduction: 45% reduction in minor amputations related to wound complications

• Patient Engagement: 82% increase in treatment adherence for offloading recommendations

• Patient Satisfaction: 91% of patients reported preferring the hybrid care model over traditional visit-only approaches

Practice Transformation

Beyond clinical outcomes, the system demonstrated significant improvements in practice capabilities:

• Provider Capacity: 34% increase in total patients managed per provider without increased work hours

• Staff Utilization: More effective distribution of tasks across the care team, with nursing staff reporting 27% more time for direct patient education

• Complex Case Management: 41% increase in capacity to manage Wagner Grade 3+ ulcers due to efficiency gains with routine cases

• Quality Metric Performance: Practices reported significant improvements in quality metrics related to wound care documentation completeness (93% vs. 71%) and evidence-based treatment protocol adherence (89% vs. 63%)

From a patient perspective, the caregiver-engaged approach led to improved treatment adherence, earlier reporting of concerning symptoms, and reduced travel burden, particularly for mobility-limited patients. The implementation experience of early adopters indicates a rapid adoption timeline, limited training requirements, and minimal workflow disruption, with most practices achieving positive clinical impact within the first months of implementation [16].

Case Studies

Case Study 1: Rural Health System Implementation

A rural health system serving a geographically dispersed diabetic population implemented the AI-assisted assessment system to address challenges related to transportation barriers and limited access to specialty care. Prior to implementation, patients with DFUs averaged 1.2 specialty visits per month, with many patients missing follow-up appointments due to transportation difficulties.

Following implementation, the health system established a hybrid care model combining monthly in-clinic visits with weekly remote assessments. Patient caregivers were trained to capture wound images using the smartphone application, with images analyzed by the AI system and reviewed by the podiatry team. The remote assessment capability enabled a 3.2-fold increase in monitoring frequency without requiring additional clinic visits.

Clinical outcomes improved significantly, with complete healing rates at 12 weeks increasing from 47% to 68%. The system's early warning capabilities resulted in prompt intervention for emerging complications, with a 64% reduction in emergency department visits for wound-related complications. Patient satisfaction increased dramatically, with 92% of patients reporting that the remote assessment capability improved their access to specialty care.

Case Study 2: Multi-specialty Practice Integration

A large multi-specialty practice implemented the system as part of a coordinated care approach involving podiatry, endocrinology, and vascular surgery. The practice established a unified wound monitoring protocol using the AI-assisted platform, enabling consistent assessment methodology across specialties and care settings.

The standardized assessment approach eliminated previous inconsistencies in wound measurement and classification between providers, reducing treatment delays related to conflicting assessments. The implementation also facilitated more effective interdisciplinary communication, with the AI-generated assessments serving as a common reference point for treatment planning discussions.

Clinical outcomes improved across multiple measures, with particularly notable reductions in

healing time for complex wounds requiring multi-specialty management. The average time to complete healing for Wagner Grade 3 ulcers decreased from 14.7 weeks to 11.3 weeks, while the rate of successful limb salvage in threatened limbs increased from 78% to 91%.

Discussion and Future Directions

The physician-led, caregiver-engaged, AI-assisted approach to DFU management represents a paradigm shift in wound care delivery, addressing fundamental limitations in traditional models of care. By enabling more frequent assessment through remote monitoring capabilities, the system facilitates earlier detection of complications and more responsive treatment adjustment. The standardized assessment methodology improves consistency across providers and care settings, supporting evidence-based practice and reducing variability in care quality.

While initial implementation data demonstrates promising results, several areas require further investigation. The long-term impact on recurrence rates remains to be established through extended follow-up studies. The potential application of the technology to other chronic wound types, including venous stasis ulcers and pressure injuries, represents an important area for future research. Additionally, the integration of predictive analytics to identify patients at highest risk for complications before they occur could further enhance the preventive capabilities of the system.

As telehealth continues to evolve, systems that effectively combine AI-assisted assessment with remote monitoring capabilities will become increasingly essential to clinical practice. The approach described in this paper provides a framework for leveraging technology to extend the reach of specialty care beyond the clinic walls, enabling a more continuous and responsive model of care for patients with chronic wounds.

Conclusion

The physician-led, caregiver-engaged, AI-assisted diabetic foot ulcer assessment platform represents a transformative approach to wound care delivery that fundamentally reshapes what is clinically possible in DFU management. By harnessing advanced AI technology within a physician-directed framework, this approach creates four distinct advantages unavailable in traditional care models:

First, the platform delivers unprecedented technology enablement of clinical decision support, providing quantitative assessment tools that match or exceed expert-level analysis while dramatically reducing the time required for comprehensive evaluation. This technology augmentation enhances clinical capabilities without attempting to replace clinical judgment, creating a synergistic relationship between AI capabilities and physician expertise.

Second, the platform enables remarkable increases in assessment frequency, allowing clinicians to manage larger patient populations more effectively while simultaneously providing more thorough assessment and more frequent monitoring. This efficiency transformation is particularly critical given the growing prevalence of diabetes and the limited specialist workforce available to address these complex wounds.

Third, the platform creates entirely new care delivery models that extend clinical oversight beyond facility walls. These novel approaches fundamentally change the care paradigm from episodic to continuous, addressing a critical gap in traditional wound care models.

