Fair AI-powered orthopedic image segmentation: addressing bias and promoting equitable healthcare.

AI-powered segmentation of hip and knee bony anatomy has revolutionized orthopedics, transforming pre-operative planning and post-operative assessment. Despite the remarkable advancements in AI algorithms for medical imaging, the potential for biases inherent within these models remains largely unexplored. This study tackles these concerns by thoroughly re-examining AI-driven segmentation for hip and knee bony anatomy. While advanced imaging modalities like CT and MRI offer comprehensive views, plain radiographs (X-rays) predominate the standard initial clinical assessment due to their widespread availability, low cost, and rapid acquisition. Hence, we focused on plain radiographs to ensure the utilization of our contribution in diverse healthcare settings, including those with limited access to advanced imaging technologies. This work provides insights into the underlying causes of biases in AI-based knee and hip image segmentation through an extensive evaluation, presenting targeted mitigation strategies to alleviate biases related to sex, race, and age, using an automatic segmentation that is fair, impartial, and safe in the context of AI. Our contribution can enhance inclusivity, ethical practices, equity, and an unbiased healthcare environment with advanced clinical outcomes, aiding decision-making and osteoarthritis research. Furthermore, we have made all the codes and datasets publicly and freely accessible to promote open scientific research.

FLASH Effects Induced by Orthovoltage X-Rays.

PURPOSE: This work describes the first implementation and in vivo study of ultrahigh-dose-rate radiation (>37 Gy/s; FLASH) effects induced by kilovoltage (kV) x-ray from a rotating-anode x-ray source. METHODS AND MATERIALS: A high-capacity rotating-anode x-ray tube with an 80-kW generator was implemented for preclinical FLASH radiation research. A custom 3-dimensionally printed immobilization and positioning tool was developed for reproducible irradiation of a mouse hind limb. Calibrated Gafchromic (EBT3) film and thermoluminescent dosimeters (LiF:Mg,Ti) were used for in-phantom and in vivo dosimetry. Healthy FVB/N and FVBN/C57BL/6 outbred mice were irradiated on 1 hind leg to doses up to 43 Gy at FLASH (87 Gy/s) and conventional (CONV; <0.05 Gy/s) dose rates. The radiation doses were delivered using a single pulse with the widths up to 500 ms and 15 minutes at FLASH and CONV dose rates. Histologic assessment of radiation-induced skin damage was performed at 8 weeks posttreatment. Tumor growth suppression was assessed using a B16F10 flank tumor model in C57BL6J mice irradiated to 35 Gy at both FLASH and CONV dose rates. RESULTS: FLASH-irradiated mice experienced milder radiation-induced skin injuries than CONV-irradiated mice, visible by 4 weeks posttreatment. At 8 weeks posttreatment, normal tissue injury was significantly reduced in FLASH-irradiated animals compared with CONV-irradiated animals for histologic endpoints including inflammation, ulceration, hyperplasia, and fibrosis. No difference in tumor growth response was observed between FLASH and CONV irradiations at 35 Gy. The normal tissue sparing effects of FLASH irradiations were observed only for high-severity endpoint of ulceration at 43 Gy, which suggests the dependency of biologic endpoints to FLASH radiation dose. CONCLUSIONS: Rotating-anode x-ray sources can achieve FLASH dose rates in a single pulse with dosimetric properties suitable for small-animal experiments. We observed FLASH normal tissue sparing of radiation toxicities in mouse skin irradiated at 35 Gy with no sacrifice to tumor growth suppression. This study highlights an accessible new modality for laboratory study of the FLASH effect.

Tumor Treating Fields: At the Crossroads Between Physics and Biology for Cancer Treatment.

Despite extraordinary advances that have been achieved in the last few decades, cancer continues to represent a leading cause of mortality worldwide. Lethal cancer types ultimately become refractory to standard of care approaches; thus, novel effective treatment options are desperately needed. Tumor Treating Fields (TTFields) are an innovative non-invasive regional anti-mitotic treatment modality with minimal systemic toxicity. TTFields are low intensity (1-3 V/cm), intermediate frequency (100-300 kHz) alternating electric fields delivered to cancer cells. In patients, TTFields are applied using FDA-approved transducer arrays, orthogonally positioned on the area surrounding the tumor region, with side effects mostly limited to the skin. The precise molecular mechanism of the anti-tumor effects of TTFields is not well-understood, but preclinical research on TTFields suggests it may act during two phases of mitosis: at metaphase, by disrupting the formation of the mitotic spindle, and at cytokinesis, by dielectrophoretic dislocation of intracellular organelles leading to cell death. This review describes the mechanism of action of TTFields and provides an overview of the most important in vitro studies that investigate the disruptive effects of TTFields in different cancer cells, focusing mainly on anti-mitotic roles. Lastly, we summarize completed and ongoing TTFields clinical trials on a variety of solid tumors.