AI-SWEEP aims to develop the use of artificial intelligence (AI) in wound care. Chronic wounds (e.g., pressure ulcers/decubitus, venous leg ulcers, and diabetic foot ulcers) are wounds that do not progress through the healing cascade in a timely and smooth manner, failing to achieve complete healing within 4 to 8 weeks, often due to stagnation in the inflammatory phase of healing [Sme21]. These chronic wounds can lead to prolonged hospital stays and additional treatments, often impacting health-related quality of life and posing a significant risk of severe infections with high mortality rates [Zar15].

Despite significant advancements in wound treatment, there has been little progress in objectively assessing wound diagnosis and treatment [Li20]. The prevalence of chronic wounds is estimated to be between 1% and 2% in high-income countries, with expectations that these numbers will rise due to the increasing incidence of comorbidities such as obesity, diabetes mellitus, and autoimmune diseases. Additionally, caring for patients with chronic wounds incurs substantial costs. In high-income countries, the costs of treating chronic wounds account for 2% to 4% of total healthcare expenditures, with the average annual treatment cost per chronic wound in Europe ranging from €6,000 to €10,000 [Sme21].

Currently, the evaluation of wound progression, wound infection, and healing timelines is primarily based on visual inspection of the wound and its surrounding tissue, leading to a purely subjective clinical assessment. Clinical symptoms, such as those associated with wound inflammation, may indicate delayed healing but are not consistently present. Wound cultures or diagnostic tests (ultrasound, X-rays, examinations under anesthesia, etc.) can be performed to identify the cause of delayed healing, but these methods are costly and results are not always available in a timely manner. At present, wound specialists have few alternative parameters available to objectively predict and anticipate healing outcomes.

Databases containing standardized wound parameters and AI-driven analyses could support more objective wound assessment, including the detection of wound progression and infection. AI-assisted wound care has the potential to reduce the workload of wound specialists, improve access to specialized medical expertise, and enhance the possibilities for digital monitoring of wound treatment and remote wound care. Specifically, AI-driven pattern recognition and algorithms could support and facilitate the determination of wound diagnosis, prognosis, and treatment using various wound parameters.

However, the quality of AI-driven outcomes is highly dependent on the availability of data—first to train these algorithms and later to implement them in clinical practice. Therefore, a range of wound parameters must be measured with sufficient accuracy. The causes of impaired wound healing are multifactorial and manifest through various abnormalities, including wound and surrounding skin temperature, pH, moisture levels, peripheral edema, vascularization, wound bed condition, presence of proteases (both human and bacterial), and oxygen tension.

This project aims to develop flexible or even stretchable digital smart dressings or e-patches with embedded microsensors that are relevant for the timely diagnosis of wound infection and impaired wound healing. These will be combined with advanced ultra-low-power microelectronics for real-time data acquisition, communication, and processing. This approach will provide AI-driven algorithms with essential information, making telemedicine-based patient monitoring a reality. Specific AI algorithms, in combination with data collected via smart dressings, can offer an efficient, minimally invasive, and patient-friendly solution to optimize and accelerate the wound healing process.