Background: Raman Spectroscopy is a non-invasive technique capable of characterising tissue constituents and detecting conditions such as cancer with high accuracy. Machine learning techniques can automate this task and discover relevant data patterns. However, the high-dimensional, multicollinear nature of Raman data makes their deployment and explainability challenging. A model’s transparency and ability to explain decision pathways have become crucial for medical integration. Consequently, an effective method of feature-reduction while minimising information loss is sought. Methods: Two new feature selection methods for Raman spectroscopy are introduced. These methods are based on explainable deep learning approaches, considering Convolutional Neural Networks and Transformers. Their features are extracted using GradCam and attention scores, respectively. The performance of the extracted features is compared to established feature selection approaches across four classifiers and three datasets. Results: We compared the proposed method against established feature selection approaches over three real-world datasets and different compression levels. Comparable accuracy levels were obtained using only 10% of features. Model-based approaches are the most accurate. Using Convolutional Neural Networks and Random Forest-assigned feature importance performs best when maintaining between 5–20% of features, while LinearSVC with L1 penalisation leads to higher accuracy when selecting only 1% of them. The proposed Convolutional Neural Networks-based GradCam approach has the highest average accuracy. Conclusions: No approach is found to perform best in all scenarios, suggesting that multiple alternatives should be assessed in each application.

Malignant diseases are the second most common cause of death worldwide.¹ The incidence rate is expected to increase by 47% by 2040, reflecting both growth and aging of the population as well as changes in other risk factors associated with socioeconomic development.² In the combat of malignant diseases, there are presently three dominating and well-developed treatment modalities: surgery, chemotherapy, and radiation therapy, whereas other modalities such as immunotherapy^3,4 are receiving increased attention.

In most developed countries, orthopedic surgery represents one of the most common surgical procedures.¹ The number of patients and procedures will only increase with aging populations. Many surgical procedures have been developed successfully to treat common degenerative conditions, including total knee or hip replacement and spinal fusion, which are the most common hospital discharge procedures in the United States with about 1.5 million cases per year. Other procedures, such as partial and total shoulder replacement, partial hip replacement, kyphoplasty/vertebroplasty, and lumbar decompression, are also regularly performed by orthopedic surgeons as well as other specialists (neurosurgeons and interventional radiologists). Minimally invasive surgery (MIS) and percutaneous techniques have gradually been introduced to reduce complications and improve recovery times:^{2 – 4} examples in spinal surgery include pedicle screw placement, spinal fusion, vertebroplasty/kyphoplasty, and decompression. There has been more recent adoption in total hip arthroplasty, with renewed interest in anterior hip approaches.^{5 – 7} Other hip arthroplasty techniques have included bone resurfacing, which has been controversial with several large registries highlighting concerns over premature wear and metallosis.^8,9

Surgical procedures involving osteotomies pose risks of breaching into critical neurovascular and musculoskeletal structures, often leading to operative morbidity or potential mortality. One related clinical situation involves total hip and knee arthroplasty (THA/TKA) where periprosthetic fracture is considered a major complication that requires revision. Studies found that 40% and 54% of primary THA and TKA patients, respectively, encountered at least one major complication, such as fracture or one minor complication, such as neurovasculature damage, or both.^1,2 Furthermore, with population ageing leading to rapidly growing incidence of primary arthroplasty, the number of revision surgeries will continue to increase substantially due to limited implant survivorship. The advanced surgical complexity and higher rates of associated complications have become a prevalent concern that remains to be improved. Within orthopedics, incidence projections of revision THA/TKA (rTHA/rTKA) to 2030 in the United States alone have indicated an increase of 43% to 70% and 78% to 182%, respectively, relative to the 50,000 rTHA and 72,000 rTKA cases in 2014. In particular, the patient group aged $>55$ years is anticipated to undergo the highest increase, suggesting that multiple revision surgeries would become necessary in the coming decades. Some common causes for rTHA include instability (28%), aseptic loosening (24%), periprosthetic fracture (18%), and prosthetic joint infection (18%).³

Transabdominal fetal pulse oximetry offers a promising approach to improve fetal monitoring and reduce unnecessary interventions. Utilizing realistic 3D geometries derived from MRI scans of pregnant women, we conducted photon simulations to determine optimal source-detector configurations for detecting fetal heart rate and oxygenation. Our findings demonstrate the theoretical feasibility of measuring fetal signals at depths up to 30 mm using source-detector (SD) distances greater than 100 mm and wavelengths between 730 and 850 nm. Furthermore, we highlight the importance of customizing SD configurations based on fetal position and maternal anatomy. These insights pave the way for enhanced non-invasive fetal monitoring in clinical application.