Background: Cachexia is a multifactorial syndrome characterized by weight and muscle loss, often linked to malnutrition and inflammation. Malnutrition affects almost 40 % of colorectal cancer (CRC) patients, contributing to worse surgical outcomes, higher morbidity, and increased mortality. This study evaluates the Cachexia Index (CXI) for malnutrition detection in CRC patients.
Patients and methods: Seventy patients (54.3 % men; median age 65) underwent bioelectrical impedance analysis, blood tests, and nutritional assessments before surgery. Patients were recruited between November 1, 2023, and October 30, 2024. CXI, calculated as [Skeletal Mass Index × Serum Albumin]/Neutrophil-to-Lymphocyte-Ratio, was compared with the Subjective Global Assessment.
Results: The median CXI was 13.7 (range: 1.9-53.8), with malnourished or at-risk patients exhibiting significantly lower median CXI values (9.5 vs. 14.9; p = 0.0262) than well-nourished individuals. Tumor location also influenced CXI; patients with sigmoid colon tumors had the highest median CXI, while those with caecum tumors had the lowest (22.2 vs. 8.2; p = 0.0202). CXI demonstrated 100 % sensitivity and 60 % specificity for malnutrition detection, with a cutoff of ≤12.09 (AUC = 0.80; 95 % CI: 0.69-0.89; p < 0.0001). Malnourished patients had a significantly lower median CXI (9.5 vs. 14.9; p = 0.0262).
Conclusions: These findings support CXI as a reliable, non-invasive biomarker for malnutrition in CRC patients, with potential applications in personalized nutritional and therapeutic care.
Background: Preoperative perforator mapping of the Profunda Artery Perforator (PAP) flap is hindered by challenges in aligning preoperative images with the intraoperative lithotomy position of the leg. Dynamic Infrared Thermography (DIRT) is a real-time, quick, and non-invasive imaging modality that is increasingly explored for preoperative perforator mapping. This feasibility study demonstrates the application of Projected Augmented Reality to project thermal images of DIRT directly on the skin for the identification of perforators before PAP flap breast reconstructions.
Methods: A portable self-aligning projection device (Anatomy Projector) was integrated with a thermal sensor to obtain thermal information and automatically project thermal images onto the patient's medial thigh before PAP flap dissection. Projected DIRT hotspots were evaluated with hand-held Doppler (HHD) and compared to locations of intraoperative perforators following a Cartesian coordinate system.
Results: Preoperative DIRT examination yielded a projection of 127 DIRT hotspots in 20 PAP flaps. All projected hotspots could be verified with Doppler (100 %) and 82.1 % of intraoperative perforators correlated with a projected DIRT hotspot within a 3 cm radius. Notably, 30.4 % of these matches involved the first appearing DIRT hotspot, and 82.6 % of these matches involved a DIRT hotspot within the first 5 appearing hotspots.
Conclusion: This feasibility study pioneers the use of Projected Augmented Reality for the display of thermal images of DIRT directly on the skin for the preoperative perforator identification before PAP flap harvest. High resemblance with HHD and intraoperative perforators was found. Future research should examine the further applications in the intraoperative and postoperative setting.