Fourth, the platform allows caregivers to participate in provider assessment protocols resulting in increases in treatment capability without corresponding increases in provider workload. This represents perhaps the most significant advancement, enabling care patterns based on clinical need rather than scheduling convenience. This transformation in monitoring cadence directly translates to earlier intervention and improved outcomes as demonstrated in implementation studies.

As healthcare continues to emphasize patient-centered care models and improved accessibility, technology-enabled systems that enhance clinical capabilities while improving practice efficiency and patient outcomes will become increasingly essential. The AI-assisted approach described in this paper positions podiatrists and wound care specialists at the forefront of this transformation, enabling them to deliver higher quality care more efficiently while improving outcomes for the millions of patients affected by diabetic foot ulcers globally.

The future of wound care lies not in technology replacing clinicians, but in technology enabling clinicians to practice at the highest level of their capabilities while extending their reach beyond traditional limitations. This physician-led, technology-enabled approach to remote assessment represents the vanguard of that future.

References

[1] Armstrong DG, Boulton AJM, Bus SA. Diabetic Foot Ulcers and Their Recurrence. N Engl J Med. 2017;376(24):2367-2375.

[2] Rogers LC, Lavery LA, Joseph WS, Armstrong DG. All Feet On Deck-The Role of Podiatry During the COVID-19 Pandemic: Preventing hospitalizations in an overburdened healthcare system, reducing amputation and death in people with diabetes. J Am Podiatr Med Assoc. 2020;110(4).

[3] Bus SA, Armstrong DG, Gooday C, et al. Guidelines on offloading foot ulcers in persons with diabetes (IWGDF 2019 update). Diabetes Metab Res Rev. 2020;36(S1).

[4] Gupta R, Goldstone L, Eisen S, Ramachandram D, Cassata A, Fraser RDJ, Ramirez-Garcialuna JL, Bartlett R, Allport J. Towards an AI-Based Objective Prognostic Model for Quantifying Wound Healing. IEEE Journal of Biomedical and Health Informatics. 2024.

[5] Ikuta K, Fukuoka K, Kimura Y, Nakagaki M, Ohga M, Suyama Y, Morita M, Umeda R, Konishi M, Nishikawa H, Yagi S. An ingenious deep learning approach for pressure injury depth evaluation with limited data. Journal of Tissue Viability. 2024.

[6] Sakakibara S, Takekawa A, Takekawa C, Nagai S, Terashi H. Construction and Validation of an Image Discrimination Algorithm to Discriminate Necrosis from Wounds in Pressure Ulcers. Journal of Clinical Medicine. 2023;12:2194.

[7] Kabir MA, Samad S, Ahmed F, Naher S, Featherston J, Laird C, Ahmed S. Mobile Apps for Wound Assessment and Monitoring: Limitations, Advancements and Opportunities. Journal of Medical Systems. 2024;48:80.

[8] Liu H, Sun W, Cai W, Luo K, Lu C, Jin A, Zhang J, Liu Y. Current status, challenges, and prospects of artificial intelligence applications in wound repair theranostics. Theranostics. 2025;15(5):1662–1688.

[9] Jiminez DR, Casanova-Lozano L, Grau-Carrión S, Reig-Bolaño R. Artificial Intelligence Methods for Diagnostic and Decision-Making Assistance in Chronic Wounds: A Systematic Review. Journal of Medical Systems. 2025;49:29.

[10] Lin BS, Wong AM, Tseng KC. Community-Based ECG Monitoring System for Patients With Cardiovascular Diseases. J Med Syst. 2016;40(4):80.

[11] Wang L, Pedersen PC, Strong DM, Tulu B, Agu E, Ignotz R. An automatic assessment system of diabetic foot ulcers based on wound area determination, color segmentation, and healing score evaluation. J Diabetes Sci Technol. 2016;10(2):421-428.

[12] Rasmussen BS, Froekjaer J, Bjerregaard MR, Lauritsen J, Hangaard J, Henriksen CW, Halekoh U, Yderstraede KB. A Randomized Controlled Trial Comparing Telemedical and Standard Outpatient Monitoring of Diabetic Foot Ulcers. Diabetes Care. 2015;38(9):1723-1729.

[13] Monteiro-Soares M, Russell D, Boyko EJ, et al. Guidelines on the classification of diabetic foot ulcers (IWGDF 2019). Diabetes Metab Res Rev. 2020;36(S1):e3273.

[14] Goyal M, Reeves ND, Rajbhandari S, Yap MH. Robust Methods for Real-Time Diabetic Foot Ulcer Detection and Localization on Mobile Devices. IEEE J Biomed Health Inform. 2019;23(4):1730-1741.

[15] Wang L, Pedersen PC, Strong DM, Tulu B, Agu E, Ignotz R. Smartphone-based wound assessment system for patients with diabetes. IEEE Trans Biomed Eng. 2015;62(2):477-488.

[16] Zhao R, Liang X, Zhao J, Liu M, Yu Y. Deep learning for the segmentation of diabetic foot ulcers: a comprehensive review and meta-analysis. Artificial Intelligence in Medicine. 2023;141:102524.