Precision Imaging to Evaluate Kaposi Sarcoma (PRIME-KS): protocol for a multicountry novel artificial...

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence

Background:

• Kaposi sarcoma (KS) is a prevalent cancer in Eastern Africa, with current treatment monitoring relying on imprecise ruler-based measurements.
• Standard methods exhibit inter-observer variability, fail to capture lesion dimensions accurately, and perform poorly on darker skin tones.
• SkinScan3D (SS3D) is a novel, portable, AI-enabled 3D imaging device designed for objective KS skin lesion assessment.

Purpose of the Study:

• To evaluate if SS3D provides more reproducible and accurate lesion measurements compared to the standard ruler-based method.
• To validate the integration of SS3D into routine clinical workflows in Kenya and Uganda.
• To assess the clinical utility of SS3D for monitoring KS treatment response.

Main Methods:

• A prospective mixed-methods study (PRIME-KS) involving diagnostic accuracy and pilot studies.
• Objective 1: Cross-sectional study comparing SS3D and ruler measurements in 50 adults with KS (150 lesions) across Kenya and Uganda.
• Objective 2: Pilot study in 100 patients evaluating SS3D usability, acceptability, and feasibility, alongside time-and-motion and micro-costing studies.

Main Results:

• The study will provide the first rigorous evaluation of a 3D imaging device for KS treatment monitoring in real-world clinical settings.
• Primary outcomes include concordance correlation coefficient for inter-rater reproducibility and coefficient of determination for accuracy.
• Formative research incorporated end-user input to refine the SS3D device and protocols.

Conclusions:

• SS3D has the potential to improve the objectivity and reproducibility of KS lesion measurements.
• If validated, SS3D could lead to earlier and more accurate assessment of treatment response.
• This could reduce unnecessary chemotherapy exposure and associated toxicities in KS patients.